Purpose or Case Report: Metastasized osteosarcoma has a poor prognosis with a two-year event free survival rate of 15-20%, highlighting the need for the advancement of efficacious therapeutics. Chimeric antigen receptor (CAR) T-cell therapy is emerging as a potent strategy for eliminating tumors by harnessing the immune system. Despite this, trials encounter challenges due to varied responses and failing therapy in patients. Development of translatable cell imaging methods would provide invaluable information regarding the biological fate of T-cells. This work aims to evaluate a new approach for CAR T-cell labeling with ferumoxytol, which enables non-invasive cell tracking with magnetic resonance imaging (MRI) and a new imaging technique, magnetic particle imaging (MPI).
Methods & Materials: In vitro characterization: We isolated human T-cells from donors, expressed the CAR gene and stimulated for expansion. A new microfluidics device was used to label T-cells with ferumoxytol nanoparticles (Feraheme) through cell volume exchange for connective transfer. The iron content of labeled T-cells and unlabeled controls was analyzed by prussian blue histology and inductively coupled plasma optical emission spectrometry. To determine the sensitivity, varying numbers (10³-10⁶) of labeled T-cells and controls were imaged using MRI or MPI. In vivo imaging: Labeled or unlabeled T-cells were intravenously administered in osteosarcoma bearing NOD SCID mice and animals were imaged using MRI and MPI. The MPI signal intensity and T2 relaxation times of different groups were compared using an analysis of variance (ANOVA) and a linear regression analysis.
Results: In vitro characterization: Retention of ferumoxytol was significantly (p<0.01) greater in labeled T-cells (0.39 ± 0.003 pg/cell) than controls (0.06 ± 0.006 pg/cell). 10⁴ ferumoxytol labeled T-cells were visually detected using MPI and MRI. T2 relaxation times of labeled T-cells were significantly shorter compared to unlabeled for cell numbers ≥10⁴. The ferumoxytol MPI signal showed a linear correlation with the T-cell number. In vivo imaging: MPI demonstrated homing of T-cells to osteosarcoma in animals administered with labeled T-cells, whilst T-cells were not visualized in animals infused with unlabeled cells.
Conclusions: This study demonstrates the first successful labeling of CAR T-cells with ferumoxytol. The labeled cells could be detected in vitro and in vivo, using MRI and MPI, thereby paving the way for in vivo monitoring of CAR T cell trafficking to osteosarcomas.