Optimization of manufacturing conditions for chimeric antigen receptor T cells to favor cells with a central memory phenotype.

BACKGROUND:Chimeric antigen receptor (CAR)-T cells are genetically engineered to recognize tumor-associated antigens and have potent cytolytic activity against tumors. Adoptive therapy with CAR-T cells has been highly successful in B-cell leukemia and lymphoma. However, in solid tumor settings, CAR-T cells face a particularly hostile tumor microenvironment where multiple immune suppressive factors serve to thwart the anti-cancer immune response. Clinical trials of solid tumor antigen-targeted CAR-T cells have shown limited efficacy, and issues for current CAR-T cell therapies include failures of expansion and persistence, tumor entry, deletion and functional exhaustion. METHODS:We compared our standard protocol for CAR-T cell manufacturing, currently used to generate CAR-T cells for a phase 1 clinical trial, with two alternative approaches for T-cell activation and expansion. The resulting cultures were analyzed using multicolor flow cytometry, cytokine bead array and xCELLigence cytotoxicity assays. RESULTS:We have found that by changing the method of activation we can promote generation of CAR-T cells with enhanced CD62L and CCR7 expression, increased interleukin (IL)-2 production and retention of cytolytic activity, albeit with slower kinetics. DISCUSSION:We propose that these phenotypic characteristics are consistent with a central memory phenotype that will better enable CAR-T cell survival and persistence after activation in vivo, and we aim to test this in a continuation of our current phase 1 clinical trial of CAR-T cells in patients with advanced melanoma.