Clinical Molecular Imaging of Chemokine Receptor CXCR4 Expression in Atherosclerotic Plaque Using ⁶⁸Ga-Pentixafor PET: Correlation with Cardiovascular Risk Factors and Calcified Plaque Burden

DISCUSSION

To our knowledge, this was the first study of the prevalence, pattern, and clinical correlates of large–arterial wall accumulation of ⁶⁸Ga-pentixafor in patients, and our results establish ⁶⁸Ga-pentixafor PET/CT for imaging of CXCR4 expression in the atherosclerotic vessel wall. This novel methodology may refine the characterization of atherosclerotic lesions and may serve as a platform for future clinical studies targeting CXCR4-positive cells in atherosclerosis.

The CXCR4/CXCL12 axis (Table 4) plays a pivotal role in atherosclerosis and arterial injury (1,20). In advanced plaque, the density of macrophages indicates the risk of an atherothrombotic event because macrophages release proteolytic enzymes degrading the fibrous cap and promote intralesional inflammation by secreting proinflammatory cytokines (8). Therefore, many commonly used tracers, including ¹⁸F-FDG and ⁶⁸Ga-DOTATATE, are applied to visualize different aspects of macrophage density, the latter with limited success (8,21). Previous immunohistochemical analyses of CXCR4 expression in human atherosclerotic plaque revealed a progressive accumulation of CXCR4-positive cells during plaque evolution, and CXCR4-positive leukocytes have been detected in culprit lesions in acute coronary syndromes (22). Taken together, CXCR4 expression by leukocytes is associated with vulnerability of plaque. Indeed, vascular ⁶⁸Ga-pentixafor uptake showed a marked independent association with established risk factors such as at-risk age (r = 0.50, P = 0.0003), arterial hypertension (r = 0.52, P = 0.0001), and history of smoking (r = 0.36, P = 0.01). More important, both the number of CXCR4-positive lesions (r = 0.70, P < 0.0001) and the TBR (r = 0.36, P = 0.009) increased as the number of risk factors increased, suggesting a role for identifying risk-bearing atherosclerotic plaques.

Hyafil et al. have recently reported that CXCR4 expression is associated with CD68-positive macrophage-rich areas on immunohistochemistry of human carotid plaques (15). Additionally in that work, atherosclerotic plaques were induced by endothelial abrasion of large arteries in rabbits fed an atherogenic diet, leading to macrophage-rich plaques that were subsequently found to be CXCR4-positive (15). However, results from this nonphysiologic model of atherosclerotic plaque requiring both a considerable arterial injury and a special atherogenic diet may not be easily translated into the human clinical setting, in which atherogenic factors are much less pronounced, inflammation is not artificially induced and is less macrophage-dominated, and a variety of physiologic mechanisms are involved in the evolution of plaque. Indeed, other cell types involved in plaque evolution, including smooth muscle cell progenitors and endothelial progenitor cells, have also been found to express CXCR4 (1). Importantly, CXCR4 is also expressed by mature endothelial and smooth muscle cells in the arterial wall (1), indicating that other cell types may relevantly contribute to the in vivo CXCR4 signal. Arterial injury has been demonstrated to induce local smooth muscle cell CXCR4 expression, contributing to intimal hyperplasia (23). This indicates that a portion of ⁶⁸Ga-pentixafor uptake could reflect mechanisms other than atherosclerosis, such as arterial injury in areas of turbulent flow. Finally, it has recently been shown that both CXCR4-positive endothelial cells and smooth muscle cells are atheroprotective in an apolipoprotein E–deficient mouse model, limiting atherosclerosis by maintaining arterial integrity and preserving endothelial barrier function (24). This finding underlines that CXCR4-positive cells have many different roles both in the atherosclerotic and in the nonatherosclerotic arterial wall (1). Although our results indicate a potential role of ⁶⁸Ga-pentixafor PET/CT as a tool not only for in vivo visualization of CXCR4-positive atherosclerotic lesions in the large arteries but also for identification of inflamed, rupture-prone plaques, the relative contribution of different pathophysiologic mechanisms and different cell types to the observed CXCR4 signal remains to be elucidated.

Of note, the ¹⁸F-FDG signal in atherosclerotic lesions may be attenuated by use of statins. Likely because of the high cardiovascular risk profile, the number of arterial uptake sites and CP burden were higher in patients receiving statin treatment. However, the signal intensity (TBR) did not differ significantly between patients who were receiving ongoing statin treatment and those who were not (P = 0.23). These results are supported by recent data that found incubation with statins to have no significant impact on CXCR4 expression by macrophages (15), although other studies suggested some effect in certain monocyte subsets (17).

When the TBR of ⁶⁸Ga-pentixafor was analyzed, a good contrast between atherosclerotic lesions and the blood pool was found. The mean TBR of 2.0 ± 0.5 for CXCR4-positive lesions is comparable to the TBR of ¹⁸F-sodium fluoride and even higher than the TBR of ¹⁸F-FDG, facilitating the identification of atherosclerotic lesions (10,16,25,26). The distribution of ⁶⁸Ga-pentixafor throughout the major arteries was consistent with established atherosclerotic topography (25–27).

Although ⁶⁸Ga-pentixafor uptake was observed in all patients, CPs were less common (17.6% PET-positive/CT-negative patients). Interestingly, these patients were considerably younger than PET-positive/CT-positive patients (34.6 ± 8.8 vs 64.9 ± 11.8 y), suggesting that the presence of CXCR4-positive cells begins at an early age and precedes arterial calcification. Consistent with that observation, only 7.3% of all arterial calcification sites showed ⁶⁸Ga-pentixafor accumulation. CP is usually considered stable, late-stage atherosclerotic disease, underlining the possibility that CXCR4 imaging may identify earlier stages of plaque development and nonstable vulnerable plaque. Indeed, CXCR4-positive cells are involved in the initiation, progression, and rupture of atherosclerotic plaque (1,7). ¹⁸F-FDG uptake representing a measure of macrophage density in plaque is likewise rarely seen in CP (10,26). Furthermore, ⁶⁸Ga-pentixafor uptake is clearly linked to the clinical cardiovascular risk profile, supporting a role for identifying risk-bearing plaque. The rate of colocalization between vascular ⁶⁸Ga-pentixafor accumulation and calcification was 33.5%, whereas for ¹⁸F-FDG, coincident calcification was reported less frequently (10,26). These results indicate that CXCR4-targeted imaging may identify additional CXCR4-positive cells in the arterial wall besides intralesional macrophages, which represent the vascular target of ¹⁸F-FDG PET. Indeed, CXCR4 is also expressed by a variety of progenitor and other cells in the arterial wall, including endothelial and vascular smooth muscle cells (1).

Some limitations of the present study should be acknowledged. First, the precise cell type contributing to the in vivo ⁶⁸Ga-pentixafor signal cannot be identified. However, previous ex vivo data evidenced upregulated CXCR4 expression by macrophages in atherosclerotic lesions (15). Second, the results of this study might not be perfectly generalizable to other patient populations. However, factors that might affect physiologic processes in the vessel wall were carefully excluded. Third, scans were performed using a routine imaging protocol for ⁶⁸Ga-pentixafor imaging optimized for other indications. Dynamic or delayed data acquisition may provide additional information for the assessment of atherosclerotic plaques (28). ⁶⁸Ga-pentixafor exhibits a relatively high blood-pool signal. However, at least some part of that signal is caused by specific binding of ⁶⁸Ga-pentixafor on CXCR4-positive cells, questioning the potential impact of delayed imaging, which will rather support renal elimination of unbound tracer. In addition, the positron range is much higher for ⁶⁸Ga than for ¹⁸F, potentially affecting the localization of the vessel wall signal and contributing to partial-volume effects. Given the facts that pentixafor is a specific marker of CXCR4-positive cell infiltration, that there are numerous preclinical studies linking plaque progression and CXCR4, and that established cardiovascular risk factors were strongly associated with tracer uptake in this study, we hypothesize that CXCR4-targeted PET imaging bears promise for stratifying the risk of impending atherothrombotic events.