Previous molecular imaging studies have exemplified that persistent pro-inflammatory response after AMI can contribute to adverse LV remodelling and heart failure [4]. Our study demonstrated that the 68Ga-Pentixafor signal in both the infarcted and remote regions elevated significantly at 5 days after AMI. In addition, an association of the remote signals and LV remodeling was initially observed in our experiments. Ex-vivo tissue work-up confirmed a similar evolutionary trend of 68Ga-Pentixafor signals and pro-inflammatory M1-macrophages in the infarcted area.
The combination of CXCR4 and its ligands, alpha-chemokine CXCL12 and migration inhibitory factor (MIF) play a pivotal role in cell migration (“homing”), recruitment, adhesion, growth and proliferation, involving in broad-spectrum diseases like hematopoietic and solid tumors, infected and non-infected inflammations, and et al [14]. CXCR4 is thus a promising target in the theranostics [9, 15]. More currently, preclinical and clinical studies proved the corresponding novel PET tracer 68Ga-Pentixafor can sensitively and non-invasively visualize the expression of CXCR4 in lymphoma [16], multiple myeloma [17], and leukemia [18]. Limited studies implicated atherosclerosis [10] and myocardial infarction [11]. Meanwhile, the corresponding radio-targeted therapeutic probes like LY2510924, 177Lu- and 90Y-pentixather are in active exploration [19–20].
The orchestrated recruitment transition from proinflammatory to reparative cell may mitigate AMI injury, promote repair, and improve prognosis [2, 21], in vivo non-invasive monitoring of these processes is important for selecting the appropriate timing and intervention protocol. The CANTOS trial and a large body of clinical trials followed in patients with AMI have proved this [21–22]. Currently, the application of 68Ga-Pentixafor in AMI is in the most initial stage. To the best of our knowledge, only single-digit relevant original articles have been published currently. Studies remain pending regarding the dynamic evolution of CXCR4-inflammatory response, clinical significance of the radioactive signal, the interaction between organs, as well as the therapeutic guidance potentials.
Regarding the prognostic value of 68Ga-Pentixafor signal in the myocardial infarction area, there are inconsistencies in the animal and clinical studies currently. For example, Reiter et al. revealed in their study concluding 13 patients 2–13 days post-AMI that the higher the 68Ga-Pentixafor signal, the smaller the scar volumes at follow-up [23]. Yet, Hess et al. demonstrated that persistently high 68Ga-Pentixafor signals in the infarcted zone within 3 days predict a higher incidence of acute LV rupture and chronic contractile dysfunction [9]. Preliminary intervention trials found that an opportune inflammation blocking (early after infarction) produce a positive outcome [9], whereas blocking at an inappropriate timing (depletion of macrophages prior to AMI) will aggravate a poor prognosis [24]. These have raised the demand for longitudinally in vivo inflammatory characterization, the success of inflammation-targeted therapy largely depends on the appropriate target cells and timing.
Integrated from previous results, the 68Ga-Pentixafor signals of the infarcted myocardium surged immediately after surgery, peaked approximately 3 days, and decreased visibly after 7 days [25–26], the persistency of 68Ga-Pentixafor signals after AMI is shorter than 18F-FDG, the latter can persist to at least 14 days after AMI [8]. The evolution of 68Ga-Pentixafor signal observed in our study was comparable to previous results, with the following updates. Firstly, previous studies used the remote region as a background for uptake correction of interindividual heterogeneity, however, this is probably not the optimal choice. According to previous studies, severe inflammatory response may affect the remote area [3] and even the kidney [25], whereas 68Ga-Pentixafor uptake in the liver is relatively stable [27]. Therefore, in the current study, the liver was used as the site for background correction, which may objectively reflect the evolution of inflammatory signals in the non-infarcted area. Secondly, the 68Ga-Pentixafor signals almost recovered to baseline on day 7 post-infarction in previous studies. Yet, in our study, although the 68Ga-Pentixafor signal in the infarct area was reduced on day 7 versus day 5 post-AMI, it was still significantly higher than that in the Sham group, which was consistent with the ex-vivo autoradiography results on day 7 by Hess et al. They concluded that the higher sensitivity and lack of background signal with autoradiography result in the positive results [9]. Besides, the correction using signals in the remote area might be another speculation. Our results may further justify the correct choice of using the liver as a correction background [25–27]. Third, previous studies have mostly set time points of 3 days, 7 days, and 6 weeks post-AMI for assessment, we set 2 extra time points at 5 days and 14 days after AMI, filling the gap of 68Ga-Pentixafor evolution in previous studies. Regarding the clinical translation in future, the extended time window may implicate more feasibilities in patients monitoring after AMI. Furthermore, the back-to-baseline signal at 14 days after AMI can be considered as the end of the early inflammatory signal after AMI in rodents, which has not been clarified in previous 68Ga-Pentixafor studies. Fourth, Hess et al. found the early infarct CXCR4 signals at 1 or 3 days after AMI correlated well with subsequent LVEF, and the correlation at 3 days started to get weakened [9]. It can be speculated that the correlation will further weaken with time. The temporal bias of monitoring (relatively later) might be an explanation of the absent correlation between 68Ga-Pentixafor signals post-operation and LVEF in the present study.
Regarding the prognostic value of remote area, previous 18F-FDG inflammatory imaging studies indicated controversial 18F-FDG signals elevated in the distal myocardium [3, 7–8]. Wollenweber et al. detected increased 18F-FDG uptake in the non-infarcted area, spleen and bone marrow of AMI patients [7]. However, a recent study by Xi et al. in infarcted pigs did not find a meaningful increase in 18F-FDG uptake in the non-infarcted region [8]. In a mice and clinical study by Werner et al, 68Ga-Pentixafor signals in the infarct and remote myocardium were both correlated with signals in the kidney, moreover, the remote signal in AMI patients within 4 days of reperfusion was the only independent predictor of renal events [25]. However, none of aforementioned studies focus on the dynamic evolution and predictive value of 68Ga-Pentixafor signal in the remote region. Our study preliminarily displayed the longitudinal 68Ga-Pentixafor evolution in the remote area after AMI, and observed the capacity of 68Ga-Pentixafor in detecting the unobtrusively increased inflammation in the remote myocardium early after AMI, as well as its potential value in predicting cardiac remodelling, which lays the foundation for further prognostic studies in the future.
As for the predictive mechanisms of the signals in the remote region, firstly, spreading of the severe inflammatory response from the infarcted area to the non-infarcted area was concerned an established mechanism [3]. Secondly, abnormal first-pass perfusion [28] or reduced microcirculatory reserve (ie. coronary flow reserve) [29] were also identified in the non-infarcted area following revascularization in AMI patients. Our previous clinical observational research also identified the presence of coronary microvascular disease (CMVD) in the non-obstructed coronary territory in patients with obstructive coronary artery disease [30]. Furthermore, amplification of extracellular matrix, upregulated matrix metalloproteinases, systolic dysfunction, and thinning of the ventricular wall were also observed in the remote myocardium after AMI, all of which may be associated with poor prognosis.
Regarding the type of inflammatory cells targeted by 68Ga-Pentixafor, an in-vitro study confirmed a broad 68Ga-Pentixafor accumulation in monocytes/macrophages and lymphocytes, and for the former, comparable uptake was observed in M1- and M2-polarized macrophages [31]. Our in-vivo studies clarified a strong rise-fall parallel between 68Ga-Pentixafor accumulation and infiltrated M1-macrophages in infarcted territory, despite there was also a correlation between M2-macrophages and 68Ga-Pentixafor accumulation. Thus, here it can be speculated that pro-inflammatory M1-macrophages attributed more to the subsequent cardiac dysfunction or adverse remodelling. Regarding the observed profitably predictive value of 68Ga-Pentixafor in AMI patients by Reiter et al. [23], the relatively later monitoring time-frame possibly reflected mainly reparative inflammatory cells, exactly as the significantly increased M2-macrophages percent around day 7–14 after AMI we observed. The controversial correspondence between macrophage phenotype and 68Ga-Pentixafor in-vivo and ex-vivo may attribute to multi-factors including species characteristics, diverse assessing time point, and heterogeneity among individuals, which warrant further studies.
Given broad spectrum of leukocytes identified by 68Ga-Pentixafor, the research for more specific radiotracers is ongoing. So far, large amount of neuro-inflammatory research reflected 18F-GE-180 that targets translocator protein (TSPO) as a novel better PET radiotracer than 11C-R-PK11195, which markedly identify M1-macrophages [32]. Over recent years, novel promising tracer such as 68Ga-mCXCL12 [33], 68Ga-DOTATATE [34], and 68Ga-DOTA-ECL1i allosterically binding to CCR2 performed a potentially valid assessment in cardiovascular inflammation [35]. More work is needed to validate the value of these novel radiotracers in guiding theranostics of cardiovascular disease.
Limitations
Certain noticeable limitations of our study need to be acknowledged. The first major limitation concerns the small sample size. Given the limited animal number in the current study, conclusions should be drawn prudently. Further verifications need to be confirmed in large samples. Second, immunohistology was not conducted to confirm additional leukocyte subpopulations beyond M1/M2 macrophages and the evolution of the macrophages in the remote myocardium, and our work has not been performed to explain the prognostic mechanism, which will be replenished in future work. Third, we did not observe long-term change of 68Ga-Pentixafor signals. For the infarcted area, Hess et al. have shown that early 68Ga-Pentixafor signal can predict chronic contractile dysfunction at 6 weeks [9]. For the remote myocardium, due to non-significant difference in signal between the AMI and Sham groups at day 7 and day 14, further assessment was not necessarily conducted. Forth, we used gated 18F-FDG metabolic imaging than the gold criteria CMRI to evaluate the cardiac function, which may trigger questioning. However, prior comparative study has proved an excellent agreement of cardiac function between gated 18F-FDG and CMRI [28], additionally, we eliminated the outlier values in the parameters of cardiac function, which may compensate for the deficiency.