Advancements in cell expansion methods and activation have led to a re-emergence of tumor-specific immunotherapy for patients.¹ However, little is known regarding trafficking patterns of infused immune cells. Cell tracking with non-invasive imaging modalities may allow improved assessments of the efficacy of cellular-based immunotherapies by enabling longitudinal detection and quantification of immune cells after infusion. Perfluoropolyethers (PFPE) have been used for labeling immune cells both in preclinical studies and recently in a clinical trial for colorectal cancer patients using fluorine-19 (¹⁹F) MRI.^2,3 However, ¹⁹F MRI suffers from inherently low sensitivity that could preclude its widespread clinical adoption. Recently, [Zirconium-89]-desferrioxamine-Bn-NCS-p (⁸⁹Zr-DBN) has been introduced as a cell labeling approach utilizing the high sensitivity of positron emission tomography (PET).⁴ This work aims to combine ⁸⁹Zr-PET with ¹⁹F-MRI to access the complementarity of information afforded by PET and MR techniques.

All animal experiments complied with our institutional animal care and use regulations.

MRI and PET/CT data were acquired on a 4.7T small animal MRI system (Agilent Technologies) and an Inveon Hybrid microPET/CT (Siemens), respectively. A 2D multi-echo spin-echo with 1.1x1.1mm² resolution, 2000/9.0ms TR/TE, 16 echoes, and 42.6min scan time was used for ¹⁹F MRI acquisitions and a T₁-weighted GRE with 0.28x0.28mm² res., 80.2/3.4 ms TR/TE, 20° FA was used for ¹H MRI. A ROI-based method with low-SNR correction was used for cell quantification.³ PET images were reconstructed using three-dimensional ordered subset expectation maximization (OSEM3D) prior to correction with a CT-derived attenuation map.

Human natural killer (hNK) cells were isolated from the lymphocyte fraction from healthy donor peripheral whole blood cells by magnetic cell separation beads (Miltenyi Biotec). After hNK expansion, cells were incubated for 24h in CS-ATM-1000 (Celsense), a commercially available emulsified PFPE tracer agent, in culture media. 7×10^{6 19}F-labeled hNK cells were then harvested and washed three times to remove excess PFPE prior to subcutaneous injection into an NSG mouse bearing a human mantle cell lymphoma (Z138) on the right flank.

⁸⁹Zr-oxalate (⁸⁹Zr, t_1/2=78.4h) was eluted from 50mg hydroxylamine functionalized resin, as previously described5 before being trapped on a Sep-Pak Light QMA cartridge, rinsed with 40mL H₂O, and eluted in 0.75mL 1M HCl. This activity was buffered by 300mg HEPES before 4µl of 5mM DBN in DMSO is added for 30min of complexation. Chelation efficiency was found to be >97% after 30 minutes (n=3). Yac-1 (mouse lymphoma) cells were washed and re-suspended in 400µl of HBSS-HEPES buffer set to pH=7.5 prior to 30min incubation in 37°C shaker 100µL ⁸⁹Zr-DBN solution. ⁸⁹Zr-labeled cells were washed with culture medium and re-suspended in PBS. 12×10⁶ Yac-1 cells were intravenously injected (~7.8µCi) into a healthy female ICR mouse.

The PET-MRI animal imaging platform was fabricated with pieces of acrylic that were laser cut and then welded together. Fiducial wells were manually machined and sealed with small laser-cut circular pieces of acrylic after filling with Cobalt-56 (⁵⁶Co, t_1/2=77.1d, β+=19.7%) solution.

Cell labeling with ¹⁹F-PFPE and ⁸⁹Zr-DBN required 24 hour and 30 minute incubation periods, respectively, and could be integrated into the clinical workflow for immunotherapies (Figure 1). ¹⁹F cell tracking enables longitudinal detection of localized cell injections with no decay of signal over days to weeks and the ability to quantify the number of cells within a voxel (Figure 2). ⁸⁹Zr cell tracking provides high sensitivity for intravenous delivery of cells with longitudinal signal detection (Figure 3). The multi-modal platform contains a long-lived radioactive fiducials visible by PET and MRI enabling co-registration between ¹H MR, ¹⁹F MR, ⁸⁹Zr PET, and CT images (Figure 4).The high sensitivity of ⁸⁹Zr PET enables detection of systemic migration patterns for intravenously infused cells. ¹⁹F MRI supports longer, longitudinal and quantitative assessment of tumor response primarily through localized injections but possibly using longer imaging times focused on the expected target lesion of immune cells. Moreover, PFPE does not utilize ionizing radiation and is more readily translated to clinical studies. A dual-label (⁸⁹Zr and ¹⁹F) approach may help to optimize methodology, or possibly enhance the ability to cell track in vivo during the early phase of cell migration, while ¹⁹F MRI would provide follow-up signal to quantify cell number within target lesion. Future work aims to combine both cell tracking approaches utilizing the dual-modality imaging platform in both small and large animals models on a PET/MRI system.Cell tracking with ⁸⁹Zr-DBN PET provides high sensitivity while ¹⁹F MRI cell tracking provides in vivo cell quantification and has been performed clinically. A dual-labeled (⁸⁹Zr and ¹⁹F) cell labeling approach may inform and potentially improve in vivo cell tracking sensitivity and clinical adoption.