Chimeric antigen receptor (CAR) T cells Imaging: Clinical Needs and Strategies for Success

Introduction: To review the recent literature on various methods used to label CAR T cells for in vivo tracking with PET and to discuss the importance of imaging and monitoring of CAR T cell trafficking.

Methods: Databases including Google Scholar and PubMed were searched to compile literature relevant to CAR T cell labeling and in vivo trafficking. Studies were separated by the method by which the CAR T cells were labeled including direct labeling and usage of reporter genes, as well as a collection of isotopes used for labeling: fluorine-18, zirconium-89, gallium-68, and copper-64. Studies without PET were excluded.

Results: Despite the successful achievement of CAR T cell therapy, there are several limitations: 1. inadequate CAR T cell expansion and trafficking to tumor sites lower the rates of long-term response, 2. dearth of objective responses in solid tumors, and 3. development of severe complications such as cytokine release syndrome (CRS), neurotoxicity, and death. Radiolabeling CAR T cells would allow in vivo characterization of their expansion and trafficking to tumor sites, in addition to early risk assessment metrics for treatment toxicity and prediction of therapeutic outcomes. Tracking the CAR T cells could thus make feasible the timely incorporation of strategies to reduce toxicities and improve response.

To gain insight into mechanisms for variable patient outcomes, it may be useful to monitor the cells’ activity (proliferation, kinetics, and spatiotemporal distribution). There are two main methods to label CAR T cells for PET imaging, each with advantages and limitations. The direct approach involves ex vivo incubation of CAR T cells with a radiolabeled probe. Commonly-used isotopes include fluorine-18, gallium-68, copper-64, and zirconium-89; the latter two have half-lives long enough to enable serial imaging and tracking of the cells for multiple days. A disadvantage of direct-labeling is signal attenuation with cell division and death. Investigations have included [89Zr]-DFO-labeled CAR T cells in mice, and ex vivo superparamagnetic iron oxide nanoparticles (SPION) copper-64 labeling for in vivo tracking in humans.

Another method is in vivo labeling by transducing CAR T cells to express a reporter gene targeted by a radiolabeled probe. Nearly all reporter systems have been performed on mouse models, including the sodium-iodine symporter (NIS) and human norepinephrine transporter (hNET). The exception is herpes simplex virus type 1–thymidine kinase (HSV1-TK) which has been investigated in clinical studies with fluorine-18-labeled trimethoprim as the probe. Unlike direct labeling, reporter labeling does not suffer from signal dilution, but challenges such as burden and cost of labor and potential immunogenicity of non-human genes must be weighed.

With both labeling methods, a serious issue to consider is whether labeling the CAR T cells interferes with their therapeutic function. The results are mixed: some mouse studies report negligible changes to CAR T cell viability and activity, while others report a reduction of cytotoxicity.

Conclusions: The importance of tracking CAR T cells in the body using PET cannot be overstated, as it avails the effectiveness of the treatments in real time based on cell localization. Monitoring the activity of the therapeutic cells over the course of treatment could be especially useful for solid tumors where CAR therapy has had limited success in comparison to hematological malignancies. The recent work on CAR T cell labeling highlights the advantageousness of cell labeling using PET/CT, and future research would benefit from similar labeling for example with transplanted stem cells in regenerative medicine. More research is warranted for the usage of cell tracking to mitigate damage from CRS and neurotoxicity as a result of off-target CAR T cell activity.