Defining the determinants of long term CAR-T persistence

About This Grant

While some patients with previously untreatable blood cancers have achieved long-term remissions following CAR T cell therapy, others relapse after months to years. While engraftment and expansion are well characterized clinical surrogates of a durable anti-tumor effect, the attributes and origin of CAR T cells with prolonged function remain unknown. Prior research on CAR T cell-intrinsic properties in the treatment of hematologic malignancies has primarily focused on the manufactured product, but did not capture or define the T cell clones that are directly responsible for functional persistence and durable remissions. We have now found that long-lived CAR T cells up to 5 years post-infusion have a distinct CD4- CD8a- “double-negative” (DN) TCRαβ+ effector memory T cell phenotype. Our multimodal single-cell analyses have revealed that (i) the clones that eradicate leukemia in the initial phase differ from those that perform long term immunosurveillance, and that (ii) long lived DN CAR T cells show specific transcriptional signatures, with remarkable similarity to persistent CAR T cells in small cohort of children with B-ALL, to rare TCRαβ+ DN T cells from healthy donors, and to melanoma patients responding to dual checkpoint blockade. The CD4 and CD8 glycoproteins are co- receptors in TCR recognition of class II or class I major histocompatibility (MHC) – peptide complexes, respectively, serving to stabilize the TCR:peptide/MHC. Since DN T cells accumulate in chronic inflammatory diseases, these observations raise the intriguing possibility that in some settings, chronic antigen exposure can lead to a unique and previously under-appreciated T cell differentiation trajectory that is distinct from that found in “classic” CD8 T cell exhaustion. This dual PI proposal unites experts in immune cell therapies and evolutionary biology to test the overarching hypothesis that the fittest CAR T cells adopt a DN phenotype to withstand the demands of long-term immunosurveillance. In Aim 1 we will define the contribution of DN T cells to the long-lived CAR T cell population in patients with hematologic malignancies using flow cytometry and single cell multiomic analyses of CAR T cells persisting >12 months after infusion in patients with CLL, NHL, and MM. By leveraging our extensive cryorepository and well-annotated clinical trial outcome data we will be able to correlate CAR T cell phenotype to their functional persistence. In Aim 2 we will identify the cellular origin and evolutionary paths of long-lived CAR T cells. To distinguish the possibilities that persistent CAR T cells originated as true DN T cells vs. conventional T cells that downregulate the CD4 or CD8 co-receptors, we will use complementary approaches to trace DN persistent clonotypes back to their infusion origin. In Aim 3 we will determine whether and how DN T cells resist exhaustion during chronic stimulation, using a novel high-fidelity T cell exhaustion model along with gain- and loss-of function interventions to illuminate this biology. Together, these studies will illuminate the lineage relationships of various endogenous and engineered T cells and guide the development of more effective CAR T cell therapies.