797. Targeted Elimination of Prostate Cancer by Genetically Directed Human T Lymphocytes

Abstract

The genetic engineering of T cells is a novel strategy designed to accelerate the generation of tumor-specific T cells and remedy the biological limitations that constrain the anti-tumoral functions of normal T cells. Chimeric antigen receptors (CARs) are essential constituents of this new armamentarium. Unlike the physiological TCR, CARs encompass immunoglobulin variable regions or receptor ligands as antigen-recognition elements, thus permitting T cells to recognize cell surface tumor antigens in the absence of HLA expression. T cell activation is mediated by the CARs cytoplasmic domain, typically derived from the CD3z chain or the FcRIgchain. The signaling function of CARs has been questioned based on findings in transgenic mice and transduced primary T cells indicating that antigen ligation was not sufficient to activate resting naive mouse T cell or elicit IL2 secretion in retrovirally transduced human peripheral blood T cells. We and others subsequently showed that z chain-based CARs could induce strong activation capable of sustaining T cell proliferation and permitting secondary antigenic restimulation in vitro, provided that antigen was presented in the context of CD28-mediated costimulation. It is not known, however, whether T cells expanded in this manner, particularly human T cells, could mediate tumor eradication in vivo, and whether further in vivo costimulation would be needed to sustain their function. Here we investigate the in vivo function of Pz1, a CAR targeting human PSMA. PSMA is a highly attractive target antigen that is found in most prostate cancer cells, including hormone refractory metastatic disease. The Pz1 receptor encompasses the z chain of the CD3 complex as its activation domain and specifically redirects in vitro cytolysis of retrovirally transduced peripheral blood T lymphocytes (PBLs) against PSMA-positive tumor cell lines. To investigate whether expanded Pz1+ T cells are active in vivo and require costimulation after adoptive transfer, we established three tumor models in SCID-bg/bg mice- orthotopic, subcutaneous and pulmonary. Direct administration of Pz1+ T cells in orthotopic and subcutaneous human prostate tumors ranging in size from 0.04 to 0.28 cm3 eliminated a majority of the tumors. T cells transduced with a control receptor had no anti-tumor effect. We next developed a systemic tumor model using a murine tumor expressing human PSMA in which mice succumbed to pulmonary tumors in 14 days. Intravenous administration of Pz1-transduced T induced objective responses in all mice and cured a substantial fraction of them. Interestingly, lung nodules continued to increase in size for several days after T cell infusion, before regressing completely. This finding demonstrates the critical importance of activating T cells so that they survive and sustain their function for at least 7-10 days after infusion. Altogether, our results strongly support the feasibility of targeting prostate cancer with autologous T lymphocytes directed against prostate-specific membrane antigen by a z chain-based receptor.