Atherosclerosis immunoimaging by positron emission tomography

Abstract

The immune system’s role in atherosclerosis has long been an important research topic and is increasingly investigated for therapeutic and diagnostic purposes. Therefore, noninvasive imaging of hematopoietic organs and immune cells will undoubtedly improve atherosclerosis phenotyping and serve as a monitoring method for immunotherapeutic treatments. Among the available imaging techniques, positron emission tomography (PET)’s unique features make it an ideal tool to quantitatively image the immune response in the context of atherosclerosis and afford reliable readouts to guide medical interventions in cardiovascular disease. Here, we summarize the state of the art in the field of atherosclerosis PET immunoimaging and provide an outlook on current and future applications.

Introduction

Despite recent advances in its management, cardiovascular disease (CVD) continues to be the leading cause of mortality worldwide. CVD’s main underlying cause, atherosclerosis, is a slowly progressing, chronic disease of the arterial wall. While originally the focus has been on lipids, mounting evidence accumulated over the past 25 years has unequivocally established inflammation as a key driver of atherogenesis. The disease process is initiated by endothelium disruption and the accumulation of oxidized low-density lipoprotein (oxLDL), which triggers immune cell recruitment to the vessel wall. In early lesions, monocytes differentiate into macrophages, which upon oxLDL uptake evolve into foam cells. Further accumulation of lipids leads to foam cell death and vessel wall thickening. These focal lesions, called plaques, can eventually rupture and the released material trigger acute cardiovascular events such as myocardial infarction or stroke. Therefore, unraveling the immune system’s role in atherosclerosis is expected to increasingly lead to viable immunotherapeutic interventions, while – similar to cancer immunotherapy – patient stratification and treatment monitoring can be facilitated by imaging approaches.

Indeed, two recent clinical trials highlight the complex role that inflammation plays in atherosclerotic CVD. The CANTOS trial concluded that anti-interleukin 1β therapy results in a reduced rate of recurrent cardiovascular events.¹ On the other hand, the CIRT study demonstrated that low dose methotrexate treatment did not result in lower levels of a number of inflammatory markers nor in fewer cardiovascular events compared to placebo.² These mixed results likely indicate the existence of specific inflammatory pathways relevant to atherosclerosis, but also prove that effectively targeting the involved pathways systemically is a plausible therapeutic avenue to treat CVD. Consequently, atherosclerosis immunology’s complexity and the resulting arterial wall remodeling mandate an integrative approach probing the hematopoietic system as well as plaque activity, morphology and composition.

Positron emission tomography (PET) is particularly suited for this purpose due to this imaging modality’s high tissue penetration and sensitivity for the detection of radiopharmaceuticals with specific biological targets. PET allows noninvasive quantification of molecular processes in a hot spot fashion using minimal amounts of injected radiotracer with no pharmacological effects. While PET necessitates anatomical reference and poses radiation concerns, its value for atherosclerosis imaging – in combination with computed tomography (CT) or magnetic resonance imaging (MRI) – is increasingly appreciated. For example, a recent study by Sanz and colleagues has shown PET/MRI’s potential to detect subclinical atherosclerosis.³ The advent of total-body PET is expected to allow simultaneous vessel wall and hematopoietic system imaging, at reduced scan times and, most importantly, at a relatively low radiotracer dose due to its substantially increased sensitivity.⁴ The latter is simultaneously facilitated by advances in probe design, consisting of molecular ligands – ranging from small molecules to complex proteins – and their labeling with radioisotopes, which can be achieved by covalent conjugation or through chelation.

In this review, we summarize the growing body of work focused on PET imaging of the immune system in atherosclerosis, highlighting its value as a tool in anti-atherosclerosis drug development, and to diagnose and guide therapeutic interventions.

Imaging the immune response in atherosclerosis by PET

Atherosclerosis PET imaging initially focused on the so-called vulnerable plaque, aiming to identify rupture-prone lesions likely to trigger acute events. Immunologically, an increasing number of cell types is being implicated in atheroprogression⁵ but monocytes/macrophages remain by far the preferred cellular target for immunoimaging of atherosclerosis given their prominent role in plaque development and rupture. While identifying high-risk plaques is still a key goal, the paradigm shift in atherosclerosis research towards a more integrative systems approach is also slowly changing how the disease is imaged. It has long been known that high numbers of circulating leukocytes are predictive of cardiovascular events.⁶ In mice, this monocytosis is also observed,^7,8 and results from increased hematopoietic activity.⁹ Moreover, CVD events like myocardial infarction aggravate atherosclerosis through heightened hematopoiesis and the ensuing monocytosis.¹⁰ These findings have been corroborated non-invasively by PET imaging. Increased hematopoiesis has been measured as augmented proliferative/metabolic activity in the spleen and bone marrow in different animal models⁹ as well as in humans¹¹ with atherosclerosis. Moreover, in another PET imaging-based study, psychosocial stress has been linked to CVD events through an association between resting amygdalar activity and medullary hematopoiesis.¹² Collectively, these findings support the role of the hematopoietic system in the low-grade inflammation observed in atherosclerosis, as well as the need to assess the disease systemically.

Many radiotracers that were initially developed for oncological PET have found application in atherosclerosis immunoimaging, including fluorine-18 (¹⁸F)-labeled 2-deoxy-2-fluoro-D-glucose (¹⁸F-FDG) and 3’-deoxy-3’-fluorothymidine (¹⁸F-FLT) or gallium-68 (⁶⁸Ga)-labeled DOTATATE. While not directly targeting immune cells or the immune function, it is worth mentioning ¹⁸F-sodium fluoride (¹⁸F-NaF) as an atherosclerosis PET radiotracer. ¹⁸F-NaF accumulates in plaques through its binding to calcium phosphate (mostly hydroxyapatite) in microcalcifications,¹³ which are produced by smooth muscle cell-derived osteoblast-like cells in response to inflammatory cytokines. Given its fast clearance and very low myocardial uptake, ¹⁸F-NaF can efficiently visualize high-risk and ruptured atherosclerotic plaques in the coronary arteries.¹⁴ However, ¹⁸F-NaF atherosclerosis PET imaging is not devoid of limitations, as its high accumulation in the bone may interfere with quantification of aortic microcalcifications.

Among the available PET radioisotopes, ¹⁸F is the most widely used for labeling of small molecules, while zirconium-89 (⁸⁹Zr) and the generator-produced gallium-68 (⁶⁸Ga) are gaining traction for labeling peptides, proteins and nanoparticles through chelation. Table 1 shows the most relevant radionuclides used for atherosclerosis immunoimaging and some key physicochemical properties. Figure 1 schematically summarizes the main atherosclerosis PET immunoimaging approaches that will be discussed in the following sections, as well as a selection of probes and labeling strategies thus far implemented.

Table 1.

Radioisotopes most frequently used for atherosclerosis immunoimaging by positron emission tomography. t_1/2 = physical half-life; R_mean = mean positron range; EC = electron capture; [C] = cyclotron; [G] = generator.

Isotope	t1/2 [h]	Production	Decay	Rmean [mm]	Probe	References
11C	0.34	14N(p,α)11C [C]	>99% β+	1.27	11C-choline 11C-acetate 11C-PK11195	27 29 40,41
18F	1.83	18O(p,n)18F [C]	96.9% β+ 3.1% EC	0.66	18F-FDG 18F-FDM 18F-FLT 18F-choline 18F-macroflor	3,11,12,18–20 25 9,60 28 58
68Ga	1.13	68Ge/68Ga [G]	89.1% β+ 10.9% EC	3.56	68Ga-DOTATATE 68Ga-DOTATOC 68Ga-nanobodies	30,31 33,34 46,50
					68Ga-pentixafor	36–39
64Cu	12.7	64Ni(p,n)64Cu [C]	17.9% β+ 38.4% β− 43.1% EC	0.56	64Cu-DOTATATE 64Cu-nanobodies	32,33 50
89Zr	78.4	89Y(p,n)89Zr [C]	22.3% β+ 76.6% EC	1.27	89Zr-HDL 89Zr-hyaluronan 89Zr-LA25	57 59 52

Figure 1. Atherosclerosis immunoimaging by positron emission tomography.

A) PET imaging allows both focal and systems imaging of the atherosclerotic process by quantitative assessment of immune activation in the plaque and other tissues. Sp: spleen; BM: bone marrow. B) Key strategies for immunoimaging of atherosclerosis implemented to date. C) Selected PET radiotracers and radiolabeling strategies used for atherosclerosis PET immunoimaging. Chelators are shown in blue. NOTA: 1,4,7-Triazacyclononane-1,4,7-triacetic acid. DFO: deferoxamine.

Immunometabolic PET imaging

Activated immune cells show a distinct metabolic profile in contrast with quiescent states.¹⁵ Similar to tumor cells, immune cells undergo metabolic reprogramming towards aerobic glycolysis upon activation by external stimuli.¹⁶ Importantly, activation of glycolytic pathways occurs both in classically and alternatively activated macrophages.¹⁷ This metabolic shift reflects the energetic and biosynthetic demands of the activated cell. Thus, most metabolic radiotracers developed for cancer PET imaging have been used for atherosclerosis imaging. Among these, the glucose analog ¹⁸F-FDG^18,19 features prominently. ¹⁸F-FDG is the most widely used radiotracer in the clinic accounting for 90% of oncological PET scans worldwide, and is also the preferred radiotracer for atherosclerosis imaging.²⁰ In the plaque, ¹⁸F-FDG uptake correlates with macrophage burden and is used as an inflammation radiotracer,¹⁹ although enhanced ¹⁸F-FDG uptake may also be due to hypoxia.²¹ Despite its widespread use, ¹⁸F-FDG suffers from lack of specificity²² and high brain and myocardial uptake, which precludes imaging of the vasculature in these tissues. However, specific patient preparation protocols^23,24 based on a ketogenic diet and fasting prior to the scan can efficiently suppress myocardial ¹⁸F-FDG uptake in most subjects and facilitate coronary artery wall imaging. As an alternative metabolic tracer to ¹⁸F-FDG, the use of ¹⁸F-labeled 2-deoxy-2-fluoro-D-mannose (¹⁸F-FDM) was explored in rabbits. In this study, both epimeric tracers were compared head to head showing very similar performance including myocardial uptake.²⁵

At the same time, activated immune cells enter a proliferative state¹⁶ that requires biosynthetic components and building blocks including nucleotides. A study in mice, rabbits and humans showed that the uptake of ¹⁸F-labeled thymidine (¹⁸F-FLT)⁹ was significantly higher in subjects with atherosclerosis compared to controls due to plaque macrophage proliferation²⁶. Moreover, ¹⁸F-FLT-PET in mice also revealed increased uptake in bone marrow and spleen, indicative of systemic immune activation in atherosclerosis. Similarly, proliferating macrophages need to synthesize membrane components like cholesterol, fatty acids or phospholipids.¹⁶ Exploiting this metabolic feature, ¹¹C- and ¹⁸F-labeled choline, a building block for membrane phospholipids, have been used for imaging inflammation in mice²⁷ and patients,²⁸ respectively, showing increased uptake in plaques that was mainly due to macrophages. Along the same line, the use of ¹¹C-acetate, which feeds fatty acid synthesis pathways, has been explored in patients.²⁹ Interestingly, most calcified regions did not show enhanced ¹¹C-acetate uptake, suggesting that macrophage activation and calcification are separate events in time.

Immune cell radioligand/receptor-based PET imaging

The abovementioned ¹⁸F-FDG’s shortcomings for atherosclerosis imaging, shared in part by all metabolic tracers, have spurred a search for higher specificity tracers and the ability to image the coronaries. A prevalent strategy focuses on targeting membrane receptors that are specifically upregulated in activated immune cells. One such example is the somatostatin receptor subtype 2 (SSR2), which is expressed on activated macrophages. A derivative of the somatostatin analog octreotide modified with the chelator DOTA, namely DOTA-octreotate or DOTATATE, has been extensively studied in the context of atherosclerosis imaging. Thus far, gallium-68 (⁶⁸Ga)-labeled DOTATATE has been evaluated in mice, in comparison with other SSR2 tracers,³⁰ and in patients, showing excellent macrophage specificity and superior coronary lesion discriminating features compared to ¹⁸F-FDG.³¹ The use of copper-64 (⁶⁴Cu) instead of ⁶⁸Ga allows imaging at later time points with improved spatial resolution – due to ⁶⁴Cu shorter positron range – albeit potentially increasing radiation exposure. Using PET /MRI, ⁶⁴Cu-DOTATATE was found to accumulate in carotid plaques, and this uptake correlated with alternatively activated macrophage burden.³² In a retrospective study in oncological patients, ⁶⁴Cu-DOTATATE uptake correlated with CVD risk factors.³³ Similarly, DOTATOC, another DOTA-modified octreotide peptide, has also been explored as a macrophage imaging agent in atherosclerosis in its ⁶⁸Ga-labeled version with opposing results. In a head-to-head comparison with ⁶⁴Cu-DOTATATE, ⁶⁸Ga-DOTATOC uptake was significantly lower in vascular regions and did not correlate with CVD risk factors.³³ A more recent study, on the contrary, reported a significant association between ⁶⁸Ga-DOTATOC thoracic aortic uptake and cardiovascular risk factors.³⁴

Similarly, the ⁶⁸Ga-labeled CXC-motif chemokine receptor 4 (CXCR4) ligand pentixafor has been proposed as an inflammation radiotracer in atherosclerosis. While CXCR4 is expressed on several immune cell subsets, in the context of atherosclerosis it is found mainly on monocytes and macrophages.^{35 68}Ga-pentixafor was first evaluated in rabbits³⁶ with promising results, demonstrating uptake in macrophage-rich areas. More recently, two retrospective studies in patients established a significant correlation between ⁶⁸Ga-pentixafor artery wall uptake and a number of CVD risk factors.^37,38 Also in humans, ⁶⁸Ga-pentixafor has been used to detect coronary artery wall inflammation by motion-corrected PET/CT.³⁹ Collectively, these results indicate the clinical value of ⁶⁸Ga-pentixafor as a macrophage-specific radiotracer to quantify arterial inflammation.

Translocator protein (TSPO) expression is upregulated in activated macrophages, representing another potential target to identify inflamed atherosclerotic plaques. To date, several TSPO ligands have been explored as atherosclerosis PET imaging agents. ¹¹C-PK11195 PET imaging was first described in mice, showing increased accumulation in atherosclerotic plaques but with high non-specific uptake.⁴⁰ In humans, ¹¹C-PK11195 uptake was higher in carotid plaques associated with ipsilateral symptoms (stroke or transient ischemia) compared to those without.⁴¹ In addition, the TSPO-targeted radioligands ¹⁸F-FEMPA⁴², ¹⁸F-GE-180⁴³ and ¹⁸F-PBR111⁴⁴ have been tested in mice. While ¹⁸F-FEMPA and ¹⁸F-GE-180 showed specific uptake in macrophage-rich areas, their overall aortic uptake in atherosclerotic samples was not different from healthy tissue. On the other hand, ¹⁸F-PBR111 uptake was significantly higher in atherosclerotic aortas, where TSPO localized to CD11b⁺ monocyte-derived macrophages.⁴⁴

Immuno-PET for atherosclerosis imaging

Immuno-PET, i.e. the use of antibody-based radiotracers for PET imaging, is an established approach in oncology.⁴⁵ Due to their high affinity and target specificity, antibodies are ideal for imaging purposes. However, the use of full-size antibodies is undesirable in the context of atherosclerosis due to blood pool signal and spillover to the arterial wall as a result of their very long circulation time. For this reason, smaller antibody fragment-based radiotracers for PET and single-photon emission computed tomography (SPECT) have been developed and tested in mice, such as nanobodies targeting the macrophage mannose receptor (MMR),⁴⁶ lectin-like oxidized LDL receptor-1 (LOX-1),⁴⁷ and vascular cell adhesion molecule (VCAM)-1,^48,49 an early marker of endothelium activation and immune cell recruitment. Nanobodies’ fast renal clearance facilitates imaging of the vessel wall due to reduced blood pool background signal, and affords highly specific imaging of their molecular targets.

More recently, we have used translational PET/MR imaging to study these three nanobodies in a head-to-head comparison using a rabbit atherosclerosis model.⁵⁰ Initially labeled with ⁶⁴Cu, the nanobodies were also compared with ¹⁸F-FDG. Further, the MMR nanobody was labeled with ⁶⁸Ga and included in an atherosclerosis progression study that also comprised ¹⁸F-FDG- and ¹⁸F-NaF-PET imaging, in addition to MR-based assessment of vessel wall thickness and permeability. Interestingly, ⁶⁸Ga-MMR nanobody uptake correlated with vessel wall area, demonstrating macrophage accumulation during atherosclerosis progression.⁵⁰

The atherosclerotic process can also be monitored indirectly through evaluation of oxidative stress. It is becoming increasingly clear that oxidized phospholipids and other lipid peroxidation breakdown products are causative of CVD.⁵¹ In a recent study, we reported the development of an antigen-binding fragment (Fab)-based ⁸⁹Zr-labeled radiotracer, derived from a natural antibody, that targets the malondialdehyde-acetaldehyde adduct oxidation-specific epitope and enables non-invasive visualization of atherosclerotic lesions.⁵² The Fab radiotracer uptake in the plaque correlated with ¹⁸F-FDG and with macrophage burden.

Immuno-PET imaging of atherosclerosis using small antibody fragments is therefore a compelling tool to non-invasively phenotype immune cells in vivo.^53,54 This information, obtained focally in atherosclerotic lesions or systemically in hematopoietic tissues, can potentially be used to stratify patients for targeted therapies and/or to monitor response to treatment.⁵⁴ Due to their fast clearance, one key strength of this approach is that it allows radiolabeling with short-lived isotopes and same-day imaging, which makes it more convenient for patients and reduces overall radiation exposure.⁵⁵ On the other hand, antibody fragment development and production is costly, and their fast clearance has the downside of producing lower signals compared to full-size antibodies. However, this issue can be mitigated through chemical modification in order to increase their size and thus prolong their circulation times.⁵⁴

Phagocyte-mediated nanoparticle uptake PET imaging

Lastly, radiolabeled nanoparticles can be used for atherosclerosis immunoimaging by PET. High-density lipoprotein (HDL), a natural nanoparticle, has an intrinsic ability to interact with macrophages through membrane receptors in reverse cholesterol transport, a feature that has been exploited for therapeutic and imaging purposes in the past.⁵⁶ More recently, we developed ⁸⁹Zr radiolabeling strategies that enabled non-invasive visualization of HDL in mouse, rabbit and swine atherosclerosis models, showing specific uptake by macrophages.⁵⁷ Importantly, these radiolabeling strategies have been subsequently used for characterization of HDL-based nanobiologics.⁵⁶ However, HDL’s long circulation time prohibits its use as an atherosclerosis imaging agent. Conversely, an ¹⁸F-labeled macrophage-targeted polyglucose nanoparticle, named ¹⁸F-Macroflor, has been evaluated in mice and rabbits.^{58 18}F-Macroflor’s small size (~5 nm) allows fast renal clearance and low background signal for vessel wall imaging. This nanotracer is taken up by macrophages, and was evaluated not only for atherosclerosis imaging, but also for myocardial infarction in mice. Similarly, we radiolabeled a hyaluronan nanoparticle with ⁸⁹Zr to image macrophage inflammation in atherosclerotic mice and rabbits.⁵⁹

Perspective/conclusion

Atherosclerosis immunoimaging by PET is a valuable approach to quantitatively assess disease burden, focally and systemically. Preclinically, it can be used to provide reliable quantitative efficacy data in early drug development. Thus, we advocate the integration of specific PET imaging readouts into multiparametric and translational PET/MR imaging protocols for evaluation of novel anti-atherosclerosis therapies in large animal models. In a recent study, we implemented this approach to escalate a nanoimmunotherapy from murine to rabbit and swine atherosclerosis models, which allowed us to assess treatment response from different angles using restricted group sizes (Figure 2).⁶⁰ In clinical trials, immunoimaging by PET can be used to interrogate specific disease aspects and obtain surrogate imaging end points to assess treatment response without the need for large cohort numbers or long follow-up periods.⁶¹ While genetic background is an important disease risk determinant, phenotypical traits are more strongly correlated with disease progression and outcome. Ultimately, clinical PET immunoimaging of atherosclerosis may be used to noninvasively quantify specific disease markers, which combined with gene screening and/or liquid biopsy biomarker data may be used to tailor personalized interventions, or to guide patient selection for immunotherapy and its subsequent monitoring. In summary, we anticipate that atherosclerosis PET immunoimaging will mature into a powerful platform that can be integrated in early drug development and in CVD precision medicine to inform personalized treatment options.

Figure 2. Integration of PET atherosclerosis immunoimaging in translational drug development.

Imaging immune cell activity in the vessel wall can be integrated into anti-atherosclerosis (nano)therapy development to provide valuable information on treatment response. A) In a recent study using a swine model, we measured the response to a 2-week simvastatin high-density lipoprotein (S-HDL) nanotherapeutic regimen using a longitudinal PET/MR imaging protocol. B) This protocol combined ¹⁸F-FDG and ¹⁸F-FLT PET immunoimaging to measure vessel wall inflammation and macrophage proliferation (top), respectively, with MRI-derived vessel wall permeability and plaque burden measurements (bottom). This approach allowed us to obtain meaningful results on the therapy’s effect on overall disease burden using restricted group sizes. Adapted from Binderup et al⁶⁰ with permission. Copyright © 2019, The American Association for the Advancement of Science.

Highlights.

• The immune system is critically involved in atherosclerosis progression

• PET is a quantitative imaging technique that allows noninvasive assessment of the immune system in atherosclerosis

• Atherosclerosis PET immunoimaging has been achieved using metabolic tracers, radioligands for specific receptors, nanoparticles and small antibody fragments targeting immune cell markers

• PET immunoimaging can be used to phenotype atherosclerotic plaques, assess hematopoietic tissue activation or guide CVD treatment

Nonstandard abbreviations and acronyms

PET

positron emission tomography

CVD

cardiovascular disease

CT

computed tomography

MRI

magnetic resonance imaging

FDG

fluorodeoxyglucose

DOTA

1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid

HDL

high-density lipoprotein

oxLDL

oxidized low-density lipoprotein