OUP user menu

Cardiovascular molecular imaging: an overview

Antti Saraste, Stephan G. Nekolla, Markus Schwaiger
DOI: http://dx.doi.org/10.1093/cvr/cvp209 643-652 First published online: 24 June 2009

Abstract

Molecular imaging is non-invasive visualization and measurement of biological processes at the molecular and cellular level within a living organism. This review provides a description of the various molecular imaging techniques for imaging cardiovascular targets and their potential clinical implications. Molecular imaging has relied mainly on nuclear imaging, but advances in nanoparticle probe development have made magnetic resonance imaging and ultrasound as emerging, radiation-free alternatives. Targeted imaging of vascular inflammation or thrombosis may allow improved risk assessment of atherosclerosis by detecting plaques at high risk of acute complications. Imaging probes detecting myocardial apoptosis, metabolic alterations, injury to extracellular matrix, angiogenesis, or innervation may provide tools for assessing risk of arrhythmias and left ventricular remodelling associated with progressive cardiac dysfunction and heart failure. Although clinical experience remains limited, careful evaluation of safety as well as validation of diagnostic and prognostic value of these techniques in clinical trials is still needed.

  • Molecular imaging
  • Atherosclerosis
  • PET
  • Heart failure
  • MRI

1. Introduction

Although traditional imaging is based on detection of changes in the anatomy and physiological features, such as blood flow or contractile function, molecular imaging aims at visualization and measurement of biological processes at the molecular and cellular levels.1 Compared with oncology, cardiovascular diseases are associated with several challenges for imaging technology, such as small size and continuous movement of the structures as well as often small amounts of pathological substrate. Despite these challenges, there is increasing evidence from pre-clinical studies documenting feasibility and clinical promise of targeted imaging of cardiovascular disease markers in the vasculature and myocardium, including altered energy metabolism, neuronal function, inflammation, apoptosis, thrombosis, and angiogenesis.

Molecular imaging can provide important insight into molecular and cellular mechanisms of diseases and thus, facilitate discovery and initial characterization of new pharmaceuticals. Molecular imaging is also seen as a potentially attractive clinical tool to provide early diagnosis and individual risk assessment. Examples of clinical problems that might benefit from molecular imaging include the identification of atherosclerotic plaques that are at high risk of causing acute complications (i.e. myocardial infarction, stroke, or sudden cardiac death),2 the identification of individuals who are at risk of developing ventricular remodelling and progressive heart failure (HF) after myocardial infarction, or the assessment of risk of arrhythmias in the presence of myocardial infarct scar or reduced left ventricular function.3 This could help to efficiently target the use of therapies, such as statins, angiotensin-converting enzyme inhibitors, and implatable cardioverter defibrillators for the individuals at highest risk. In addition to diagnosis, molecular imaging could provide surrogate endpoints to evaluate efficacy of therapies in clinical trials and monitor therapy responses. For example, trials based on clinical endpoints have provided conflicting results on the benefits of cell therapies in ischaemic heart disease.4 By performing these tasks non-invasively, molecular imaging could offer patient-friendly and safe alternatives to invasive diagnostic tests, such as endomyocardial biopsy and intravascular imaging.

This review provides an overview of imaging techniques and targeted imaging probes for molecular imaging of the major cardiovascular diseases and emerging therapies. Opportunities and challenges in translation of molecular imaging into clinical practice are discussed.

2. Techniques and imaging modalities

Molecular imaging is based on the use of molecular probes or biomarkers in very low concentrations to detect biological processes without disturbing their function.1 The probe is composed of a label system that can be visualized by imaging and a ligand that recognizes and binds to the target, for example, an antibody, peptide, or small molecule derived from candidate screening methods. To achieve images with high target-to-background ratio, the probe should include favourable pharmacokinetics, high binding efficacy, and specificity combined with lack of toxicity and feasibility of synthesis.

Features of non-invasive, clinical imaging modalities applicable for molecular imaging are outlined together with their applications in imaging anatomy and physiology in the cardiovascular system in Table 1.5 In order to facilitate translational research, dedicated small animal imaging systems have become available.

View this table:
Table 1

Overview of the clinical non-invasive imaging modalities for anatomical, functional, and targeted molecular imaging in the cardiovascular system

ModalityContrast agents (approx. sensitivity)Conventional applicationsExamples of molecular targeting (for details and references, see text)
UltrasoundMicrobubblesaCardiac structure, LV systolic and diastolic functions, Doppler haemodynamics, ischaemia, and viability (contrast-perfusion and Doppler-coronary flow)Inflammation and angiogenesis (VCAM-1-, αvβ3 integrin-, and VEGFR-targeted microbubbles)
MRIGadolinium chelates, iron oxide (10−3–10−9 M)Cardiac structure, ventricular functions, DE of myocardial infarction (coronary angiography, perfusion)Apoptosis, inflammation, proteolysis, ECM remodelling (various targeted nanoparticles, liposomes, micelles, and small molecules), metabolic pathways (MRS, hyperpolarized MRI)
CTIodineaCoronary angiography (LV structure and systolic function)Macrophage-targeted iodine-filled micelles
SPECTPhoton emitters 201Tl, 99mTc, 111In … (10−10–10−11 M)Perfusion and viability, LV and RV systolic functionsInnervation (mIBG), stem cells (reporter genes), apoptosis (annexin A5), angiogenesis (RGD peptides)
PETPositron emitters 18F, 15O, 13N, 11C, 64Co … (10−11–10−12 M)Myocardial blood flow, 18F-FDG of viabilityMetabolic pathways (BMIPP, FDG, acetate), innervations (HED), inflammation (FDG), angiogenesis (RGD peptides), gene expression and stem cells (reporter genes)
  • LV, left ventricle; VCAM-1, vascular cell adhesion molecule 1; VEGFR, vascular endothelial growth factor receptor; MRI, magnetic resonance imaging; DE, delayed enhancement; MRS, magnetic resonance spectroscopy; CT, computed tomography; SPECT, single photon emission computed tomography; mIBG, meta-iodobenzylguanidine; RV, right ventricle; FDG, fluorodeoxyglucose.

  • aNot defined yet.

2.1 Nuclear imaging

Nuclear imaging techniques [i.e. single photon emission computed tomography (SPECT) and positron emission tomography (PET)] have been the main molecular imaging modalities. The strengths of nuclear imaging are its high sensitivity and availability of a wide variety of targeted, radioactively labelled tracers in experimental and clinical applications.1,5 However, the exposure to ionizing radiation is a limitation that can be a major issue in screening of asymptomatic individuals or repeated, follow-up imaging. Currently, the leading molecular imaging modality is PET, which acquires data with higher spatial resolution than SPECT (4–5 vs. 10–16 mm with clinical scanners). Owing to its good temporal resolution and well-established correction of patient-specific photon attenuation, PET provides accurate delineation of regional tracer kinetics, which is used in combination with validated kinetic models to quantify biological processes in absolute terms.1,6 List mode data acquisition provides an opportunity to correct for motion during cardiac and respiratory cycles.7 Although availability remains a limitation, the exponential growth in the number of PETs, attributable primarily to the technology's widely accepted role in clinical oncology, is likely to increase the availability of this technique also for cardiac imaging in the future.

2.2 Magnetic resonance imaging

Magnetic resonance imaging (MRI) has the advantages of providing very high spatial and temporal resolution along with excellent soft tissue contrast for characterization of detailed cardiac anatomy and function without ionizing radiation and even without contrast agents. The major limitation of MRI in molecular imaging is its lower sensitivity compared with nuclear imaging. However, nanoparticles, such as iron oxide particles, micelles, liposomes, and emulsions that can deliver large amounts of contrast-generating material, i.e. gadolinium chelates (Gd, signal enhancement on T1-weighted images) or iron oxide (signal reduction on T2-weighted images), have been used as a platform for incorporation of various targeting ligands.8 Iron oxide particles have already undergone extensive clinical evaluation that has not revealed safety problems. The potential risk of retention of Gd in the tissue resulting in subsequent systemic fibrosis requires rigorous evaluation of safety. Another novel approach uses perfluorocarbon nanoparticles carrying fluorine (19F) that can be specifically detected by 19F MRI.9 This method detects signal that is distinct from any background signal and therefore quantifiable.

2.3 Ultrasound

Ultrasound molecular imaging is mainly based on the use of gas-filled, acoustically active intravascular microspheres or ‘microbubbles’ engineered to bind to specific endothelial targets and that can be detected using clinical ultrasound systems.10 The targeted approach is an extension of the clinically approved indication of microbubbles as a blood pool contrast agent. Nanoparticle ultrasound contrast agents, such as perfluorocarbons, can exit blood vessels and have been demonstrated to allow detection of extravascular targets.10,11 Ultrasound is inexpensive and currently the most widely used clinical cardiac imaging modality, without biological risks. Due to its very high temporal resolution, it provides images at real time. Developments of ultrasound pulse sequences may increase sensitivity and specificity of microbubble signals facilitating their cardiovascular molecular imaging applications.12

2.4 Optical imaging

Optical fluorescence and bioluminescence imaging allow high-speed and high-sensitivity detection of multiple fluorescent tracers, complementing well other imaging modalities.5 The limited penetration depth of optical imaging signals has so far limited their use for in vivo imaging.

2.5 Multi-modality imaging

Use of multiple imaging modalities is often required for localization and true quantification of signals from molecular probes that are typically seen as ‘hot spots’ with limited information about anatomy and physiology of the organ. For example, hybrid SPECT or PET and computed tomography (CT) scanners offer the possibility to integrate targeted PET images with high-resolution morphological images provided by CT to obtain an anatomic distribution of the probe and a correction for photon attenuation by body tissues and account for partial volume errors that would cause underestimation of the true regional radiotracer activity. It has also been demonstrated that it is possible to do simultaneous acquisition of PET and MRI in a hybrid scanner.13 Three-dimensional scanning techniques are likely to help co-registration of ultrasound images with other imaging modalities.14

There is increasing interest in the development of hybrid nanoparticles including more than one detection system in order to combine advantages of the different imaging modalities.1517 For example, studies on in vivo biodistribution by PET and/or MRI can be followed by high-resolution studies of cellular localization in tissue samples by optical methods. Quantum dots are small fluorescent semiconductor nanocrystals that represent an attractive fluorescent label system for nanoparticles since they offer high stability of photoemission, narrow emission spectrum, and bright fluorescent signals.16,18 However, due to their toxic components, clinical application may not be straightforward.

3. Atherosclerosis

The majority of myocardial infarctions, strokes, and sudden cardiac deaths result from rupture of the thin fibrous cap in atherosclerotic plaques.2,19 The plaques at high risk of rupture are typically large and contain a large necrotic core covered by thin and inflamed fibrous caps. However, due to outward remodelling, they often do not cause significant luminal obstruction or symptoms before the acute event, therefore making plaque rupture unpredictable using current diagnostic tools.

In the future, it is likely that the use of non-invasive coronary angiography, either MRI or CT, will increase in the diagnostic work-up of CAD.19 Although it is possible to non-invasively image coronary artery plaques at submillimetre resolution, their small size, for example, thickness of the fibrous cap being <100 µm, makes their characterization challenging on an anatomical basis alone. Molecular imaging may provide an opportunity to detect plaques at high risk of rupture with selective enhancement of contrast-to-noise ratio.2,19

3.1 Macrophages

Inflammation is a key feature of plaques at high risk of rupture.2,19 Imaging can directly detect accumulation of circulating monocytes that were labelled with radioactive probes before intravenous injection into atherosclerotic plaques in murine models.20 In the plaques, monocytes differentiate into macrophages and cholesterol-filled foam cells constituting a main component of the plaque lipid core. Macrophages secrete inflammatory cytokines and produce proteolytic enzymes that can weaken the fibrous cap of plaque and make it vulnerable to rupture. Prevalence of macrophages is high in ruptured plaques and in plaques of patients with acute coronary syndromes.2,19

Fluorine-18 (18F)-labelled fluorodeoxyglucose (FDG) PET is currently the best characterized approach for imaging plaque inflammation (Figure 1). 18F-FDG is a glucose analogue that is transported into cells that are active in glycolysis. However, inability to pass through the glycolytic pathway leads to its intracellular accumulation.1 Uptake of 18F-FDG has been demonstrated in the aorta, carotid artery atherosclerotic plaques of patients with ischaemic symptoms, and peripheral atherosclerotic arteries.2123 Histological analysis demonstrates that its accumulation in atherosclerotic lesions is associated with the degree of macrophage infiltration.21,22 Importantly, 18F-FDG PET signal in carotid arteries was highly reproducible in repeated PET scans and it was attenuated by simvastatin treatment indicating potential for monitoring of treatments.23,24

Figure 1

Hybrid 18F-FDG PET CT imaging of plaque inflammation. (A) Focal 18F-FDG uptake (arrow, left panel) that co-localizes with carotid artery in CT image (arrows, middle and right panels) in a patient with recent cerebral ischaemic event. There was high 18F-FDG uptake also in brain, jaw muscles, and facial soft tissues. (B) Incidental 18F-FDG uptake co-localizing with the proximal left coronary artery (arrows) and the aorta in a patient undergoing PET CT study for assessment of malignancy. Reproduced from Rudd et al.21 with permission from Wolters Kluwer Health and from Alexanderson et al.26 with kind permission from Springer Science+Business Media·Production.

In coronary vessels, intense tracer uptake in the adjacent myocardium combined with limited spatial resolution of PET limits plaque imaging with 18F-FDG PET. However, incidental uptake of 18F-FDG has been demonstrated in the coronary arteries of patients undergoing an 18F-FDG PET scan due to malignancies (Figure 1).25,26 An initial report of a prospective study employing coronary CT angiography for anatomical localization suggested that 18F-FDG uptake can be detected in the stented culprit lesions of patients with acute coronary syndromes if background 18F-FDG uptake in the myocardium is suppressed by metabolic intervention.27 Development of gating techniques may help to overcome inaccuracies related to motion of coronary arteries during the respiratory and cardiac cycles in the future.7 However, since 18F-FDG is a relatively non-specific marker of metabolic activity, other, more specific tracers are needed to image plaque inflammation and vulnerability.

A promising approach for imaging macrophages involves the use of small iron oxide particles that are often efficiently phagocytosed by macrophages and can be detected in atherosclerotic plaques as negative MRI contrast on T2-weighed images.28,29 The value of iron oxide particle accumulation in carotid arteries as a potential risk marker of atherosclerosis is being evaluated in clinical trials. New MRI probes for imaging of macrophages in atherosclerotic lesions include macrophage scavenger receptor-targeted immunomicelles loaded with Gd, which produces a positive contrast on T1-weighed images.16 A recent study introduced iodine-loaded micelle for imaging macrophages in atherosclerotic rabbit aorta with the use of CT. This could potentially present an adjunct to the clinical evaluation of coronary arteries with contrast CT.30

3.2 Leucocyte adhesion molecules

Endothelial expression of the leucocyte adhesion molecules is increased even in early phases of atherosclerosis and may serve as a biomarker for vascular inflammation. Targeted microbubbles and magnetofluorescent nanoparticles have allowed visualization of adhesion molecule expression in experimental mouse models of atherosclerosis by ultrasound and MRI.12,31

3.3 Apoptosis

High rate of macrophage apoptosis is a typical feature of fibrous caps of vulnerable and ruptured atherosclerotic plaques. Apoptosis can be imaged with labelled annexin A5 that shows selective, high-affinity binding to phospholipids exposed on the surface of apoptotic cells that triggers uptake of cell remnants by neighbouring cells. Technetium (99mTc)-labelled annexin A5 showed uptake in human carotid arteries in patients with recent ischaemic syndromes and co-localized with apoptotic macrophages.32 However, binding of annexin A5 may not be specific for apoptosis as indicated by its binding to platelets in thrombi.33

3.4 Angiogenesis

Intraplaque angiogenesis by proliferation of medial vasa vasorum has been implicated in rapid plaque growth, intraplaque haemorrhage, and plaque rupture.34 A potential marker of inflammation and angiogenesis in atherosclerotic lesions is αvβ3 integrin, a cell surface glycoprotein receptor highly expressed by macrophages and endothelial cells.35,36 MRI of αVβ3 integrin-targeted paramagnetic nanoparticles demonstrated signal enhancement in the rabbit aorta with early atherosclerotic changes including expansion of the adventitial vasa vasorum.35 More recently, cyclic peptides that contain the Arg-Gly-Asp (RGD) attachment site labelled for optical imaging or PET have demonstrated focally increased uptake in advanced, macrophage-rich atherosclerotic lesions of hypercholesterolaemic mice.36,37

3.5 Extracellular matrix remodelling

Recent experimental studies have demonstrated imaging of activated matrix metalloproteinase (MMP) enzymes in the vasculature of atherosclerotic mice or rabbits using MMP inhibitors labelled for SPECT,38 MRI,39 or optical imaging.40 These studies provide the proof-of-principle of imaging MMP activation that is considered as an attractive target since it may directly contribute to degradation of the protective fibrous cap of atherosclerotic plaques.2,19

Extracellular matrix (ECM) constitutes a major component of atherosclerotic neointima and contrast agents targeting ECM proteins, such as collagens, proteoglycans, and elastin, may be useful for assessment of the extent of atherosclerosis, staging of lesions, and following their progression or regression.41,42 The ECM proteins are present in such large amounts that the use of small molecular MRI contrast agents with kinetics similar to non-targeted Gd may be feasible. This would facilitate translation of their use into clinical studies.

3.6 Fibrin-targeted MRI contrast agents

Thrombus formation may be a useful target for identification of complicated atherosclerotic lesions as well as sources of thrombo-embolism.2,19 Fibrin is abundant in thrombi and has been targeted for thrombosis imaging. Most of the experience is obtained using the small-molecular Gd-based fibrin-specific MRI contrast agent EP2104R. The tracer has undergone extensive experimental validation and initial clinical results in 11 patients with suspected thrombosis indicate that the agent provides enough signals for visualization of intracardiac, extracardiac arterial, and venous thrombi.43,44

4. Myocardial infarction, remodelling, and HF

In addition to acute myocardial insults, such as myocardial infarction, chronic left ventricular remodelling plays an important role in the development and progression of HF.3 Remodelling involves activation of the neurohormonal mediators (e.g. inflammation, sympathetic nervous system, and renin–angiotensin system), ongoing loss of myocytes, maladaptive changes in the surviving myocytes (i.e. accumulation of glycogen and reduced high-energy phosphate stores), and injury to ECM (i.e. fibrosis and activation of metalloproteinases).3 Molecular imaging of the mechanisms underlying myocardial injury and remodelling may provide diagnostic and prognostic markers for the identification of high-risk patients and may predict response to therapeutic interventions.

4.1 Myocardial ischaemia

Rest and stress myocardial perfusion imaging is a well-established technique for diagnosis and risk stratification of coronary artery disease that is the most common underlying cause of HF in the western world.45 By the use of radiopharmaceuticals, which are retained in the myocardium in proportion to blood flow, relative and absolute measures of myocardial blood flow can be obtained.6 Comparison of tracer distributions at rest and exercise or pharmacological stress can describe the presence, severity, and extent of myocardial ischaemia caused by flow-limiting coronary stenoses. A normal myocardial perfusion study identifies a low-risk population independently of the presence or absence of anatomical coronary artery disease.45 In addition to nuclear imaging, dynamic MRI of the first-pass signal changes in the myocardium after injection of a fast bolus of Gd has shown potential for assessment of myocardial perfusion and the presence of coronary stenosis.46 Tracers that are avidly taken up in hypoxic myocardium, such as 18F-fluoromisonidazole, have been proposed for determining the presence and extent of jeopardized, but salvageable myocardium in acute ischaemic syndromes, since their uptake is present in viable, hypoxic tissue, although there is no uptake in irreversibly injured myocardium.47 Hybrid imaging with PET CTs combines almost simultaneous anatomic, functional, and metabolic imaging and may provide new quantitative tools to study the pathophysiology of HF.

4.2 Metabolism

18F-FDG traces myocyte glucose uptake and phosphorylation and can be used to quantify regional myocardial glucose metabolism.48 18F-FDG uptake provides a clinically useful measure of tissue viability within hypoperfused segments in patients with advanced coronary artery disease and impaired left ventricular function.49 Although limited by the lack of large randomized clinical trials, a meta-analysis of retrospective data indicates that such patients with ischaemic, but viable myocardium are at substantial risk of death, which can be effectively reduced by revascularization.50

Myocardial kinetics of carbon-11 (11C)-labelled palmitate describe uptake and metabolism of long-chain fatty acids. Their oxidation represents the major source of cardiac high-energy phosphates, but complex metabolism allows only semi-quantitative measurements of fatty acid metabolism with PET imaging of 11C-palmitate in patients.51 More recently, iodinated fatty acids such as β-methyl-iodophenylpentadecanoic acid (BMIPP) have provided a sensitive SPECT marker of altered fatty acid transport into the myocyte for patient studies.52 Fatty acid imaging may gain clinical acceptance in HF if clinical trials confirm that a metabolic switch from fatty acids to glucose utilization in the failing myocardium improves the outcome of patients.51 A transient ischaemic event results in a metabolic shift from fatty acids to glucose as the preferred substrate for energy production that persists for a prolonged time. Imaging of this feature called ischaemic memory may reveal the ischaemic origin of recent chest pain and define the extent of myocardium compromised by ischaemia in patients with suspected acute coronary syndromes.52

PET imaging of carbon 11C-labelled acetate allows assessment of cardiac oxidative metabolism without the complexity of substrate interaction between glucose and fatty acids.53 The early, rapid clearance of acetate correlates closely with myocardial oxygen consumption. Thus, the relationship of myocardial 11C-acetate kinetics to cardiac work offers a non-invasive parameter for cardiac efficiency, i.e. the fraction of total expended energy that is converted into external work, that has been used to demonstrate the effect of pharmacological and pacing interventions on cardiac energetics in HF patients.54,55

Magnetic resonance spectroscopy (MRS) is able to measure energy metabolism, intermediary metabolism, drug metabolism and ion homeostasis through analysis of nuclei, such as hydrogen (1H), carbon (13C), fluorine (19F), sodium (23Na), and phosphorous (31P).56 The ratio of phosphocreatine to ATP, which can be measured by 31P-MRS, provides an index of the energetic state of the heart.56 However, 31P-MR spectra may be difficult to quantify, and although the technique has been performed with standard 1.5-T systems, the signal-to-noise ratio is more favourable at high magnetic field strength. Although MRS is likely to shed light on important mechanisms in HF, its clinical adoption seems unlikely at present due to long acquisition times and low spatial resolution. Hyperpolarization and chemical shift imaging with 13C-labelled substances, such as 13C-pyruvate may enhance the sensitivity of MRI measures of cardiac substrate metabolism and become a useful tool for experimental and clinical research.57

4.3 Apoptosis

Apoptosis is a mechanism of cardiomyocyte death in myocardial infarction and remodelling that can contribute to progression of HF if functional myocytes are lost by apoptosis.3 Increased myocardial uptake of 99mTc annexin A5 has been demonstrated in acute ischaemia–reperfusion injury, allograft rejection, myocarditis, and in a small group of patients with advanced, non-ischaemic cardiomyopathy.5861 Annexin-labelled magnetofluorescent nanoparticle for MRI of apoptosis has also been characterized.62 The specificity of annexin for apoptosis may be impaired in the presence of negatively charged phospholipids in necrotic cells and low rate of apoptosis in HF makes the assessment of differences in the rate of apoptosis by imaging challenging.

4.4 Inflammation

Inflammation is an important mediator of acute myocardial ischaemia–reperfusion injury and continues to play a role in the subsequent scar formation and left ventricular remodelling.3 In addition to these conditions, imaging of inflammation could be useful for detection of myocarditis and transplant rejection. Myocardial infiltration of monocytes labelled with nanoparticles before intravenous injection has been imaged by MRI in experimental myocardial infarction.62 Similarly, ischaemia-induced expression of leucocyte adhesion molecules by myocardial microvasculature has been imaged with ultrasound and targeted microbubbles in experimental models.10 Most recently, a Gd-based probe allowing MRI of increased activity of myeloperoxidase (MPO), an enzyme secreted by neutrophils and macrophages, was applied to experimental infarction model.63 The probe becomes polymerized in the presence of MPO resulting in increased relaxivity, protein binding, and slow wash-out, all contributing to signal enhancement in areas of high MPO activity.

4.5 Angiogenesis and ECM remodelling

Induction of angiogenesis, i.e. formation of new capillaries from pre-existing vasculature, is an important part of the myocardial repair process following ischaemic injury, the mechanisms of which have been recently reviewed.64 Although both myocardial perfusion and functional recovery are endpoints in the clinical evaluation of angiogenetic therapy, specific imaging markers may monitor the effect of new drugs or interventions more directly.64 Potential imaging approaches include probes that target up-regulation of vascular endothelial growth factor (VEGF) receptors65 or αvβ36670 integrin. The former has been targeted with radiolabelled VEGF peptides and the latter with radioactively labelled cyclic RGD peptides, MRI nanoparticles, and ultrasound microbubbles. Experimental studies and initial results in patients show regionally increased VEGF and RGD signals in ischaemic myocardium (Figure 2). Further experimental and clinical research will focus on the specificity of the new imaging probes for angiogenesis and their relationship to remodelling after myocardial infarction. Both RGD peptides and probes for MMP activation have also been proposed to reflect ECM remodelling following myocardial infarction.70,71

Figure 2

18F-galacto-RGD PET of αvβ3 integrin expression in a patient who had 2 weeks old inferior reperfused infarction. 18F-galacto-RGD was injected 1 h before imaging. PET images in short axis of the left ventricle demonstrate 18F-galacto-RGD uptake (arrows in C and D) that co-localizes with a perfusion defect in 13N-ammonia PET image (arrow in A) and delayed enhancement in MRI (arrows in B and C) indicating that 18F-galacto-RGD uptake is present in the area of myocardial infarction.

4.6 Innervation imaging

Activation of the sympathetic nervous system is a hallmark of HF.3,72 It has been associated with progression of HF, arrhythmias, and sudden cardiac death, which is a mode of death in a substantial proportion of HF patients. Nuclear imaging tracers targeting cardiac sympathetic nerve terminals have undergone extensive clinical testing and provided promising results in HF.

Extraneuronal sympathetic neurotransmitter concentrations are regulated by an efficient amine uptake mechanism, the noradrenaline transporter (NAT), which can be used to visualize the sympathetic nerve terminal by radiolabelled norepinephrine analogues. The most commonly used SPECT tracer, iodine 123 (123I)-labelled meta-iodobenzylguanidine (mIBG), undergoes avid uptake and storage in the cardiac nerve terminals but is not metabolized by monoamine oxidase. Uptake of this tracer is specific for the integrity of sympathetic nerve terminals. Cardiac 123I-mIBG uptake is measured by defining the heart-to-mediastinum uptake ratio, assessing washout kinetics, or quantifying the apex-base gradient of denervation in the failing heart.72 123I-mIBG imaging appears to have prognostic value in patients with left ventricular dysfunction, and it may be adopted clinically if prospective multicentre studies confirm its incremental prognostic value.73,74 Individual risk profiles for HF patients may also permit more selective use of costly interventions, such as implantable defibrillators.75

The PET tracers 11C-hydroxyephedrine and 11C-epinephrine allow quantification of the density of sympathetic nerve terminals (Figure 3).76 Myocardial 11C-hydroxyephedrine uptake correlates with norepinephrine tissue concentration and density of NAT. Post-synaptic β-receptor density can be assessed with a radiolabelled β-receptor antagonist 11C-CGP12177. Combined assessment of pre-synaptic neurotransmitter uptake and post-synaptic receptor density has provided new insights into the mechanisms of neuronal dysfunction in HF and arrhythmic conditions.77,78 The use of radiopharmaceuticals in imaging various aspects of neuronal function (transmitter uptake, release, metabolism, and storage) may permit guidance of pharmacological interventions and their effects on cardiac autonomic innervation. Moreover, imaging of dysfunction of cardiac sympathetic innervation has been proposed for differential diagnosis of non-cardiac diseases, such as diabetes and neurodegenerative disorders.79

Figure 3

11C-meta-hydroxyephedrine (HED) PET of myocardial sympathetic innervation in a patient with HF due to dilated cardiomyopathy. PET images obtained 40 min after injection of HED in short axis (upper row), vertical long axis (middle row), and horizontal long axis (lower row) show inhomogeneous retention of the tracer, despite preserved myocardial perfusion as assessed by 13N-ammonia. Quantitative HED retention index representing myocardial activity at 40 min after tracer injection normalized to arterial input function is reduced (normal values >12%).

5. Gene and cell therapies

Cell transplantation is extensively studied as a potential therapeutic option for patients with impaired cardiac function due to cell death. Methods to monitor cell migration, homing, survival, and engraftment may facilitate the understanding of the heterogeneous results from early clinical investigations of intracoronary injection of bone marrow-derived cells.4 The tasks and strategies for imaging can be divided into short-term cell labelling, cell survival assays, and long-term monitoring of cell differentiation.

Short-term visualization after transplantation may be obtained by direct labelling of stem cells with indium 111 (111In)-labelled oxine, 99mTc-exametazime (HMPAO), or 18F-FDG, a process achievable without detectable changes in viability, functionality, migration, and proliferative capacity.4,80 Only a small fraction of radioactivity (<5%) is observed in the myocardium, which suggests regional retention by only a few transplanted cells.81 The duration of cell tracking ranges from a few hours with 18F-FDG-PET in human infarctions to 7 days after injection with 111In with SPECT CT in dogs.

Paramagnetic or fluorinated nanoparticles can be used to facilitate non-toxic labelling of stem cells before transplantation, which would enable repetitive MRI. High spatial resolution and direct correlation of cell signals to regional function or delayed enhancement make this approach very attractive for experimental and clinical research.82,83 However, the signal is not directly related to cell viability, so it may lose specificity for transplanted cells after cell death and macrophage phagocytosis (Figure 4).

Figure 4

Multi-modality imaging of dual-labelled endothelial progenitor cells with nuclear reporter probe (human sodium/iodide symporter) and iron oxide particles (Resovist). Labelled cells were injected into rat myocardium and imaged at 1, 3, and 7 days (arrows indicate the injection site). Note the stable MRI signal, whereas 124I uptake in transplanted cells decreased rapidly, most likely due to rapid cell death.

Monitoring of cell survival and differentiation requires imaging techniques linked to the integrity of cells and tissue-specific protein expression. Reporter gene imaging with optical or bioluminescence imaging approach is successful in small animal research, whereas scintigraphic techniques may be applicable to large animals and patients.84 Most studies label an enzyme or transporter (herpes simplex virus type I thymidine kinase) or the human iodide/symporter (Figure 4).4,84 The application of these approaches to clinical research will require further documentation of suitability and safety to pass regulatory hurdles.

6. Conclusions and future directions

Molecular imaging has provided a large number of attractive new probes for cardiovascular imaging. Owing to high sensitivity and availability of tracers with low risk of toxicity, the most clinical experience in cardiovascular molecular imaging is mainly based on the use of nuclear imaging techniques. New nanoparticle contrast agents together with advances in scanner technology can provide the required signal amplification enabling molecular imaging using modalities with limited inherent sensitivity, such as MRI and ultrasound. They could provide radiation-free alternatives to nuclear imaging. However, pharmacokinetics and safety of nanoparticle contrast agents capable of delivering large amounts of Gd require further careful evaluation in humans.

Nuclear imaging techniques of myocardial metabolism, viability, and innervation have been available for many years and have demonstrated their usefulness in clinical studies. Evaluation of cardiac metabolism using PET and SPECT tracers has proved to be a valuable technique for initial clinical evaluation of new pharmacological and device therapies for HF.54,55 PET with 18F-FDG is the most accurate non-invasive technique to assess myocardial viability, which has well-documented prognostic value in selection of patients for revascularization therapies.50 Evaluation of sympathetic innervation with SPECT and PET may be adopted for wide-spread clinical use provided that ongoing large prospective studies confirm their value in the assessment of prognosis of patients with myocardial infarction or HF.73,74

Many more important clinical questions can potentially be addressed with the new targeted imaging probes. For example, molecular imaging of atherosclerotic plaques with markers of inflammation may improve assessment of risk for acute complications. Although imaging of carotid arteries seems to be feasible, imaging of coronary arteries is technically challenging and further improvement in contrast-to-noise ratios and motion correction of the signals are needed. Then, prospective, long-term studies are needed to determine how molecular imaging tools could be combined with clinical evaluation and serum markers, such as C-reactive protein, to guide management of high-risk individuals with aggressive risk factor modification. The importance of diagnostic tests to provide information that leads to a more appropriate choice of therapy and thus, improved patient outcome has been emphasized in recent discussions related to increasing costs of diagnostic studies. Large clinical trials documenting their diagnostic and prognostic value are needed before potential new probes can be adopted clinically. This requires standardized and robust image analysis protocols and collaborative efforts of multiple centres.

Funding

Financial assistance was received from EC-FP6-project DiMI (LSHB-CT-2005-512146), Finnish Foundation for Cardiovascular Research, The Academy of Finland Centre of Excellence in Molecular Imaging in Cardiovascular and Metabolic Research, and Bristol–Myers Squib Medical Imaging.

Acknowledgement

We acknowledge the editorial assistance of Jasmine Schirmer.

Conflict of interest: none declared.

References

View Abstract