Introduction

Glioblastoma multiforme (GBM) is the most common and most aggressive brain cancer for which classical treatment options remain limited. Recent advances in adoptive cell transfer applied in other cancers hold great promise for the treatment of GBM. The use of engineered chimeric antigen receptor (CAR) T cells is experiencing a period of intense research and development activity [1]. It is critical to develop tools to quantitatively assay T-cell biodistribution and survival after delivery to correlate with putative therapeutic effects. Our preliminary results show that ¹⁹F NMR cytometry [2] is a promising tool for preclinical CAR T cell detection. In this approach, T-cells are quantitatively labeled in culture with a perfluorocarbon (PFC) nanoemulsion and then infused to animals; at the experimental endpoint, organs and tissues are necropsied and subjected to intact-tissue ¹⁹F NMR spectroscopy to measure total ¹⁹F content and apparent T cell numbers. This approach yields a high-throughput method to quantify CAR T cell distribution in a model of glioblastoma multiforme (GBM) using intact tissue samples, which can be correlated to tumor growth.

Methods

Human PBMC-isolated T cells were transduced with a CAR lentiviral vector to express a surface antibody against EGFRvIII, a common receptor in GBM [3]. Transduction efficacy and phenotype of the T cells was confirmed by FACS. Additionally, U87-EGFRvIII human cancer cells were transduced with an established luciferase reporter to monitor tumor progression (Referred to as U87-EGFRvIII-Luc). In vitro apoptosis induction by co-culture of CAR T cells with U87-EGFRvIII cells was evaluated at 6 h, 12 h and 24 h. Female SCID mice (N=30) received bilateral sub-cutaneous injections of 5 million U87-EGFRvIII-Luc cells. Once tumors were established, the mice were divided into three groups. One group received 20 million CAR T cells intravenously (i.v.), another received untransduced T cells and the third remained untreated. CAR T cells and naïve T cells were intracellularly labeled with a PFC emulsion (CS-1000, Celsense, Inc.) ex vivo and the mean cell uptake was measured by ¹⁹F NMR to yield the ¹⁹F/cell [4]. Tumor growth was monitored over 7 days with bioluminescence imaging and caliper measurements. Intact organs were then harvested for fluorine-19 NMR measurements to quantify their total fluorine content. The apparent cell number per organ was deduced from the average PFC loading per cell and did not take in vivo cell division into account. Histopathological staining against HLA (human marker) and CD3 (T lymphocyte marker) were performed to correlate T-cell homing in tissues.

Results

Two weeks after transduction, 85% of the human T cells expressed the CAR receptor (Fig. 1.A). Labeling experiments with PFC nanoemulsion showed optimum uptake when incubated with 10 mg/ml overnight. PFC labeling did not impair T-cell viability or phenotype (Fig. 1.B and C). In co-incubation experiments with U87-EGFRvIII-Luc cells, CAR T cells were significantly more cytotoxic than untransduced T cells. CAR T cells caused 93% cell death at 24 h compared to 53% for untransduced T cells. Longitudinal bioluminescence acquisitions showed considerable tumor regression 7 days after CAR treatment with an average radiance of 5.02x10¹⁰ photons/sec, which was twice lower than the luminescence measured for both naïve T-cell treated and untreated groups (p=0.0001, Fig. 2.A and B). Tumor volume averages showed 50% reduction in tumor growth for the CAR T-cell treated group (p=0.0007) and no significant difference between untransduced T-cell treated and untreated groups (p=0.53). NMR measurements at day 2 post treatment showed greater CAR T-cell homing to the spleen and tumors compared to untransduced T cells (p=0.01 and 0.04 respectively). The number of CAR T cells found in the tumors remained stable until day 7, whereas naïve T cells could no longer be detected. The liver signal likely represented the dead T cell fraction, averaging ~15% of the injected dose in both groups. No significant differences were found regarding lymph node biodistribution between groups. The tail signal represents the mis-injected cell portion. Histopathological staining confirmed the presence of CAR T cells and untransduced T cells in the spleen and tumors, consistent with the NMR results.

Conclusions

¹⁹F NMR imaging is a rapid and quantitative technique to evaluate adoptive cell transfer biodistribution on intact tissues. This technique provides more comprehensive information compared to histopathology, which relies on indirect detection via antibodies and quantification approximation from a limited number of slides. Overall, ¹⁹F NMR imaging, in conjunction with bioluminescence imaging, may accelerate the timeline to evaluate new immunotherapeutic cell candidates by providing a rapid and straightforward method to evaluate cell therapy distribution, cell fate, and efficacy in preclinical studies.

Acknowledgements

This work was supported by the NIH grant R01-EB017271 and CIRM grant LA1-C12-06919

References

[1] Jackson HJ, et al. Nat Rev Clin Oncol. Jun;13(6):370-83, 2016; [2] Ahrens ET et al. NMR Biomed. Jul;26(7):860-71, 2013; [3] Johnson LA et al. Sci Trans Med. Feb 18;7(275):275ra2, 2015; [4] Zhong J et al. Plos One. Oct 20;10(10):e0140238, 2015