169 : Direct in vitro generation of antigen specific, memory phenotype, Cd8+ T cells using a cocktail of small molecules and gamma-chain cytokines

Abstract

Adoptive T cell transfer is an emerging cancer immunotherapy that has yielded complete responses in solid tumors such as B cell lymphoma and melanoma that are incurable by standard therapies. Prior studies have demonstrated that transfer of less differentiated T cells with greater proliferative capacity is associated with better clinical responses. Conventional expansion of antigen specific CD8 T cells uses high dose IL-2 and generates effector cells with high cytolytic and little proliferative capacity. We hypothesized that altering culture conditions by adding known cytokines and small molecules that inhibit acquisition of effector function and encourage memory formation would be associated with greater efficacy of transferred cells. We have used peptide antigen pulsed dendritic cells combined with a cocktail of homeostatic and memory-inducing cytokines that bind to the common gamma chain (IL-7 and IL-21) along with small molecules that inhibit glycolysis and enhance Wnt signaling to rapidly generate tumor specific, less-differentiated CD8 T cells with memory properties. Cells grown by this method are highly antigen specific (>90%) and are ∼50% central memory (CD62L+CD44+) or naive (CD62L+CD44−) phenotype versus virtually undetectable levels of these cells in cultures grown under standard conditions (i.e. with large amounts of IL-2). Compared to highly activated effector cells grown in IL-2, these memory-like cells engraft more efficiently into lymphopenic tumor bearing mice, maintain memory (CD62L+) phenotype for several weeks, and have greater anti-tumor efficacy in vivo. This approach is particularly useful since rapid advances in bioinformatics will soon make practical the identification of CD8 T cell epitopes derived from patient specific mutations. Since this method represents a way to generate more effective T cells that target tumor specific mutations, these results have significant implications for extending the usefulness of adoptive transfer to non-immunogenic (i.e. non-TIL bearing) solid tumors that could be targeted with personalized immunotherapy.