Positron emission tomography imaging of atherosclerosis.

Atherosclerosis-related cardiovascular events are the leading causes of death in the industrialized world. Atherosclerosis develops insidiously and the initial manifestation is usually sudden cardiac death, stroke, or myocardial infarction. Molecular imaging is a valuable tool to identify the disease at an early stage before fatal manifestations occur. Among the various molecular imaging techniques, this review mainly focuses on positron emission tomography (PET) imaging of atherosclerosis. The targets and pathways that have been investigated to date for PET imaging of atherosclerosis include: glycolysis, cell membrane metabolism (phosphatidylcholine synthesis), integrin αvβ3, low density lipoprotein (LDL) receptors (LDLr), natriuretic peptide clearance receptors (NPCRs), fatty acid synthesis, vascular cell adhesion molecule-1 (VCAM-1), macrophages, platelets, etc. Many PET tracers have been investigated clinically for imaging of atherosclerosis. Early diagnosis of atherosclerotic lesions by PET imaging can help to prevent the premature death caused by atherosclerosis, and smooth translation of promising PET tracers into the clinic is critical to the benefit of patients.

Introduction

Atherosclerosis is a leading cause of death in developed countries and it is crucial to distinguish the unstable plaques from the stable ones to increase the survival of patients with early intervention 1. The initial abnormality in the pathogenesis of atherosclerotic plaques is the fatty streak, a white/yellow linear discoloration that is visible macroscopically on the endothelial surface of an artery, caused by accumulation of lipids and macrophages. This initial lesion matures into an atherosclerotic plaque with the accumulation of lipid and connective tissue, in particular collagen. As the plaque grows, the vessel expands to preserve the blood flow, a process known as positive remodeling. The artery eventually can expand no farther, and the plaque begins to occlude the lumen of the vessel causing circulatory obstruction 2-4 (Figure ). Unstable plaques that are prone to cardiovascular accidents demonstrate a thin fibrous cap, large lipid core, paucity of smooth muscle cells and abundance of inflammatory cells (Figure ) 5, 6.

Clinical imaging of atherosclerosis

Traditionally atherosclerosis is visualized based on the anatomical changes in the vessel walls such as calcification and stenosis. Many techniques have been used for imaging of atherosclerosis in the clinic, such as X-ray angiography, optical coherence tomography (OCT), intravascular ultrasound (US), computed tomography (CT), magnetic resonance imaging (MRI), single photon emission computed tomography (SPECT), positron emission tomography (PET), etc. 4. Angiography yields an excellent resolution but it is highly observer dependent and not very reproducible 7-9. Moreover diffuse arterial narrowing without obvious luminal irregularity can be easily missed in angiographic images 10. Even if angiography could demonstrate the luminal narrowing accurately, it has been documented that acute coronary syndromes often result from plaque rupture at sites with no or only modest luminal narrowing based on angiography 11-13.

Molecular imaging of atherosclerosis

Many biological changes take place in atherosclerotic plaques preceding the appearance of anatomical disturbances 24, 25. Given that a large proportion of people who suffer a sudden cardiac event have no prior symptoms 26, it is crucial for diagnostic procedures to go beyond the simple assessment of the vessel lumen to identify rupture-prone vulnerable plaques 6. When compared with traditional methods, molecular imaging approaches have a number of advantages, such as enabling noninvasive study of cells in their natural microenvironment hence providing information on the whole biological process rather than a small part of it. In terms of experimental studies it is possible to perform longitudinal studies in the same animals with molecular imaging techniques, which will reduce the cost and can provide better statistical power.

Since the beginning of the 21st century, the field of molecular imaging has expanded tremendously 27-34. Many of the preclinical imaging studies focused on atherosclerosis involved the use of MRI and/or optical techniques, for which excellent recent reviews are available 4, 6, 11, 35-37. A few other imaging techniques have also been employed for imaging of atherosclerosis such as Raman spectroscopy, which measures the light-scattering effects 23, 38. In one study, Raman spectroscopy was used for imaging atherosclerosis associated inflammation with protease-activated fluorescent probes (representing capthesin-B and matrix metalloprotease (MMP)-2/9 expression) 39.

Molecular vascular imaging with PET can enable early detection of these changes, thereby decreasing the dependence to invasive biopsies or surgical procedures to characterize diseased tissues 11, 20, 23, 24, 40, 41. PET can detect tracer concentrations in the picomolar range, providing 4-5 mm resolution with clinical PET scanners and 1-2 mm with small animal PET scanners, and has been routinely used in clinical cancer patient management and preclinical research 42-48. Although there is no optimal PET tracer for routine clinical imaging of atherosclerosis available to date 10, various imaging agents have been investigated for PET imaging of atherosclerosis, targeting different components of the atherosclerotic plaque (Table ).

In this review, we will summarize the targets and tracers used for PET imaging of atherosclerosis in experimental and clinical studies. The targets and pathways that have been investigated to date for PET imaging of atherosclerosis include: glycolysis, cell membrane metabolism (phosphatidylcholine synthesis), integrin αvβ3, low density lipoprotein (LDL) receptors (LDLr), natriuretic peptide clearance receptors (NPCRs), fatty acid synthesis, vascular cell adhesion molecule-1 (VCAM-1), macrophages, platelets, etc. (Table ). To create experimental models for atherosclerosis, animals are typically kept on a high fat diet. Commonly used animal models for studying atherosclerosis are: C57BL/6 Apolipoprotein (Apo) E-/- mice, C57BL/6 LDL receptor deficient mice (LDLr-/- mice), heterozygous LDL receptor deficient rabbits (LDLr+/- rabbits), and homozygous LDL receptor deficient rabbits (LDLr-/- rabbits) 49-51. The aorta is typically used as the human coronary artery analog in rodents since the diameter of mouse aorta is ~1 mm, comparable to the size of small arteries in humans 11.

PET imaging of atherosclerosis with 18F-FDG

2-deoxy-2-18F-fluoro-D-glucose (18F-FDG) is one of the most frequently used PET tracers for in vivo imaging of atherosclerosis 4. Deoxyglucose is a glucose analog that competes with glucose for uptake into metabolically active macrophages in atherosclerotic plaques. After being labeled with 18F, the resultant 18F-FDG is taken up into metabolically active cells but is not metabolized and thereby accumulates in atherosclerotic plaques (Figure ) 4. More than 15 years ago, studies have demonstrated uptake of 18F-FDG in the region of the aortic arch in a rabbit model of atherosclerosis 52, while normal rabbits did not show increased uptake above background levels. Ex vivo analysis of the aortic arch confirmed 18F-FDG uptake in areas of atherosclerosis that were rich in macrophages. In another early study, Lederman et al. 53 documented a 4-fold increase in 18F-FDG uptake by atherosclerotic plaques in iliac arteries of New Zealand White rabbits fed with high cholesterol diet.

A number of studies have been reported in patients. Yun et al. 10 reported the increased uptake of 18F-FDG in 137 consecutive patients correlated with atherogenic risk factors. Among all risk factors, age was found to be the most significant factor. In another study, Rudd et al. 54 documented the ability of 18F-FDG PET to image inflammation within carotid artery atherosclerotic plaques in 8 patients with symptomatic carotid atherosclerosis, who had experienced a recent carotid territory transient ischemic attack (TIA) and had an internal carotid artery stenosis of at least 70%. Several other clinical studies followed 55-58.

However, in a study carried out by Laurberg et al. 59, the authors did not observe increased aortic or carotid tracer accumulation in hypercholesterolemic, ApoE-/- mice. Moreover, Menezes et al. 60 retrospectively examined 250 PET/CT images of 50 patients covering a 5 year period and found that they could not reproduce the similar results in subsequent images in patients with initially increased 18F-FDG uptake in carotids and the aorta. Therefore, it was concluded that increased 18F-FDG uptake in arterial lesions is a transient phenomenon. The discrepancy between the results of these studies and others remains unexplained 61, 62. In an interesting report, it was shown that different mechanisms for stimulation of macrophages, smooth muscle cells, and endothelial cells may contribute to increased 18F-FDG uptake in atherosclerotic plaques 63. Experimental evidence suggested that hypoxia may play an important role in 18F-FDG accumulation in atheromata, which deserves to be investigated in more detail in animal models and clinical settings to further our understanding of the biological mechanisms and significance of 18F-FDG uptake in atheromata.

PET imaging of atherosclerosis with radiolabeled choline

Choline is taken up into cells by specific transport mechanisms, phosphorylated by choline kinase, metabolized into phosphatidylcholine, and eventually incorporated into the cell membrane 61. Increased uptake of choline has been observed in activated macrophages 64, which led to the idea of using radiolabeled choline for PET imaging of atherosclerotic plaques. For example, Matter et al. 65 used 18F-Choline as the tracer to detect atherosclerotic plaques in ApoE-/- mice and reported superior results when compared with 18F-FDG. Laitinen et al. 66 observed a high 11C-choline uptake in the aortic plaques of atherosclerotic mice deficient for both LDLr and apolipoprotein B48. Kato et al. 67 reported an increased 11C-choline uptake in vessel walls of 93 consecutive male patients between 60 and 80 years old (Figure ). However, PET tracer uptake and calcification are rarely co-localized. Therefore, it was concluded that 11C-choline has the potential to provide information about atherosclerotic plaques independent of calcification measurement.

PET imaging of atherosclerosis targeting integrin αvβ3

Integrin αvβ3 is a cell adhesion molecule expressed by macrophages and endothelial cells in atherosclerotic lesions 4. Laitinen et al. 68 evaluated integrin αvβ3 deposition in atherosclerotic plaques in mice with 18F-galacto-RGD PET (RGD denotes arginine-glycine-aspartic acid, potent antagonists of integrin αvβ3). They dissected the aorta of mice 2 hours after tracer injection and found that biodistribution of 18F-galacto-RGD was higher in the atherosclerotic than in the normal aorta, and 18F-galacto-RGD uptake was directly correlated with macrophage density in the plaques. In a previous clinical study, Beer et al. 69 investigated the dosimetry of 18F-galacto-RGD in 18 human subjects and concluded that it demonstrates high metabolic stability, a favorable biodistribution, and a low radiation dose. Therefore, this tracer can safely be used for noninvasive PET imaging of molecular processes (e.g. atherosclerosis) involving integrin αvβ3.

PET imaging of atherosclerosis with 18F-labeled LDL

Oxidation of LDL is regarded as a crucial event in atherogenesis. Langer et al. 70 conjugated 124I to the scavenger receptor CD68 (soluble CD68-Fc) and used this molecule for the detection of atherosclerotic plaques in wild type or ApoE-/- mice. Since CD68 binds to oxidized LDL (ox-LDL) molecules and mediates their uptake, this tracer could be a useful tool for noninvasive imaging of atherosclerosis. In another study, Pietzsch et al. 71 reported PET imaging with 18F-labeled native LDL (nLDL) and ox-LDL after injection into male Wistar rats. It was suggested that the use of 18F-labeled LDL could be an attractive alternative to iodinated LDL. Future studies are warranted to clarify the potential of radiolabeled LDL for the detection of atherosclerotic changes.

PET imaging of atherosclerosis targeting NPCR

Natriuretic peptides (NPs) have potent antiproliferative and antimigratory effects on vascular smooth-muscle cells and participate in vascular remodeling in atherosclerosis. The expression of NPCRs is upregulated both in endothelium and in vascular smooth-muscle cells during the formation of atherosclerotic plaques 72. Liu et al. 72 investigated the potential of radiolabeled C-type atrial natriuretic factor (C-ANF) to noninvasively image developing plaque-like lesions. C-ANF was linked with 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA) and labeled with 64Cu for noninvasive PET in a hypercholesterolemic rabbit model, which demonstrated that 64Cu-DOTA-C-ANF was a promising candidate for PET of NPCR in atherosclerotic plaques (Figure ). In a follow-up study, the same group developed a C-ANF-integrated, 64Cu-labeled nanoprobe for in vivo PET imaging of NPCR during angiogenesis in a mouse hindlimb ischemia model 73.

PET imaging of atherosclerosis with other tracers

In addition to the abovementioned examples, many other tracers have been investigated for imaging of atherosclerosis. Derlin et al. 74 used 11C-acetate PET/CT for the evaluation of atherosclerotic plaques in 36 patients (Figure ). Fatty acids are a common constituent of atherosclerotic plaque and may be synthesized in the plaque itself. Fatty acid synthesis requires acetyl-coenzyme-A (CoA) as a main substrate, which is produced from acetate. Therefore, 11C-acetate, just like 18F-FDG, can be used as a non-specific tracer for in vivo imaging of atherosclerotic plaques.

VCAM-1 is a monocyte and lymphocyte adhesion factor expressed by endothelial cells under pro-atherogenic conditions 75, 76. Nahrendorf et al. 77 have investigated the potential of VCAM-1 as a target for PET imaging of vulnerable plaques. A tetrameric VCAM-1 affinity peptide (V4), which can be internalized into endothelial cells after VCAM-1 binding, was labeled with 18F and tested in atherosclerotic ApoE-/- mice receiving a high-cholesterol diet. The tracer was found to be suitable for noninvasive PET imaging of VCAM-1 in inflammatory atherosclerosis, as well as monitoring the effects of therapeutic intervention. In another study, Nahrendorf et al. 78 also reported the use of a 64Cu-labeled, dextranated, diethylenetriaminepentaacetic acid (DTPA)-modified magnetofluorescent nanoparticle for imaging of atherosclerotic plaques. The agent was termed 64Cu-TNP (i.e. trireporter nanoparticle), where its uptake in atherosclerosis was based on the avidity of lesional macrophages for polysaccharide-containing supramolecular structures. It was found that 64Cu-TNP accumulated predominantly in macrophages in atherosclerotic plaques in ApoE-/- mice (Figure ).

Silvola et al. 79 investigated the uptake of ionic 68Ga in atherosclerotic LDLr-/-ApoB100/100 mice. PET/CT imaging revealed an elevated 68Ga uptake in the aortic atherosclerotic plaques of mice, especially at the sites that are rich in macrophages. Ex vivo biodistribution studies and autoradiography of aortic cryosections confirmed the results of in vivo imaging. However, it was stated that the slow blood clearance may limit the clinical usability of 68Ga as a PET tracer for imaging of atherosclerosis. In another study, Yano et al. 80 suggested that 68Ga-labeled platelets could be useful for PET imaging of thrombosis or atherosclerosis. However, no in vivo studies were carried out, although in vitro labeling of rabbit platelets with 68Ga was successful.

Gaemperli et al. 81 recently suggested the use of 11C-PK11195, a selective ligand of the translocator protein which is highly expressed by activated macrophages, for PET imaging of atherosclerosis. Initial results obtained from 32 patients with carotid stenosis were encouraging. Elmaleh et al. 82 evaluated PET imaging with 18F-AppCHFppA, a competitive inhibitor of adenosine diphosphate-induced platelet aggregation, to detect atherosclerotic lesions in male New Zealand White rabbits. It was reported that the accumulation of 18F-AppCHFppA in macrophage rich atherosclerotic plaques could be quantified noninvasively.

Conclusion and future perspectives

A wide variety of targets and tracers have been investigated for PET imaging of atherosclerosis, which has significant potential for translation and future clinical imaging of atherosclerosis. Among these PET tracers, inflammation-targeted tracers seem particularly promising. Monocytes, macrophages, and foam cells play important roles in the evolution and complications of atherosclerosis 83. Following their migration to the lesion site, these cells scavenge lipids, secrete cytokines that further amplify inflammation and produce proteases that can be noninvasively visualized with the use of suitable tracers 11, 83. The most commonly used PET tracer for imaging of atherosclerosis is 18F-FDG, with many other tracers already tested clinically: 11C-choline 67, 18F-galacto-RGD 69, 11C-acetate 74, and 11C-PK11195 81. Since 11C-PK11195 specifically binds to macrophages through translocator proteins that are highly expressed on activated macrophages, it may provide more specificity for imaging of atherosclerosis than 18F-FDG.

An important contribution of PET imaging of atherosclerosis will be to determine the at-risk individuals earlier in the disease process by targeting one of the abovementioned components at the early stage of disease. It may also prove to be useful in determining the biological responses to certain therapeutic intervention. The key strengths of PET include its excellent sensitivity, limitless depth of penetration, and quantitation capabilities 27. However, the small size of the atherosclerotic lesions and their vicinity with blood, the unbound/circulating tracer activity, and the continuous respiratory and cardiac movements during the acquisition of the images are the unsolved issues in PET imaging of atherosclerosis 11, 12, 84. The availability of PET/MR systems 85, 86, where MRI can provide high spatial resolution and exquisite soft tissue contrast to complement PET, may dramatically facilitate the translation of promising PET tracers into the clinic for atherosclerosis imaging.

Measurement of regional and global calcification of the heart and major arteries using 18F-NaF PET/CT may also be useful for early detection of atherosclerosis 87. This technique may provide highly relevant information about the state of calcified plaque before structural calcification is detectable by standard CT techniques, thereby allowing for earlier intervention for risk reduction in cardiovascular diseases 87, 88. Another potential target for the PET imaging of atherosclerosis is CD105 (i.e. endoglin), where it was demonstrated that precursor atherosclerotic lesions exhibit intimal neovascularization that are associated with increased CD105 expression on endothelial cells 89-91. In addition, CD105 also plays a role in restenosis after stent placement 92, therefore can be used as a marker for the progression of coronary atherosclerosis. We have recently developed a series of CD105-targetted PET and/or optical imaging probes 93-96, which may be applied for PET imaging of atherosclerosis in the near future. In conclusion, PET imaging of atherosclerosis is still at a nascent stage which needs significant future research effort. Smooth clinical translation of promising PET tracers into the clinic is critical to the benefit of patients.

Table 1

Summary of targets and tracers used for PET imaging of atherosclerosis.

Target	Tracer	Preclinical	Clinical
Glucose metabolism	18F-FDG	52, 53, 59	10, 54-58, 60, 63
Membrane metabolism	18F-choline	65	-
	11C-choline	-	67
Integrin αvβ3	18F-galacto-RGD	68	69
LDLr	124I-CD68-Fc-ox-LDL	70	-
	18F-nLDL & ox-LDL	71	-
NPCR	64Cu-DOTA-C-ANF	72, 73	-
Fatty acid synthesis	11C-acetate	-	74
VCAM-1	18F-V4	77	-
Macrophages	64Cu-TNP	78	-
	68Ga	79	-
	11C-PK11195		81
Platelets	18F-AppCHFppA	82	-