Atherosclerosis of the carotid artery is a major cause of ischemic stroke, accounting for approximately 10%-20% of ischemic strokes.1 In patients with carotid artery disease, treatment decisions related to surgical interventions are mainly based on an association with a recent clinical event (symptomatic vs asymptomatic carotid stenosis) and the degree of stenosis. These stratification parameters may not represent the most complete predictors for stroke risk as they only incorporate structural changes of the extracranial vessels. The number-needed-to-treat to prevent one stroke with revascularization further increases due to improvements in best medical therapy. This highlights the need for improved patient selection with additional methods to identify those who are most likely to benefit. Additionally, recent data also suggest that arterial stenosis of less than 50% may play an etiological role in acute ischemic stroke,2 yet these strokes would not classify as a large-artery atherosclerosis stroke subtype based on the TOAST and similar grading systems. High-risk plaque features such as plaque ulceration and plaque thickness were found more often ipsilateral to the index stroke side when compared with the contralateral side in patients with embolic stroke of undetermined source (ESUS).3,4

On a cellular and molecular level, different pathophysiological processes leading to an increase in thrombo-embolic events have been described.5 Inflammation, neovascularization, hypoxia, and microcalcification can contribute to atherosclerotic plaque rupture and thereby stroke.6

Molecular imaging of atherosclerosis has the potential to improve risk stratification by quantifying these physiological processes.7 In combination with structural data, molecular imaging holds promise for targeted prevention and personalized treatment of patients.8 In various stages of atherosclerosis, patients with vulnerable plaques or those at increased risk for progression may be identified.5,9 Additionally, in randomized clinical trials identifying the therapeutic effect of antiatherogenic drugs, molecular imaging may be included as a specific intermediary outcome marker for stroke.10 Probes for molecular imaging include radiopharmaceuticals for PET and single photon emission computed tomography (SPECT) or iron oxide nanoparticles for MRI. These probes are further engineered to target moieties that interact with specific cellular processes and molecular biomarkers of inflammation in atheroma.5,11

18F-FDG is the most studied radiotracer for the identification of infiltration of macrophages in the atherosclerotic plaque. After being taken up by all cells that metabolize glucose, 18F-FDG is phosphorylated and accumulates within the cell as it cannot undergo glycolysis. 18F-FDG uptake corresponds to real-time metabolic activity, which in the inflamed plaque correlates mainly to macrophage activity. This specific property makes 18F-FDG PET an important noninvasive technique to assess inflammation in atherosclerosis.12, 13, 14 18F-sodium fluoride (NaF) is a radiotracer incorporated into areas of calcium deposition capable of detecting microcalcifications in the arterial wall. These microcalcifications are associated with inflammation and may themselves contribute to atherosclerotic plaque rupture. In various retrospective studies using 18F-NaF-PET/CT, NaF uptake was higher in culprit versus nonculprit carotid plaques.15 However, no correlation with arterial 18F-FDG retention was observed which suggests that NaF provides different information regarding plaque vulnerability to 18F-FDG.16,17

These tracers have some limitations. The 18F-FDG signal is not macrophage-specific and is affected by large and variable physiological uptake in the myocardium and brain. In addition, hypoxia influences 18F-FDG uptake. Circulating glucose levels also interfere with 18F-FDG uptake, necessitating a fasting period before imaging.13 As for 18F-NaF, continued calcification may be associated with plaque stability rather than vulnerability, potentially making serial imaging with a calcification-dependent imaging tracer less valuable.14 Therefore more specific tracers have been developed to potentially better characterize the carotid inflammation. In this systematic review, we focus on clinical studies using novel molecular imaging techniques targeting carotid plaque inflammation in relation to structural imaging and 18F-FDG-PET. We discuss these novel tools/targets in relation to other markers of inflammation and predictive potential.

Nonmolecular Imaging of Carotid Atherosclerosis

Computed tomography angiography (CTA) and high-resolution magnetic resonance angiography (MRA) enable the noninvasive monitoring of carotid plaque progression and vulnerability by imaging plaque morphology and composition in addition to anatomy.18 Based on CTA and/or MRA, characteristics that predispose to plaque rupture can be described: the presence of intraplaque hemorrhage, a large lipid-rich necrotic core, and a thin fibrous cap with decreased smooth muscle cell content.19 Contrast-enhanced MRA can provide 3-dimensional images in the absence of radiation or the need for iodinated-based contrast materials. Another advantage of MRA is that it can be combined with MRI of the brain to visualize silent infarction. However, despite intensive research over the past 15-20 years, the use of this technique in clinical practice remains limited both due to its expense as well as its susceptibility to motion artifacts due to long acquisition times.20,21

Ultrasmall superparamagnetic iron-oxides (USPIO) contrast agents are long-circulating nanoparticles initially devised for imaging the reticuloendothelial system.6 USPIOs can access inflammatory lesions, such as atherosclerotic plaques, presumably through dysfunctional endothelium.22 The USPIO is taken up by activated macrophages and concentrated in the phagolysosomes, where it produces signal decrease on T2-weighted MRI.23 Iron sequestration by pro-inflammatory M1 macrophages in inflammation is thought to play a role in USPIO uptake.24 A relationship between USPIO uptake and CD68 staining (a marker of macrophage infiltration) within carotid plaque has been described.25, 26, 27 Two studies on USPIO-enhanced MRI found differences in USPIO uptake between the culprit carotid artery plaque and the asymptomatic contralateral artery.26,28 In patients with stroke due to large vessel disease, asymptomatic plaques contralateral to the symptomatic side showed higher USPIO uptake compared to plaques in individuals without prior stroke. These contralateral asymptomatic plaques showed higher USPIO uptake despite having a lower mean grade of carotid artery stenosis (46% compared to 63% in truly asymptomatic patients).29 A similar association between carotid tracer uptake and cardiovascular disease was found when comparing patients with prior coronary artery disease vs individuals without any cardiovascular disease.30 Only one study prospectively assessed clinical outcomes by follow-up of 62 patients for a median of 48 months to analyze the risk of subsequent cardiovascular events after an initial USPIO-enhanced MRI, but no association with USPIO uptake was demonstrated.31 In a randomized controlled trial, aggressive lipid-lowering therapy resulted in a significant reduction of USPIO-defined inflammation at both 6 (change from baseline in signal intensity of 0.131, P-value = 0.0003) and 12 weeks (0.203, P-value < 0.0001) after treatment initiation.10 No significant correlation was found between 18F-FDG uptake and USPIO-signal change.22 Together, several studies demonstrate the feasibility of using USPIOs as a marker of vulnerable atherosclerotic plaques. However, their use have not yet been incorporated in clinical practice.