Label-free imaging of CAR T cell metabolism

PROJECT SUMMARY / ABSTRACT The goal of this proposal is to develop non-invasive single-cell technologies to improve the potency of T cell therapies against cancer. The first 6 chimeric antigen receptor (CAR) T cell therapies were recently approved and >800 CAR and T cell therapies are in clinical trials. However, barriers remain in achieving durable remissions

(>1-year) for ~50% of patients who receive CAR T cell therapy. Due to the rapid development of these therapies and a great need for process optimization, we focus on improving three translational roadblocks to effective CAR T cell therapy: (1) screening patients whose T cells are unfit for CAR T cell manufacturing, (2) optimizing in vitro

CAR T cell production for higher potency, and (3) identifying metabolic features of potent CAR T cells in vivo. CAR T cell therapy could be improved by enriching for naïve and stem cell memory (SCM) T cells in starting materials and final products. Deficiencies in naïve and SCM T cells occurs in ~50% of untreated cancer patients,

and manufacturing autologous CAR T cell products from these sources has been unsuccessful. Even if SCM T cells can be isolated, after CAR incorporation, the expansion process typically diminishes potency through T cell exhaustion. After infusion, the presence of memory-like phenotypes in vivo correlate with better responses. To

date, there are no robust, non-destructive technologies to monitor CAR T cell manufacturing to optimize production and assess potency in vivo at a single-cell level. These issues limit the impact of CAR T cell therapy. Current approaches to measure T cell function are labor-intensive, destructive, or lack single-cell resolution,

which limits the frequency or specificity of these measurements. For CAR T cell therapy to realize its clinical potential, new methods are needed to monitor T cells for optimal potency throughout manufacturing and post- infusion. Changes in cell metabolism provide an attractive yet under-explored assay to track T cell potency.

Previous studies, including our own, show that T cells undergo drastic metabolic changes with activation, and that naïve, exhausted, and memory-like T cells have distinct metabolic features. Our preliminary data shows that non-invasive single-cell imaging of the fluorescence intensity and lifetime of NAD(P)H and FAD (optical

metabolic imaging, or OMI) can predict CAR T cell manufacturing conditions that produce a more vs. less potent anti-tumor response in vivo. Given these metabolic features of CAR T cell potency, we propose to determine whether label-free OMI of T cell autofluorescence and multivariate models can identify patient T cell

fitness, optimal in vitro expansion conditions, and in vivo cell biomarkers of potent and persistent CAR T cell response. Overall, these technologies will streamline processes and interventions for consistently potent T cell therapy and increase our knowledge of CAR T cell metabolism in vitro and in vivo.