Molecular Imaging of Atheroma: Deciphering How and When to Use 18F NaF and 18F FDG

Despite significant therapeutic advances, atherosclerotic disease (i.e. stroke and myocardial infarction) remains the largest cause of death in the United States. Molecular imaging provides insights into the biological processes within the atherosclerotic plaque, improves risk stratification and enhances the assessment of novel therapies. Accordingly, molecular imaging has the potential to facilitate further gains in cardiovascular care. Today, two radiotracers are widely available to assess atherosclerotic plaque biology: 18F sodium fluoride (NaF), and 18F fluorodeoxyglucose (FDG). While there is overlap in their potential applications, important differences exist between them. It is critically important to understand the relative strengths and limitations of NaF and FDG in order to know how and when to optimally use them.

FDG is a glucose analogue that enters living cells via glucose transporters. Soon after the radiotracer enters the glycolytic pathway, itis phosphorylated by hexokinase and is thereafter unable to be further metabolized via glycolysis. Accordingly, its accumulation in tissues, which can be measured non-invasively using positron emission tomography (PET) with computed tomography (CT) or magnetic resonance (MR), provides a quantitative measure of glucose utilization.^{1, 2} Clinically, FDG PET is frequently performed to assess for malignancies and myocardial viability. Another important use of FDG PET imaging is the identification of inflammatory foci. FDG PET imaging of inflammation leverages the fact that inflammatory cells, notably activated macrophages and neutrophils, have particularly high glycolytic rates.³ To that end, cardiac FDG PET imaging is clinically used to evaluate for the presence of cardiac sarcoidosis,⁴ prosthetic valve endocarditis,⁵ and suspected implanted device infections.⁶ Moreover, since inflammatory cells are among the most metabolically active cells in atheroma, FDG PET is also useful for measurement of arterial inflammation. Indeed, human studies have repeatedly shown that assessment of arterial FDG uptake provides a reliable index of arterial inflammation.^7–9

Given the importance of inflammation to the pathobiology of atherosclerosis, the presence of arterial inflammation has a number of implications and uses. Arterial FDG uptake has been shown to co-localize to plaques with high-risk features¹⁰ and to identify atheromas that are likely to progress.¹¹ Moreover, arterial FDG uptake provides an independent measure of CVD risk that is incremental above the Framingham Risk Score or the extent of coronary artery calcium.¹² Furthermore, since atherosclerotic inflammation can be modified by anti-inflammatory therapies, arterial FDG uptake is, likewise, modifiable by therapeutic interventions. The observation of reduced arterial inflammation has potentially important implications; short-term decreases in arterial FDG uptake presage long-term decreases in the rate of progression of the underlying atheroma.¹³ Many studies have employed arterial FDG imaging to evaluate the impact of anti-atherosclerotic drugs on atherosclerotic plaque biology. However, only five classes of drugs have been subjected to both FDG PET imaging to evaluate the impact on atherosclerotic inflammation and clinical end-point trials to evaluate their impact on CVD events. Thus far, for each of these classes, there is apparent concordance between imaging and clinical outcomes.^{1, 14}

However, the uptake of FDG is not specific to inflammatory cells. Several tissues, including the myocardium, bone marrow, and brain take up FDG. This limits the tracer’s utility to evaluate arterial inflammation in the coronaries, as background myocardial FDG uptake may be difficult to suppress and can interfere with coronary assessment (aside from the non-epicardial left main).¹⁵ Interestingly, the fact FDG can be simultaneously assessed in several tissues can also be leveraged. A recent study demonstrated a link between metabolic activity in the amygdala to increased arterial inflammation and CVD events.¹⁶ This observation of a neural-hematopoietic-arterial axis was enabled by the simultaneous assessment of metabolism across tissues, providing novel insights into the mechanisms linking psychosocial stress to CVD. Accordingly, FDG PET imaging has a wide range of potential uses for evaluating the cardiovascular system, including imaging myocardial metabolism, evaluating inflammatory and infectious foci, delineating atherosclerotic inflammation, refining risk, assessing likelihood of atheroma progression, and evaluating treatment response.

Another commonly employed radiotracer, NaF, is used to identify metastases to bone, where it binds to hydroxyapatite. This affinity results in the accumulation of NaF in areas of active micro-calcification within the atheroma. Vascular calcification has long been known to associate with cardiovascular risk. Larger deposits of calcium can be detected with CT. However, micro-calcifications (< 50 μm) cannot be measured using CT; hence, NaF imaging has the potential to provide unique prognostic information that is not available from CT.

Arterial NaF uptake is increased in high-risk and/or symptomatic atheroma. In a study of patients with recent myocardial infarction, NaF uptake was higher in culprit than in non-culprit plaques,¹⁷ indicating a useful niche for identifying recent culprits. Moreover, the relatively low uptake by myocardial cells renders the tracer more useful for coronary imaging than FDG.¹⁸

In this issue of Circulation: Cardiovascular Imaging, Vesey et al report a case-control study examining the use of NaF and FDG PET/CT after transient ischemic attack or minor stroke. They evaluated 26 patients following a recent neurovascular event. Eighteen patients were found to have a culprit carotid stenosis awaiting carotid endarterectomy and eight “controls” lacked an identifiable carotid atheromatous culprit. All individuals underwent PET/CT scanning using FDG and NaF. In the subset of patients who underwent carotid endarterectomy, histological analysis was performed on the excised plaques.¹⁹

The authors observed increased NaF uptake within culprit lesions in comparison to the contralateral artery and arteries from “controls”. They also note that NaF uptake correlated with cardiovascular risk and high-risk plaque features (i.e. remodeling index and plaque burden). For FDG, there also was an association with cardiovascular risk. However, while FDG uptake appeared to be visually increased in several culprit carotid lesions, there was no statistically significant increase within culprit vessels versus contralateral or “control” vessels. The authors conclude that NaF PET/CT imaging appears useful for identifying culprit and phenotypically high-risk carotid plaque. In comparison, the authors found that FDG did not reliably identify culprit plaques and did not correlate with high-risk morphological features.

Several limitations should be considered while interpreting the study’s findings. One important limitation derives from the construction of the control group; each of the control subjects had a recent stroke. This is relevant, as animal and human studies have confirmed that arterial inflammation and FDG uptake increase in non-culprit lesions after strokes or myocardial infarctions.^{20, 21} Accordingly, arterial inflammation is expected to be increased in non-culprit vessels of recently symptomatic individuals, hence reducing the difference between culprit and non-culprit vessels. Had the authors chosen controls that were not recently symptomatic, they would likely have found a larger difference between symptomatic and asymptomatic vessels. Further, the authors reported a high proportion of uninterpretable carotid FDG PET/CT images, (5/16, or 31%), which is quite unusual and likely limited power to test their hypothesis. These limitations notwithstanding, the study does provide some useful observations. Namely, it once again demonstrates that NaF nicely accumulates within culprit lesions, thus highlighting a particularly useful feature of this tracer.

So, how best to use NaF and/or FDG for imaging atheroma? There is compelling data to demonstrate that NaF effectively accumulates within both carotid and coronary culprit plaques, while FDG might not be as reliable for that purpose. On the other hand, FDG provides a reproducible measure of arterial inflammation, which is useful for assessing overall risk and for investigating the impact of therapies targeting arterial inflammation. Molecular imaging of atheromatous plaques is maturing; a growing number of approved and available tracers are providing unique insights into an individual’s CVD biology.² As these molecular imaging techniques become more widely employed, our approach to using the various tools needs to become more sophisticated.