Cellular imaging: a bright future for 19F-CMR

In the study entitled ‘Monocyte Imaging after Myocardial Infarction with 19F-fluorine MRI at 3T—A Pilot Study in explanted porcine hearts’ of this issue of the journal, Boenner and colleagues present several key innovations in the field of 19F-CMR.1 With their approach, they were able to specifically visualize and quantify macrophage infiltrations in the ischaemia-reperfused myocardium on a clinical 3T MR system. While the conventional 1H-proton-based CMR is well known as a most powerful tool for tissue characterization, the 19F-fluorine technique can boost CMR to the next level, i.e. to that of cellular tissue characterization.2 Conventional CMR can identify tissue characteristics such as scar vs. normal viable myocardium, necrosis, oedema, ischaemia, iron overload, intramyocardial haemorrhage, microvascular obstruction, and others.3–5 However, the understanding of disease processes and the prediction of prognosis may be facilitated by characterizing the cellular composition of the tissue.6,7 19F-CMR can visualize 19F-containing molecules, which are incorporated into specific cell types, i.e. it can detect and track 19F-labeled cells non-invasively in the intact body (Figure 1). As 19F-fluorine is not present in the human body, 19F-MRI can detect this label with an excellent contrast-to-noise (CNR), as this technique lacks any background signal. If the magnetic resonance (MR) scanner is acquiring the 19F signal at its specific resonance frequency, it can detect 19F-containing molecules [so-called perfluorocarbons (PFC)] similar to nuclear medicine techniques. Unlike nuclear tracers, however, the signal emanating from the 19F-PFC is not decaying over time, and consequently, this technique can monitor 19F signals of migrating cells over hours up to days without radiation. To take full advantage of the 19F-MRI technique, further development of PFCs, of labeling strategies, and of 19F-MRI pulse sequences is needed. In their study, Boenner et al. used a particular PFC, perfluoro-octyl bromide (PFOB) with a biological half-life of 3–8 days,8 which is ideal to monitor cells over hours to days following injection. Interestingly, this compound is also considered safe, as it passed already Phase III clinical trials in the field of surgery, as this compound can facilitate oxygen transport in blood. It was therefore tested as a plasma expander by administration of large volumes of PFOB to patients.9 It is also possible to follow migration of different cell types by 19F-MRI simultaneously by using various PFCs which differ in their resonance frequency spectrum.10 Another research activity is exploring the effect of shortening the relaxation time of PFC by introducing gadolinium into the PFC, thereby increasing the signal of the PFCs.11 In the present study by Boenner et al., the PFOB was injected intravenously and the source of the 19F signal in the heart was the macrophage population which had incorporated the circulating PFOB. Other PFCs are designed for labelling various cell types in vitro. Once labelled in vitro, these specific cell types can be injected into the body and their fate is trackable by 19F-MRI in vivo. Currently, a first clinical study in patients is underway, where dendritic cells designed for colorectal cancer treatment are 19F-labeled ex vivo, are injected intradermally, and then tracked successfully in vivo by 19F-MRI.12 Another field of PFC research is exploring the SPIT technique (sterol-based post insertion technique), a promising platform that allows to attach different ligands (peptides, antibodies, etc.) to PFC nanoemulsions for in vivo19F-MRI. During thrombus formation, α2-antiplasmin cross-links with fibrin. To visualize this process, the SPIT technique was used to generate α2-antiplasmin-labelled PFC nanoemulsions (α2AP-PFC). In an animal model of deep vein thrombosis and pulmonary embolism, these targeted PFC emulsions were successful in detecting thrombi in vivo.13 While the PFOB features a favourable safety profile and a biological half-life to monitor cells over several days, its resonance spectrum is composed of several peaks which is challenging from a MR imaging perspective. To solve this problem, the authors of the current paper developed an elegant MR pulse sequence, which can handle the multi-peak spectrum of PFOB, and thus, they present in their paper a major step forward in the 19F-MRI development. Nevertheless, many MR pulse and reconstruction options are still to explore. For example, the so-called compressed sensing technique allows for acceleration of acquisition if sparsity in the data set is identifiable. This technique was successfully applied in patients for highly accurate LV volume and function analysis.14 This compressed sensing approach appears ideally suited for the reconstruction of 19F data as typically signals are sparsely represented in k-space as demonstrated by Ahrens et al., who accelerated 19F-MRI by a factor up to 8.15 As the simultaneous tracking of various cell types appears desirable, Bauer et al. presented a multi-spectral 19F technique. Such novel 19F-MR pulse sequences can probe PFCs with different resonance peaks that allow to track various cell types simultaneously and non-invasively in the body. At this point, it should be mentioned that the 19F signal obtained by 19F-MRI is fully quantitative as the 19F signal correlates linearly with the number of 19F-fluorine atoms in the imaged tissue volume.16,17 Finally, Boenner and co-workers addressed another important aspect of the 19F technique. Up to now, most studies utilizing this 19F technique were performed on high-field MR scanners (of 9T and higher) using small animal models. This current study is one of the first to demonstrate the feasibility to achieve a cellular characterization of the myocardium by the 19F technique on a clinical 3T MR scanner. In their study, they convincingly confirmed their 19F-MRI results with immunohistological and flow cytometry data. As no radiation is involved in this type of cell tracking, there appears a bright future for this technique to successfully be translated into clinical routine. In summary, the study of Boenner and colleagues convincingly demonstrates the potential of the PFOB for cellular characterization, when combined with a novel MR pulse sequence scheme and by translating this approach onto a 3T clinical MR machine. Hopefully, the presented findings will further accelerate the transition of this technique from high-field small animal experiments to clinical studies in patients. Once available in patients this 19F-MRI technique could represent a powerful tool to localize the site of inflammation and to quantify and monitor the response to treatment, e.g. in myocarditis but also in the setting of fever of unknown origin and in other inflammatory diseases. This tool may also allow to studying auto-immune diseases in more detail and could be of great value in the design and monitoring of immunotherapies in cancer patients. Finally, the concept of targeted PFC emulsions could substantially enhance our armamentarium to detect thrombi or other specific targets in the human body. As this technique is radiation free and can monitor cells over hours and days, there is a bright future for this technique to improve diagnosis and to monitor treatment responses in patients.