FormalPara Key Summary Points

Early diagnosis of atherosclerosis is a clinical necessity in order to allow for timely intervention before irreversible changes have taken place.

18F-fluorodeoxyglucose (FDG) is a PET tracer with many uses and was first described as a marker of vascular inflammation associated with atherosclerosis over 20 years ago.

More recently, PET imaging with 18F-sodium fluoride (NaF) has gained attention for the purpose of imaging vascular micro-calcification, which is a component of atheroma progression.

Although both FDG and NaF both appear promising for the assessment of atherogenesis, the close association between NaF uptake and established cardiovascular risk factors as well as increased specificity of NaF uptake allowing for visualization of coronary artery involvement suggest that NaF may be the optimal PET tracer for atherosclerosis imaging.

Introduction

The 2020 American Heart Association (AHA) statistical update reported that cardiovascular diseases (CVD) are attributed to more than 1000 deaths daily worldwide [1]. Atherosclerosis, a multi-factorial and progressive process, is a major cause of CVD and related complications, including stroke and peripheral vascular disease [1]. Medical and lifestyle interventions are only effective at reversing disease until a certain point; therefore, it is of utmost importance to detect atherosclerosis in the early stages.

The development of atherosclerosis usually begins with endothelial cell dysfunction, triggered by underlying cardiovascular risk factors including hyperlipidemia, aging, hypertension, or diabetes. This is followed by accumulation of proinflammatory cells and fatty streak formation [2]. Later, molecular micro-calcification of atheromatous plaque forms and progresses into structural macro-calcification, gradually leading to high risk of plaque rupture and vessel occlusion [3]. In the past, numerous conventional imaging techniques have been used in identifying macro-calcification, including cardiac magnetic resonance imaging (CMR), coronary computed tomographic angiography (CTA), intravascular ultrasonography, optical coherence tomography, and coronary artery calcium score (CACS) via CT; however, none of these modalities demonstrate the ability to detect micro-calcification in the early stages of the disease [1, 4].

Molecular imaging with positron emission tomography (PET), more common in the clinical setting as the hybrid modality PET/computed tomography (CT), is capable of detecting early-stage disease by identifying specific changes in metabolism [5]. To date, 18F-fluorodeoxyglucose (FDG) and 18F-sodium fluoride (NaF) are among the two most widely studied PET tracers in atherosclerosis imaging (Fig. 1) [6,7,8]. In addition, many other tracers are in current use within cardiac molecular imaging [9]. We aim to review the properties and clinical applications of emerging PET radiotracers in nuclear cardiology, with an emphasis on early detection of atherosclerosis with FDG and NaF. This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

Fig. 1
figure 1

The development of a complicated atheromatous plaque with hypothesized changes visible by imaging. Although it is established that NaF uptake precedes coronary artery calcium (CAC) visible on CT, new data suggest that NaF uptake may be present earlier relative to inflammation portrayed by FDG than previously thought (blue arrows). Image reprinted without changes from McKenney-Drake et al. [8] under the Creative Commons Attribution 4.0 International License (CC BY). https://creativecommons.org/licenses/by/4.0/

Full size image

18F-Fluorodeoxyglucose (FDG)

Myocardial Injury and Inflammation

Myocardial tissue primarily utilizes fatty acids as source of energy, but under ischemic conditions, myocardial cells consume glucose as their primary energy source. Cardiomyocytes with diminished blood supply can undergo an irreversible pattern of infarction and scar tissue formation with total loss of contractile abilities or can bear the stress with temporary, reversible stunning with loss of contractile properties, which is considered hibernating myocardium [10].

As the most widely used PET tracer with major applications in oncology, neurology, infection, and inflammation, FDG serves as a glucose analog and becomes trapped within cells that demonstrate a high rate of glycolysis [11]. This occurs when FDG is converted to FDG-6-phosphate by hexokinase, accumulating intracellularly without being further metabolized. Since the 1980s, FDG has been used to characterize myocardial viability in patients with coronary artery disease [12]. Clinically, FDG cardiac imaging is routinely performed in combination with myocardial perfusion imaging using either positron-emitting or single photon-emitting radiotracers. Ischemic cardiomyopathy can be associated with either myocardial scar or hibernating myocardium. Preserved myocardial perfusion and viability are seen in healthy tissue, whereas both are reduced in non-viable areas of myocardium. A region with preserved (or elevated) FDG uptake but reduced myocardial perfusion indicates viable hibernating myocardium, which can be treated by vascular intervention [13].

FDG uptake in the myocardium has also been used to assess for cardiac sarcoidosis, where it was found to be superior to perfusion single-photon emission computed tomography (SPECT) and delayed enhanced CMR to not only diagnose but also monitor therapeutic responses [14]. Methods of suppressing physiologic myocardial glucose uptake prior to FDG-PET imaging have become a standard practice to facilitate identification of abnormal activity. Specifically, low-carbohydrate and ketogenic dietary protocols have been shown to improve detection of cardiac sarcoidosis [15,16,17]. Further metabolism suppression can be accomplished with the administration of heparin prior to imaging [18]. Hybrid imaging with FDG-PET/magnetic resonance imaging (MRI) has been proposed to assess cardiac sarcoidosis, representing a combination of two sensitive modalities to assess for myocardial involvement. PET-based methods can help differentiate between acute and post-inflammatory reactions in myocarditis and may also guide endomyocardial biopsies or tailored treatment strategies, such as escalating immunosuppressive therapies during acute exacerbations of inflammation.

Suspected cardiac infections, which are often FDG avid due to inflammatory cell activation, may warrant further evaluation by PET in cases of inconclusive echocardiography [19]. While the sensitivity and specificity of FDG-PET/CT are both high for the diagnosis of prosthetic valve endocarditis, the sensitivity is somewhat lower for native valve endocarditis [20, 21]. High sensitivity and prognostic utility of FDG-PET/CT has been demonstrated in the setting of left ventricular assist device infections [22, 23]. Overall, there is increasing acceptance for clinical use of FDG-PET/CT for the evaluation of infection, and the endorsement of FDG-PET/CT by the European Heart Rhythm Association as a major criterion in the diagnosis of cardiac implantable electronic device infection emphasizes the fact that cardiac infections are not excluded from this trend [24].

Vascular Inflammation

The significance of vascular uptake of FDG as a potential indicator of atherosclerotic disease was first observed by Yun et al. in 2001 [25]. In the context of atherosclerosis, FDG tends to accumulate in macrophages due to their increased glucose utilization that results from activation [26, 27]. The ability to identify clinically relevant atheromas by FDG has been validated by several studies [28, 29]. A study by Ogawa et al. provided the histopathologic evidence that vascular FDG uptake is associated with macrophage infiltration and foam cell formation [26]. The study further showed that FDG uptake was higher when macrophages were actively differentiating into foam cells, rather than in the stage of completely differentiated foam cells.

A strong correlation between FDG-avid lesions and atherosclerotic plaques in different arteries has been observed, including in the aorta, coronary arteries, vertebral arteries, and carotid arteries [6, 29,30,31]. When Rudd et al. used autoradiography to compare the difference in the FDG uptake between symptomatic and asymptomatic carotid lesions, the authors found that the accumulation of FDG is 27% higher in symptomatic carotid lesions than in contralateral asymptomatic lesions, and no measurable uptake was detected in normal carotid arteries [32]. Studies have shown promising utility of FDG in evaluating the risk of ischemic events [33, 34]. FDG-PET/CT has also shown potential in the ability to assess the effects on medical therapy on the evolution of atheromas [35, 36].

Despite the significant correlation between FDG and atherogenesis as noted by multiple studies, discrepancies have been found between the accumulation of FDG and structural changes detected by the anatomic imaging modalities, especially CT [37, 38]. According to a study by Meirelles et al., disease progression in advanced stages may be accompanied by subdued inflammation, reducing the sensitivity of FDG in such cases [39]. As such, FDG uptake is not generally associated with macro-calcification visible on CT. Results reported by Ben-Haim et al. [40] and Dunphy et al. [37] show only 7% and 2% of cases with CT calcification had corresponding FDG uptake, respectively. In addition to the aforementioned nuances in FDG association with disease activity, the correlation between cardiovascular risk factors and FDG uptake is poorly understood. Risk factors including age, dyslipidemia, smoking history, hypertension, diabetes, etc., are known to play a vital role in the process of atherosclerosis [2]. To date, studies have only been able to successfully prove the clear association between age and FDG avidity [41, 42]. A major limitation of using FDG-PET/CT to assess atherosclerosis is the impracticality of determining disease in the coronary arteries [43]. Expected physiologic uptake of FDG by the myocardium is much higher than the subtle changes in uptake resulting from plaque formation in the coronary arteries. Thus, given the limitations of FDG in this domain, it is expected that NaF-PET/CT will play a major role in PET-based atherosclerosis imaging [44, 45].

18F-Sodium Fluoride (NaF)

Vascular Micro-calcification

Unlike the nonspecific nature of FDG uptake, NaF localizes specifically to areas of ongoing calcification [46]. Therefore, NaF-PET images are not affected by physiologic myocardial uptake and therefore can detect early signs of coronary plaque micro-calcification in atherosclerosis. Because micro-calcification is mainly composed of inorganic hydroxyapatite, NaF-PET/CT is able to identify thin-cap fibroatheromas, which represent rupture-prone areas (Fig. 2) [47]. NaF was originally introduced as a radiotracer in nuclear medicine by Blau et al. for skeletal imaging given its high uptake by bone [48]; it is therefore important to avoid osseous structures when attempting to quantify vascular uptake of NaF.

Fig. 2
figure 2

CT, fused NaF-PET/CT, and IVUS images from three patients with unstable angina. While lesions with spotty calcification (a–d) and a large necrotic core (e–h) were NaF avid, the fibrotic lesion (i–l) did not demonstrate focal uptake. Image reprinted without changes from Li et al. [47] under the Creative Commons Attribution 4.0 International License (CC BY). https://creativecommons.org/licenses/by/4.0/

Full size image

NaF-PET has also shown higher sensitivity and specificity than FDG-PET in detecting and characterizing atherosclerosis [49,50,51,52,53,54,55,56,57,58,59,60,61,62]. Furthermore, NaF uptake in atherosclerotic plaque correlates with cardiovascular risk factors but not with coronary calcium scoring [63, 64]. This emphasizes the role of NaF-PET in assessing early rather than late atherosclerotic disease burden. Strong correlations have been found between NaF activity and different cardiovascular risk factors including advancing age, dyslipidemia, diabetes mellitus, and hypertension [65,66,67,68]. Moreover, NaF uptake has also been found to be correlated with vital and laboratory values such as blood pressure and the triglycerides-to-high-density lipoprotein ratio [60, 61]. Associations between vascular NaF uptake and cardiovascular scoring systems including Framingham Risk Score (FRS) [69], atherosclerotic cardiovascular disease (ASCVD) risk scores, and CHADS-VASc score have been described as well [70]. However, there is a debate regarding the effect of gender on these correlations [65, 66]. According to McKenney-Drake et al., cardiovascular micro-calcification as shown by NaF-PET/CT has a potential role in risk stratification and the ability to detect arterial calcifications earlier than by CT [49].

In a prospective clinical study by Joshi et al. that compares both FDG and NaF radiotracers, the latter was shown to be superior in identifying culprit lesions from non-culprit lesions among patients with acute coronary syndrome [71]. In another study by Doris et al., it was shown that coronary NaF activity has the ability to provide insight into disease within the coronary circulation by identifying coronary segments with more rapid progression of coronary calcification (Fig. 3) [72]. Kwiecinski et al. found that increased coronary NaF activity is able to serve as a reliable prediction model of both fatal or non-fatal myocardial infarction, independently from age, sex, risk factors, segment involvement, coronary calcium scores, presence of coronary stents, coronary stenosis, REACH, SMART scores, the Duke coronary artery disease index, and recent myocardial infarction [73]. To date, stress testing and conventional coronary angiography can only identify obstructive lesions but do not have the ability to detect high-risk plaques which could potentially cause acute thrombosis. When compared with the high-risk plaque features on intravascular ultrasound (positive remodeling, micro-calcification, and necrosis of the lipid core), Lee et al. demonstrated a significant association with coronary plaques with high focal NaF avidity [74].

Fig. 3
figure 3

Baseline and repeat imaging of two coronary plaques in the same patient. Contrast-enhanced CT coronary angiography (A, C) and fused PET/CT (B) demonstrate NaF activity localized to calcification present in the left main stem. CT coronary angiography 1 year later (D, E) demonstrates increased CT calcification at this site. By contrast, CT coronary angiography (F, H) and PET/CT (G) showed a non-NaF avid calcified plaque in a proximal obtuse marginal branch, which was not found to have increased calcification after 1 year (I, J). Image reprinted without changes from Doris et al. [72] under the Creative Commons Attribution 4.0 International License (CC BY). https://creativecommons.org/licenses/by/4.0/

Full size image

The question of quantification of radiotracer activity in an atherosclerotic plaque on PET is difficult given the small size of atheromas and the low resolution of PET [75]. The maximum standardized uptake value (SUVmax) is the most commonly used parameter to measure tracer activity. The use of the target-to-blood pool ratio (TBR) in NaF imaging, a calculation derived by dividing the raw standard uptake value (SUV) to the venous blood pool SUV remains controversial [30]. Global assessment with NaF-PET/CT could potentially overcome the difficulty in quantifying atherosclerotic plaques by defining a region of interest that encompasses the entire affected vessel [43, 76].

Valvular Calcification

Calcification is a major mechanism causing aortic stenosis [77], the most frequent form of valvular heart disease in the Western world [78]. On CT, the degree of calcification of the aortic valve can be determined quantitatively and precisely, and it appears to correspond with the hemodynamic severity of the aortic stenosis [79, 80]. Hence, the ESC/EACTS Guidelines on Valvular Heart Disease suggest that CT calcium scoring should be used as a first-line test in cases with discordant grading [81]. However, NaF-PET/CT imaging has the advantage of being able to determine the degree by which active calcification is occurring [82]. In addition, histological markers of calcification activity coincide with NaF [83]. Thus, patients with more severe disease have the highest degree of tracer uptake [12], explaining why disease progression is faster in patients with severe AS. New calcification could be identified in the areas of elevated NaF activity reported on the baseline scan if patients undergo repeat CT calcium scoring [83]. As a result, baseline NaF uptake can predict disease progression and related adverse cardiovascular events. Hence, it might also be used to assess the effects of novel drugs on disease activity besides assessing the durability of transcatheter aortic valve replacements. Additionally, PET/MRI offers simultaneous imaging of calcification activity in the aortic valve using NaF in addition to detailed assessment of markers of myocardial decompensation. Hence, further testing for signs of calcification in bioprosthetic valves as an early marker of valve degradation using PET/MRI may be possible.

Cardiac Amyloidosis

The two types of cardiac amyloidosis are acquired monoclonal immunoglobulin light-chain (AL) and transthyretin-related (familial and wild-type/senile) amyloidosis (ATTR). CMR is routinely used in the assessment of cardiac amyloidosis. However, CMR is unable to differentiate between AL and ATTR amyloidosis. Because there is higher NaF activity in ATTR cardiac amyloidosis but not AL amyloidosis, NaF-PET/CT may be a viable non-invasive method to discriminate ATTR from AL [84]. This variation in tracer uptake shows that the kind of amyloid deposits in the myocardium influences calcium homeostasis locally. In patients with ATTR amyloidosis, NaF uptake could be effective for disease monitoring and amyloid deposition localization [85].

In addition to NaF, PET tracers which target amyloid fibrils, including 11C-Pittsburgh compound B (11C-PiB), 18F-florbetaben, 18F-florbetapir, and 18F-flutemetamol, have been proposed to image cardiac amyloidosis [86]. A study by Park et al. utilizing autoradiography suggested that 18F-florbetapir binds AL to a higher degree than to ATTR [87]. A similar pattern of differential uptake among cardiac amyloidosis subtypes has been described using 18F-florbetaben [88]. A meta-analysis by Kim et al. reported high sensitivity and specificity of amyloid PET in the diagnosis of cardiac amyloidosis, suggesting that amyloid PET imaging may play an additional role in identifying amyloid subtypes [89].

Other Radiotracers

Myocardial Perfusion Imaging

82Rb is a radiotracer that is absorbed from the blood by the myocardium via the Na+/K+-ATPase [90]. It has been widely utilized with PET and has become an important technique for determining myocardial blood flow and coronary reserve (Fig. 4) [91, 92]. As such, it is widely accepted as a standard for coronary artery disease imaging [93]. In a recent prospective study done by Gaudieri et al. on 517 hypertensive subjects of which 26% had resistant hypertension, 82Rb-PET was able to perform risk stratification by assessing coronary vascular function [94]. According to the authors, subjects with resistant hypertension and coronary vascular dysfunction have the highest risk of cardiovascular events. However, the practical usage of 82Rb-PET is limited by a short half-life and a long positron range. Other PET tracers proposed for perfusion imaging suffer from similar technical concerns which limit their widespread adoption. For example, 15O-water produces images with a low signal-to-noise ratio, and 13 N-ammonia still has a longer positron range compared to 18F-labeled tracers [95, 96].

Fig. 4
figure 4

99mTc-sestamibi SPECT (A) and 82Rb-PET/CT (B) studies performed in a 56-year-old woman. Ischemia involving the circumflex artery (arrow) was evident by 82Rb-PET after administration of dipyridamole, demonstrating the superior sensitivity of this technique. Image reprinted without changes from Chatal et al. [92] under the Creative Commons Attribution 4.0 International License (CC BY). https://creativecommons.org/licenses/by/4.0/

Full size image

18F-flupiridaz is a pyridazinone derivative that targets the mitochondrial complex I of the electron transport chain with high affinity. The method of localization for this radiotracer is deemed cardioselective because mitochondria make up roughly 20–30% of the cellular volume of the myocardium, resulting in higher myocardial uptake compared to the lungs and liver [97]. 18F-flurpiridaz has shown to be advantageous in detecting myocardial blood flow [98]. It has demonstrated a high level of uptake in the heart, and it has a high myocardial extraction fraction, permitting easier quantification using PET [99]. Myocardial perfusion imaging with 18F-flurpiridaz-PET was found to have superior diagnostic ability compared to 99mTc-labeled SPECT in women, obese individuals, and those undergoing pharmacological stress testing in a phase III clinical trial [100]. Therefore, abnormalities in myocardial perfusion may be assessed by 18F-flupiridaz, which has the potential to become a routine clinical test.

18F-Flubrobenguane

18F-flubrobenguane, also referred to as 18F-LMI1195, is a 18F-labeled norepinephrine (NE) transporter ligand demonstrating clinical promise for its ability to evaluate cardiac sympathetic function as a PET radiotracer [101]. 11C-metahydroxyephedrine (11C-HED) is a NE transporter ligand PET tracer similar to 18F-flubrobenguane but has limited use due to its short half-life. Uptake of 18F-flubrobenguane has been shown to closely mimic the physiologic NE uptake mechanism, thereby allowing for evaluation of cardiac neuronal integrity and function. Physiologic NE is stored in presynaptic vesicles and released upon neural impulse into the synaptic cleft via vesicular exocytosis. At the post-synaptic cleft, adrenoreceptors are activated. NE is then transported to the presynaptic cleft for reuptake and is stored before the next firing impulse. It has been demonstrated that blockage of the NE transporter via a selective NE transporter inhibitor, desipramine, reduces uptake of 18F-flubrobenguane considerably, suggesting that 18F-flubrobenguane closely mimics NE turnover [102]. Furthermore, early data have shown that 18F-flubrobenguane provides a well-tolerated non-invasive assessment of cardiac regional denervation and heterogeneity of innervation [103].

PET imaging with 18F-flubrobenguane allows for evaluation of damage to the cardiac nervous system, thereby identifying individuals at high risk for cardiac events. For example, heart failure has been associated with increased sympathetic activity within the heart. This increase in NE release downregulates the NE transporter responsible for its reuptake, thereby increasing synaptic concentrations of NE. Higher concentrations of NE desensitize beta-adrenoceptors, which would otherwise slow progression of cardiac disease [104]. Since regional cardiac sympathetic dysfunction has been associated with incidence of ventricular arrhythmias and sudden cardiac death (SCD), PET can be used to guide management and stage patients for implantable cardioverter-defibrillator placement [105]. In the Prediction of ARrhythmic Events with PET (PAREPET) trial, patients were imaged with echocardiography, 11C-HED PET, 13N-ammonia PET for perfusion, and FDG-PET for viability to assess predictors of SCD [106]. Results from the PAREPET trial showed that sympathetic denervation imaged by 11C-HED PET could predict cause-specific SCD [107]. A later analysis showed that regional rather than global parameters derived from 11C-HED PET were superior in determining SCD risk [108]. The PAREPET II trial aims to evaluate the role of 18F-flubrobenguane in characterizing left ventricular sympathetic denervation, with SCD serving as the primary endpoint [109]. The results of this ongoing trial may clarify the role of imaging sympathetic function by PET in determining which patients would benefit from an implantable cardioverter-defibrillator.

Conclusions

Early detection of atherogenesis with FDG and NaF represents an exciting development within cardiac PET. Furthermore, evaluation of inflammation and ATTR amyloid deposition affecting the myocardium can be detected with FDG and NaF, respectively. Cardiac sarcoidosis in particular is a domain that can benefit from the additional characterization only available by FDG-PET. A large number of additional tracers, both in constant routine use as well as newly introduced, are under investigation into their utility in assessing the consequences of ischemic cardiomyopathy. As future studies provide further validation for these tracers and clarify their optimal usages, it is expected that patient care and our scientific understanding of CVD will continue to benefit from the knowledge provided by cardiac PET imaging.