The Year in Molecular Imaging

Abstract

Molecular imaging is devoted to the discovery and application of specific biological imaging approaches that complement traditional anatomical imaging. This field continues to witness impressive growth, particularly in the study of oncology, neurology, and cardiovascular disease (CVD). Interest in molecular imaging technologies stems from its great potential, not only to heighten our understanding and diagnostic capability of common CVD scenarios, but also to offer the prospect of personalized treatment and early monitoring of therapeutic response. Targeted imaging reporters are now spawning the development of combined diagnostic and therapeutic agents that can deliver therapy to individual cells in affected tissues. Recently, diagnostic imaging probes for myocardial infarction (MI), stem cell tracking, and atherosclerotic vascular disease have demonstrated significant advances in preclinical research and development (also see the Online Appendix). Clinical molecular imaging studies have further expanded into the areas of aortic dissection and aneurysm disease, and have provided new insights into aspects of heart failure and transplant medicine. In this review, we highlight outstanding CVD molecular imaging studies published over the past year. A summary of important clinical and preclinical imaging agents and applications is provided in Table 1.

Clinical Investigations

A major goal of molecular imaging is to foster translation of innovative technology into the clinical arena. Clinical molecular imaging studies of human atherosclerosis remain prominent. In addition, new strategies have been developed for imaging myocardial disease, as well other areas of vascular disease.

Atherosclerosis

The overarching goal of atherosclerosis molecular imaging is to identify features associated with at-risk (vulnerable) plaques that are prone to rupture. Such plaques may harbor inflammation, hemorrhage, thrombus, neovascularization, or apoptotic cells, aspects suitable for molecular imaging. Studies investigating atherosclerosis based on 18-fluorodeoxyglucose (¹⁸F-FDG) positron emission tomography (PET) imaging of plaque metabolism/inflammation, and possibly hypoxia, lead the efforts in clinical CVD molecular imaging.

¹⁸F-FDG PET imaging of carotid plaque inflammation

¹⁸F-FDG is a radiolabeled glucose analog transported into metabolically active cells, and its presence is correlated with plaque macrophage content. ¹⁸F-FDG PET imaging, therefore, is a surrogate reporter of this critical cell involved in atherogenesis and plaque rupture. On the basis of the commercial availability of the ¹⁸F-FDG tracer and clinical PET scanners, ¹⁸F-FDG PET has become one of the most widely employed clinical molecular imaging strategies to study the presence and evolution of endovascular inflammation, particularly in the larger arterial beds (i.e., carotid, aorta).

CAROTID PLAQUE INFLAMMATION AND RISK OF MICROEMBOLI

Microemboli emanating from carotid plaques are associated with an increased risk of stroke, as well as larger macrophage-rich areas in carotid atheroma. In a cross-sectional study of 16 patients with recent anterior circulation transient ischemic attack or minor stroke and 50% to 99% ipsilateral carotid stenosis, ¹⁸F-FDG PET imaging was performed (1). Carotid microemboli signals (MES) were detected on transcranial Doppler ultrasound (middle cerebral artery, 1-h duration, same day as PET imaging). ¹⁸F-FDG PET discriminated between MES⁺ and MES⁻ plaques (p = 0.005) on the basis of their degree of local inflammation. Notably, plaque percent stenosis did not provide discriminatory power between MES⁺ and MES⁻ plaques in the pilot study (p = 0.48). Prospective studies are indicated to determine whether ¹⁸F-FDG PET imaging of carotid plaque inflammation/metabolism will improve risk prediction of stroke beyond percent stenosis derived from anatomical imaging.

PLAQUE INFLAMMATION AND CARDIOVASCULAR RISK FACTORS

Extending the application of ¹⁸F-FDG PET imaging to patients with diabetes mellitus, Kim et al. (2) studied 90 age- and sex-matched subjects without known cardiovascular disease, who were divided evenly between normal, impaired glucose tolerance, and type 2 diabetes. Images revealed that carotid artery ¹⁸F-FDG uptake increased incrementally with the degree of glucose intolerance, and the mean and maximum plaque target-to-background ratios (TBRs) correlated with several cardiovascular risk factors and metabolic biomarkers. The same subjects stratified by Framingham Risk Scores of <10%, 10% to 20%, and >20% also showed increasing maximum TBRs of 1.2, 2.0, and 2.3, respectively, (p = 0.001). Interestingly, however, concomitantly obtained carotid intima-media thicknesses were similar among all 3 groups, providing further evidence that inflammation cannot be routinely evaluated using structural-based imaging approaches.

RELATIONSHIP BETWEEN CAROTID PLAQUE INFLAMMATION AND PLAQUE ECHOGENICITY

The association between carotid plaque ultrasound echogenicity and ¹⁸F-FDG PET inflammation was evaluated prospectively in 33 patients with symptomatic carotid arterial disease (3). Although echogenic lesions demonstrated minimal ¹⁸F-FDG uptake, echolucent atheroma, a putative ultrasound marker of vulnerable carotid plaques, exhibited a wide range of 18F-FDG–assessed inflammation. Plaque inflammation in vivo correlated positively with CD68⁺ macrophages on endarterectomy histopathology specimens (r = 0.501). Therefore, within a range of suspected high-risk plaques based on echolucency, ¹⁸F-FDG PET may further stratify at-risk lesions by their degree of local inflammation. Stated another way, this study shows that all echolucent plaques are not inflamed. In addition, similar to other studies, stenosis was not significantly associated with the degree of plaque inflammation, or with plaque echogenicity.

PLAQUE INFLAMMATION AND NEW BIOMARKERS

In a validation study, Yoo et al. (4) positively correlated elevated circulating blood adipocyte fatty acid binding protein with carotid ¹⁸F-FDG PET uptake in a cross-sectional study of 87 men without previously diagnosed CVD or diabetes. Multiple regression analysis accounting for other CVD risk factors further identified adipocyte fatty acid binding protein as an independent predictor of ¹⁸F-FDG– determined vascular inflammation. This study demonstrates the ability of molecular imaging strategies to validate potential biomarkers in human subjects.

NATURAL HISTORY OF ¹⁸F-FDG PET INFLAMMATION SIGNALS

A prospective study has demonstrated the stability of 18F-FDG plaque signal over a 2-week period; however, longer-term data are limited. In a retrospective study, Menezes et al. (5) examined the records of 50 patients that had at least 4 PET/computed tomography (CT) scans performed over a mean follow-up period of 27 months. The authors measured aortic and carotid ¹⁸F-FDG PET uptake, as well as CT-determined Hounsfield unit tissue density (i.e., calcification). Notably, no initial ¹⁸F-FDG–positive site remained positive on subsequent PET studies, suggesting that inflammation in a particular plaque may not persist for years. Prospective studies are needed to better understand the natural history of plaque inflammation, and these data may arise from ongoing serial imaging trials assessing the inflammatory-modulating effects of new pharmacotherapeutics.

¹⁸F-FDG PET imaging of coronary plaque inflammation

A noninvasive method to detect coronary arterial inflammation could be highly useful. Major challenges for ¹⁸F-FDG PET imaging include relatively low spatial resolution and high background signal from metabolically active myocardium underlying the coronary bed. Encouragingly, there has been recent progress in this area, aided by dietary manipulations that can suppress background myocardial ¹⁸F-FDG signal.

CORONARY PLAQUE INFLAMMATION IN PATIENTS WITH ACUTE CORONARY SYNDROMES AND STABLE ANGINA

Rogers et al. (6) examined ascending aorta and left main coronary artery ¹⁸F-FDG PET uptake in 25 patients 1 to 2 weeks after presenting with acute coronary syndrome (ACS) or stable angina. In the left main coronary artery, greater ¹⁸F-FDG–based inflammatory activity was present in subjects with ACS compared with stable angina (Fig. 1) (plaque TBR: 2.48 vs. 2.00, p = 0.03). In stented epicardial coronary artery segments, similar findings were observed, with plaque TBRs in ACS patients higher than those with stable angina (TBR: 2.61 vs. 1.74, p = 0.02). Impressively, 96% of coronary PET segments were analyzable after dietary-based suppression of background myocardial signal, including 89% of stented culprit lesions. The results demonstrate that hybrid molecular PET/CT may have utility for noninvasive evaluation of left main coronary arterial inflammation, as the left main is larger, more remote from myocardium, and less mobile, leading to improved image quality and co-registration with coronary CT angiography.

Figure 1. Co-Registered ¹⁸F-FDG PET/CT Imaging of Inflammation in Human Coronary Arteries.

(A) In a patient with an acute coronary syndrome (ACS), the left main (LM) coronary artery and culprit stented lesion show enhanced 18-fluorodeoxyglucose (¹⁸F-FDG) uptake. (B) A patient with stable angina and recent percutaneous coronary intervention reveals less stented segment ¹⁸F-FDG signal. Note also the focal ¹⁸F-FDG signal in the mixed composition LM coronary artery plaque. (C) A lesion stented months earlier in a patient with stable angina shows comparatively modest ¹⁸F-FDG activity. (D) ¹⁸F-FDG uptake at the LM coronary artery trifurcation in a patient with acute coronary syndrome. In panels A to D, the dashed arrow identifies the LM coronary artery and, when present, the solid arrow points to the stented region. CT = computed tomography; PET = positron emission tomography. Reproduced, with permission, from Rogers et al. (6).

LEFT ANTERIOR DESCENDING ARTERIAL INFLAMMATION

In an attempt to extend ¹⁸F-FDG to the proximal coronary arteries, Saam et al. (7) examined ¹⁸F-FDG uptake in the proximal left anterior descending coronary artery (LAD) in a consecutive series of oncology patients. The authors found a modest, but significant, correlation with various cardiac risk factors (R ≈ 0.2), and further noted that patients with a TBR in the upper tertile had a larger calcified plaque burden and higher pericardial fat volume than patients in the lower tertile. However, LAD data from 45% of the patients were not usable due to background myocardial 18F-FDG uptake, limiting image analysis. Further advances, possibly such as electrocardiogram-based triggering of CT acquisitions, are needed to extend ¹⁸F-FDG reliably into the non–left main coronary tree.

Statin pharmacotherapy and 18F-FDG PET plaque activity

The effect of increasing statin dose on ¹⁸F-FDG PET atheroma uptake was prospectively evaluated in 30 coronary artery disease patients randomized to atorvastatin 5 mg versus 20 mg daily (8). All patients were scheduled for percutaneous coronary intervention to treat stable angina symptoms, and none of them had received lipid-lowering medications for at least the previous 12 months. The day after percutaneous coronary intervention, baseline 18F-FDG PET/CT imaging was performed, and the ¹⁸F-FDG activity in the aorta and femoral artery quantified. The patients were then randomized to high- or low-dose statin therapy, with a follow-up 18F-FDG PET imaging study performed 6 months later. The arterial ¹⁸F-FDG TBR was significantly reduced in the high-dose statin arm (aorta baseline TBR: 1.15 vs. 1.05, p = 0.007; femoral baseline TBR: 1.12 vs. 1.02, p = 0.012). Although there were similar significant reductions in low-density lipoprotein and high-sensitivity C-reactive protein levels in both the high-dose and the low-dose statin groups, the low-dose group did not show significant changes in arterial ¹⁸F-FDG uptake. This suggests that higher statin doses may be required to reduce plaque inflammation.

Near-infrared fluorescence imaging of carotid plaque protease activity

Augmented inflammatory protease activity is a marker of vulnerable plaques. Using near-infrared fluorescence (NIRF) imaging systems in concert with protease-activatable NIRF imaging agents, Kim et al. (9) investigated ex vivo matrix metalloproteinase and cysteine protease activity in samples obtained from 56 subjects under-going either carotid endarterectomy or carotid stenting with filter devices. On macroscopic NIRF imaging, enhanced cathepsin B and MMP 2/9 protease activity were identified both in retrieved emboli and in bifurcations of carotid endarterectomy specimens (Fig. 2). Plaques demonstrated unstable histopathological features (rupture with thrombus, hemorrhage, and thin, inflamed cap). Symptomatic patients (ipsilateral stroke within 1 month) harbored plaques with greater protease activity compared with asymptomatic patients. Intriguingly, cathepsin B protease activity was diminished in patients taking statins, consistent with the anti-inflammatory properties associated with this medication class. Protease signal correlated modestly with anatomic stenosis or ultrasound echolucency, a suggested marker of plaque instability, indicating that protease-identified inflammation likely provides independent, complementary data on plaque stability. Although NIRF molecular imaging agents are not yet approved for clinical study, the current data, along with a recent report of a clinical targeted fluorescence imaging trial (10), provide momentum for future clinical testing of NIRF protease sensors in human subjects.

Figure 2. NIRF Imaging of MMP and Cysteine Protease Activity in Resected Human Carotid Plaques.

(A) Matrix metalloproteinase (MMP) near-infrared fluorescence (NIRF) signal colocalizes with hemorrhagic, ulcerated carotid plaque regions (lines 1 to 3), but is only weakly present in uncomplicated segments without these features (line 4). Of note, similar MMP activity is present in echolucent (lines 1 and 2) and echogenic locations (line 3) identified by duplex ultrasonography (Duplex U.S.). (B) Tissue immunohistochemistry (IHC) at representative regions corresponding with lines 1 to 4 in A reveal enhanced CD68⁺ (CD-68) macrophage content that colocalizes with MMP-9 immunoreactivity and macroscopic MMP activity on NIRF imaging. (Inset) MMP-9 staining at the region labeled by a hash mark not apparent in the low power view is evident at higher magnification. (C) MMP-2 immunoreactive regions overlap with that of MMP-9 (compare the regions labeled with an asterisk and a hash mark in B and C). (D and E) MMP-9 colocalizes with MMP activity on NIRF microscopy. The arrowhead in B identifies the location of the magnified region in D and E. White scale bars = 5 mm; black scale bars = 2 mm; blue scale bars = 200 μm. Cy5.5 F-micro= cyanine 5.5 fluorescence microscopy; Intra-OP = intraoperative. Reproduced, with permission, from Kim et al. (9).

Myocardial Infarction

Following irreversible myocardial damage, necrotic muscle is replaced by collagen-dense scar that may promote maladaptive tissue remodeling, myocardial dysfunction, and clinical heart failure and malignant arrhythmias. Typically, the true degree of dysfunction and scar burden are not fully realized until months after the index event, at a time when there may be little reversibility. Therefore, early detection of these changes may allow subsequent intervention to prevent adverse outcomes.

Single-photon emission computed tomography imaging of fibrogenesis post-MI, and its relationship to final infarct size at 1 year

In 10 patients presenting with MI, Verjans et al. (11) serially evaluated changes in interstitial extracellular matrix via an intravenous injection of an integrin receptor– targeted arginine-glycine-aspartic acid–radiolabeled imaging peptide (RIP). RIP has been shown previously to target up-regulated integrin expression (e.g., α_vβ₃) found on myofibroblasts that manufacture collagen. At 3 and 8 weeks post-MI, ^99mTc-RIP–enhanced single-photon emission computed tomography (SPECT) imaging revealed uptake in 70% of patients involving the infarct zone identified on myocardial perfusion imaging. Importantly, RIP-based SPECT signal typically extended be-yond the fixed perfusion defect. At 1-year follow-up, the pattern of early myocardial RIP uptake was observed to be similar to the final infarct size as measured by cardiac magnetic resonance (CMR) late gadolinium enhancement (Fig. 3). RIP imaging early after infarction may therefore predict the final extent of post-MI myocardial scarring, and offer opportunities for earlier intervention.

Figure 3. Clinical SPECT ^99mTc-RIP Imaging of Myofibroblast Activity 3 Weeks Post-MI, With Follow-Up CMR-Measured Myocardial Fibrosis at 1 Year.

(Top) In 2 different patients, scar in the left anterior descending coronary artery distribution (yellow arrows) is identified with late gadolinium-enhanced cardiac magnetic resonance (CMR) at 1 year. (Middle) ^99mTc-radiolabeled imaging peptide (RIP) uptake (blue arrows) in similar views for the same patients months earlier, with yellow arrows placed for comparison to the CMR-identified scar. Single-photon emission computed tomography (SPECT)/CMR fusion images are shown in the bottom row. MI = myocardial infarction; SA = short axis; VLA = vertical long axis. Reproduced, with permission, from Verjans et al. (11).

Heart Failure and Transplant Medicine

Sarcoid cardiomyopathy and inflammation imaging by ¹⁸F-FDG PET

Congestive heart failure is a wide-reaching condition and a leading cause of hospitalization and death. Sarcoid cardiomyopathy is a relatively uncommon, but important, cause of heart failure that may be successfully treated if detected early. Tahara et al. (12) studied cardiac inflammation in 24 patients with systemic sarcoid disease, of which half had recognized cardiac involvement by endomyocardial biopsy or clinical criteria. Compared with healthy normal subjects or age-matched controls with dilated cardiomyopathy, fasting ¹⁸F-FDG uptake in cardiac sarcoidosis patients showed significantly more heterogeneous myocardial signal. Heterogeneity was quantified by the coefficient of variation, defined as the ¹⁸F-FDG standard deviation divided by the mean uptake in a 17-segment cardiac model. At a cutoff coefficient of variation value of 0.18, this metric demonstrated excellent operating characteristics, with 100% sensitivity and 97% specificity. Following corticosteroid therapy, serial repeat 18F-FDG PET assessment over 12 months follow-up confirmed that the 18F-FDG uptake variation returned to that of control subjects (Fig. 4). Thus, 18F-FDG PET imaging might provide useful diagnostic and treatment responsiveness in patients with sarcoid cardiomyopathy.

Figure 4. Serial Cardiac ¹⁸F-FDG PET Imaging in Cardiac Sarcoidosis Patients.

(A) Compared with the strong baseline signal heterogeneity (high COV), a patient treated with corticosteroids reveals complete resolution of heterogeneous myocardial ¹⁸F-FDG activity (low COV). (B) An untreated patient at a similar follow-up time point exhibits more heterogeneous ¹⁸F-FDG uptake. COV = coefficient of variation; Rx = treatment; other abbreviations as in Figure 1. Reproduced, with permission, from Tahara et al. (12).

Detecting infected left ventricular assist devices using integrated SPECT/CT of leukocyte foci

The noninvasive diagnosis of left ventricular assist device (LVAD) infection, a potentially life-threatening condition, remains challenging. Imaging modalities such as CT or magnetic resonance imaging (MRI) are limited due to metallic artifacts and contraindications. A noninvasive imaging approach may avoid additional testing and treatment that in certain cases may be costly and unnecessary. Furthermore, such an imaging approach could help assess the response to antibiotic therapy in the LVAD itself. Within this context, Litzler et al. (13) performed a clinical cellular imaging study in 8 consecutive patients with suspected LVAD infection, using ^99mTc-exametazime–labeled radioactive leukocytes (≥99% viability with 90% labeling efficacy) and hybrid SPECT/CT imaging. At 4 and 24 h after injection of the radiolabeled autologous leukocytes, pockets of infection were identified in all patients. Images discriminated superficial from deep infection, and depicted the extent of LVAD involvement and any remote involved sites. On the basis of the location and severity of the ^99mTc-labeled leukocyte uptake, the duration of antibiotic therapy could be effectively tailored, with follow-up scans in 5 subjects demonstrating infection resolution in the LVAD or in a non–LVAD-related source.

Aortic Disease

Abdominal aortic aneurysm

Inflammation and weakening of the vessel wall may lead to frank rupture, especially in aneurysms that rapidly expand or reach a critical size. New molecular imaging methods are being directed to evaluate aneurysm markers of instability such as inflammation, hemorrhage, and neovascularization.

CMR OF MACROPHAGES IN AORTIC ANEURYSM

Fourteen patients with infrarenal abdominal aortic aneurysm (4.0- to 5.5-cm diameter) underwent CMR of aneurysm inflammation following intravenous injection of Sinerem (ferumoxtran, Guerbet, Villepinte, France) 2.6 mg Fe/kg, an ultrasmall superparamagnetic iron oxide (USPIO) nanoparticle agent phagocytosed by resident macrophages (14). Compared with pre-USPIO injection, post-injection CMR aneurysm signals were significantly lower in macrophage-rich atherosclerotic areas. If validated prospectively, USPIO-enhanced CMR could identify high-risk subjects with aneurysm inflammation more likely to expand or rupture, motivating increased surveillance and/or earlier treatment.

A similarly sized study by Nchimi et al. (15) detected biologically active macrophages within intraluminal thrombi of abdominal aortic aneurysms via the superparamagnetic iron oxide (SPIO) agent Endorem (ferumoxide, Guerbet, Villepinte, France). Compared with USPIO, SPIO have relatively larger size and shorter circulation half-life due to rapid hepatic clearance, restricting distribution to luminal surface targets and other well-perfused tissues. Exploiting these properties, SPIO localized in leukocyte-dense intraluminal thrombi 1 h after injection, much earlier than the 24 to 36 h necessitated by longer-circulating USPIOs. Ex vivo CMR and histopathology of excised surgical specimens obtained within 2 weeks of in vivo imaging confirmed thrombus inflammatory cell infiltration with iron particle retention, as well as elevated matrix metalloproteinase 9 (MMP-9) enzymatic activity.

Aortic dissection

Disruption of the aortic wall is a life-threatening event that often requires immediate surgical intervention. However, autopsy series indicate that incidental aortic dissection is present in up to 1% to 2% of subjects, suggesting that many chronic aortic dissections can be managed conservatively. Because anatomic imaging modalities do not reliably distinguish acute from chronic dissections, new imaging strategies are needed.

¹⁸F-FDG PET IMAGING OF INFLAMMATION IN ACUTE VERSUS CHRONIC AORTIC DISSECTION

Reeps et al. (16) evaluated ¹⁸F-FDG PET imaging to discriminate acute from chronic type B aortic dissection in 18 patients. ¹⁸F-FDG enhancement was observed at the injured aortic wall in all acute dissection patients 3 to 13 days after presentation (Fig. 5). Quantitative analysis revealed that the standardized uptake value (SUV) ratio, or TBR, performed significantly better in discriminating acute from chronic stable dissection (SUV ratio: 2.20 vs. 1.24, respectively, p < 0.001) than simply measuring maximum SUV at the site of injury, which suffered from false-negative and false-positive outliers. This preliminary study suggests that 18F-FDG PET may help distinguish acute from chronic aortic dissection on the basis of the degree of vessel wall inflammation.

Figure 5. ¹⁸F-FDG PET Imaging of Arterial Wall Inflammation in a 37-Year-Old Male With an Acute Type B Aortic Dissection.

(A) Sagittal and (C) coronal hybrid PET/CT fusion images with corresponding (B) sagittal and (D) coronal CT reconstructions. Abbreviations as in Figure 1. Reproduced, with permission, from Reeps et al. (16).

¹⁸F-FDG AND AORTIC DISSECTION PROGNOSIS

The prognostic value of ¹⁸F-FDG uptake in medically managed acute aortic dissection was addressed in 28 patients. PET scans were obtained approximately 2 weeks after CT diagnosis (17). All patients had type B dissections except for 2 poor surgical candidates with type A. Over 6 months, there were 8 unfavorable outcomes (death, surgical repair, or dissection progression), predicted by the initial inflammatory 18F-FDG uptake at the dissection site with an odds ratio of 7.72 for adverse events on multivariate analysis. ¹⁸F-FDG provided a sensitivity and specificity of 75% and 70%, respectively, for mean SUV > 3.029. Furthermore, in patients with favorable outcomes, a lower ¹⁸F-FDG uptake corresponded to a greater likelihood of dissection healing or regression. These results provide intriguing preliminary evidence that 18F-FDG PET imaging may risk-stratify aortic dissection populations that benefit from aggressive earlier surgical intervention.

Vasculitis

Giant cell arteritis and Takayasu arteritis are large-vessel inflammatory diseases that can result in arterial occlusion, thrombosis, aneurysm, and accelerated atherosclerosis. New approaches are needed to understand the pathogenesis of these diseases in vivo.

PET IMAGING OF MACROPHAGES IN VASCULITIS USING [¹¹C]-PK11195

Pugliese et al. (18) tested the PET imaging probe [¹¹C]-PK11195, an established human neuroimaging agent that selectively binds the benzodiazepine receptor on activated macrophages. Six patients with giant cell arteritis or Takayasu arteritis were compared with 9 asymptomatic controls. Symptomatic subjects exhibited enhanced [¹¹C]-PK11195 signal on vascular PET scans (TBR: 2.41 vs. 0.98 for asymptomatic patients, p = 0.001). In 1 patient treated with 20 weeks of corticosteroid therapy, a follow-up [¹¹C]-PK11195 PET scan demonstrated markedly reduced vascular PET signal (TBR decreased from 1.63 to 0.87), associated with symptomatic improvement and decreased inflammatory biomarkers. Compared with 18F-FDG, which reports on all metabolically active cells, [¹¹C]-PK11195 PET may provide a more macrophage-specific readout, given its ligand selectivity. In addition, [¹¹C]- or [¹⁸F]-PK11195 may also improve coronary atherosclerosis PET imaging of inflammation, because background uptake by metabolically active myocytes should be lower than in ¹⁸F-FDG PET imaging.

Advances in Preclinical Imaging Approaches

Atherosclerosis

In the last year, new atherosclerosis-targeted imaging agents and devices have been developed to sense plaque inflammation, angiogenesis, and oxidative stress. In vivo studies of pharmacotherapeutic efficacy, as well as the area of integrated therapeutic and diagnostic targeted imaging (“theranostics”), particularly via nanotechnological approaches, are also in a growth phase.

Inflammatory macrophages

Macrophages are pivotal cells in atherogenesis, driving plaque expansion and plaque complications. Macrophages are the most established imaging target for atherosclerosis molecular imaging studies, with clinical detection of plaque macrophages enabled by ¹⁸F-FDG PET and USPIO-enhanced CMR. Although these approaches have proven valuable, new reporter agents are diversifying imaging platforms and aiming to improve sensitivity.

PET AGENTS

The positron emitter [¹¹C]-PK11195 has been studied to image activated macrophages in clinical vasculitis, as discussed in the previous text. ¹¹C-choline, a source of cell membrane lipids found in high levels in proliferating cells, including plaque macrophages (19), was shown to colocalize with murine plaques by autoradiography, and was able to distinguish noninflamed and inflamed, macrophage-rich plaques (20). A head-to-head comparison with ¹⁸F-FDG may be informative in understanding any potential advantages with respect to achievable TBRs.

SPECT AGENTS

Folate receptor beta is expressed on activated macrophages and thus offers another molecular imaging target for atherosclerosis. The SPECT tracer ^99mTc-labeled folate-targeted probe (^99mTc-EC20) was investigated in murine atherosclerosis (21). In vivo SPECT imaging demonstrated greater SPECT signal in animals on a high-cholesterol diet compared with a regular chow diet. Uptake of ^99mTc-EC20 was also reduced in animals treated with clodronate liposomes, a macrophage apoptosis–inducing therapeutic agent. Whether the additional macrophage specificity conferred by a folate receptor–based SPECT approach is advantageous compared with ¹⁸F-FDG remains to be tested in clinical subjects.

CMR AGENTS

Based on the natural affinity of macrophages for ferritin, targeted imaging agents comprised of bioengineered polypeptide apoferritin (iron-free ferritin) cages were developed as NIRF or CMR agents by coupling apoferritin to the NIR fluorochrome cyanine 5.5 or to magnetite nanoparticles, respectively (22). In diabetic, high-fat diet– fed mice, the imaging agents localized to the macrophage-rich lesions of ligated carotid arteries. Loading of other imaging agents or even therapeutic drugs into the apoferritin protein cage architecture is possible, providing versatility for this macrophage-avid agent.

Maiseyeu et al. (23) synthesized gadolinium-containing anionic vesicles comprised of phosphati-dylserine and a novel synthetic oxidized cholesterol ester derivative that avidly binds low-density lipoprotein. Reporter vesicles were internalized by macrophages via scavenger receptors in culture, and in rabbit atheroma in vivo, as shown by 1.5-T CMR. This liposomal agent can also be formulated with fluorescently labeled lipids, and has potential advantages for a good safety profile, easy manufacturing, and scalability that could facilitate clinical testing.

DUAL-MODAL PET-MRI REPORTERS

Extending prior work with PET-MRI high-resolution/high-sensitivity macrophage reporters based on 64Cu-labeled crosslinked iron oxide (CLIO-⁶⁴Cu), Jarrett et al. (24) tested CLIO-⁶⁴Cu, as well as gadolinium/⁶⁴Cu-labeled maleylated bovine serum albumin for targeting of the macrophage scavenger receptor A (SRA). Macrophages in injured carotid arteries in 3 different rodent models were visualized with both ⁶⁴Cu-based PET reporters, as well as by 7-T MR (T1-weighted for gadolinium-based agents, T2-weighted for CLIO-based agents) for higher-resolution imaging.

CT AGENTS

Although coronary CT imaging has demonstrated substantial clinical utility, few molecular imaging agents are available for CT platforms. Using a high-density lipoprotein (HDL)-based macrophage-targeted gold nanoparticle coupled with spectral (“multicolor”) CT, a multienergy imaging approach based on the discrimination of x-ray energy attenuation differentials, Cormode et al. (25) imaged plaque macrophages in apoE^−/− mice. Capitalizing on the multispectral capabilities, in addition to macrophage detection, simultaneous co-registered images of calcification (bones) and iodinated contrast (angiography) were obtained in the same mice (Fig. 6). These results expand the selection of current nanoparticle CT agents for macrophages (e.g., N1177), and may accelerate the use of CT as a molecular and anatomical imaging approach for coronary atherosclerosis.

Figure 6. Thoracoabdominal Spectral CT Imaging of Murine Atherosclerosis via the Gold-Based Nanoparticle Au-HDL.

(A–C) Spectral computed tomography (CT) images in an apolipoprotein-negative (apoE^−/−) mouse after Au-HDL injection 24 h earlier. (D and E) Spectral CT images in the aortic bifurcation after co-injection of Au-HDL and iodinated contrast. HDL = high-density lipoprotein. Reproduced, with permission, from Cormode et al. (25).

Myocardial Infarction

Post-MI healing involves an orchestrated immune response. Hypo- or hyperimmune cell infiltration following MI can impair myocardial healing, and may promote mechanical complications (rupture) or adverse left ventricular (LV) remodeling (progressive dilation). Methods to study key effector cells and molecules in post-MI healing may shed important insight into this process in vivo, and potentially offer new therapeutic strategies.

Inflammation and myocardial healing

THE EFFECTS OF HYPERLIPIDEMIA ON POST-MI INFLAMMATION AND HEALING

In wild-type and apolipoprotein E–negative (apoE^−/−) hyperlipidemic mice, Panizzi et al. (26) first quantified monocyte recruitment post-MI by mapping the pro- and anti-inflammatory subtypes Ly-6C^hi and Ly-6C^lo via flow cytometry. ApoE^−/− mice showed significantly increased levels of inflammatory Ly-6C^hi monocytes, cells that secrete degradative enzymes. Using noninvasive, dual-channel fluorescence molecular tomography–CT molecular imaging of post-MI mice (Fig. 7), the authors found that in vivo inflammatory cathepsin enzymatic activity (activatable agent ProSense680, PerkinElmer, Waltham, Massachusetts) and macrophage activity (reporter agent CLIO-750) was also greater in apoE^−/− mice (26). Atherosclerotic mice had 10-fold greater Ly-6C^hi monocyte infiltration in the infarct zone 5 days post-MI than wild-type controls, as assessed by histopathology, and was associated with increased inflammatory markers, proteolysis, and phagocytosis. In addition, serial CMR showed that apoE^−/− mice with MI developed greater ventricular dilation compared with wild-type mice, despite having similar infarct sizes initially. Furthermore, induction of Ly-6C^hi monocytosis alone by intraperitoneal lipopolysaccharide injection in wild-type mice without hyperlipidemia or atherosclerosis recapitulated a similar post-MI inflammatory state. This work concludes that the Ly-6C^hi monocyte may be a therapeutic cellular target in the post-MI setting.

Figure 7. Noninvasive FMT/CT Dual-Target Molecular Imaging of Inflammation During Infarct Healing in Atherosclerotic and Control Mice.

(A) Five days post MI, CT images reveal apical infarct (arrows). Corresponding fusion 2-dimensional and 3-dimensional in vivo fluorescence molecular tomography (FMT)/CT images of inflammatory activity (ProSense680 for cysteine protease activity, and CLIO-750 for infarct macrophages) in control C57Bl/6 mice (top) and apoE^−/− mice (bottom). (B) Ex vivo fluorescence reflectance imaging corroborates the in vivo FMT/CT imaging results. *p < 0.05. TBR = target-to-background ratio; other abbreviations as in Figures 1, 3, and 6. Reproduced, with permission, from Panizzi et al. (26).

Please see the Online Appendix for additional high-impact preclinical imaging studies.

Outlook

Molecular imaging studies are shedding important light on the cellular and molecular biology underlying important cardiovascular diseases. Excitingly, this last year witnessed a growth in clinical molecular investigations, particularly in the areas of atherosclerosis, aneurysm disease, and MI. In the short term, we anticipate continued expansion of ¹⁸F-FDG PET– based molecular imaging applications of vascular disease. Insights from 18F-FDG PET should prove useful in predicting which atherosclerosis therapeutics will likely be efficacious and safe. Across many areas of CVD, molecular imaging data studies will continue to elucidate the in vivo pathogenesis of disease, and will provide the foundation for improved risk assessment of key CVD events, such as acute plaque rupture and adverse LV remodeling.

Due to lower imaging agent dose requirements, new strategies for PET and SPECT may be clinically available in the next 2 years. In addition, select fluorescence agents may enter the clinical arena within the next 1 to 3 years. These trends will enable new clinical molecular imaging studies of CVD, and will be spurred further by advances in clinical imaging systems, especially in the emerging areas of fluorescence imaging and multispectral CT imaging. As more agents transition from the bench to the patient bedside, molecular imaging studies should prove helpful in the clinical management of CVD.

Table 1.

Promising Agents for Molecular Imaging of Cardiovascular Disease

Agent	Modality	Primary Target	Application	Clinical
18F-FDG	PET	Glucose transporter-1, hexokinase	Atherosclerosis, aneurysm (metabolism)	Yes
USPIO	CMR	Macrophages	Atherosclerosis (inflammation)	Yes
99mTc-annexin A5	SPECT	Annexin-A5/macrophages	Atherosclerosis, myocardial infarction  (apoptosis)	Yes
Au-HDL	Spectral CT	Macrophages	Atherosclerosis (inflammation)	No
CLIO-64Cu	PET + CMR	Macrophages	Atherosclerosis (inflammation)	No
99mTc-EC20	SPECT	Folate receptor/macrophages	Atherosclerosis (inflammation)	No
Apoferritin cages	CMR, NIRF	Macrophages	Atherosclerosis (inflammation)	No
Synthetic Gd anionic vesicles	CMR, NIRF	Macrophages	Atherosclerosis (inflammation)	No
L-PLP	CMR	Macrophages	Atherosclerosis (inflammation) [Theranostic]	No
CLIO-THPC-AF750	CMR, NIRF	Macrophages	Atherosclerosis (inflammation) [Theranostic]	No
N1177	CT	Macrophages	Atherosclerosis (inflammation)	No
ProSense	NIRF	Cysteine proteases	Atherosclerosis (inflammation)	No
MMPsense	NIRF	Matrix metalloproteinases	Atherosclerosis (inflammation)	No
99mTc-RP805	SPECT	Matrix metalloproteinases	Atherosclerosis (inflammation)	No
P947	CMR	Matrix metalloproteinases	Atherosclerosis, aneurysm (inflammation)	No
αvβ3-targeted nanoparticles	CMR, NIRF	αvβ3 integrins	Atherosclerosis (neovascularization)	No
99mTc-RIP	SPECT	αvβ3 integrins	Myocardial infarction (fibrosis)	Yes
cNGR peptide–labeled paramagnetic  quantum dot	CMR	CD13 aminopeptidase	Myocardial infarction (neovascularization)	No
111In-RP782	SPECT	Matrix metalloproteinases	Aneurysm, myocardial infarction, vascular  injury (inflammation)	No
EP-2104R	CMR	Fibrin	Thrombosis (coagulation)	Yes
NanoK	Spectral CT	Fibrin	Thrombosis (coagulation)	No
P975	CMR	GPαIIbβ3 receptor/platelets	Thrombosis (platelet activity)	No
[11C]-PK11195	PET	Benzodiazepine receptor/macrophages	Vasculitis (inflammation)	Yes
OsteoSense	NIRF	Hydroxyapatite	Valvular disease (calcification)	No
HSV-tk	PET	Herpes simplex virus thymidine kinase	Stem cell therapy (cell tracking)	No
99mTc-exametazime–labeled leukocytes	SPECT	White blood cells	Heart failure/transplant (inflammation)	Yes
MPO-Gd	CMR	Myeloperoxidase activity	Heart failure/transplant (oxidative stress)	No

¹⁸F-FDG = 18-fluorodeoxyglucose; CLIO = crosslinked iron oxide; CMR = cardiac magnetic resonance; cNGR = cyclic Asn-Gly-Arg; CT = computed tomography; Gd = gadolinium; HDL = high-density lipoprotein; HSV-tk = herpes simplex virus thymidine kinase; L-PLP = liposomal prednisolone phosphate; MPO = myeloperoxidase; NIRF = near-infrared fluorescence; PET = positron emission tomography; RIP = radiolabeled imaging peptide; SPECT = single-photon emission computed tomography; USPIO = ultrasmall superparamagnetic iron oxide.