71. Validation of Manufacturing NKG2D Chimeric Antigen Receptor T Cells for a First-In-Human Clinical Trial in AML/MDS and Multiple Myeloma

Abstract

Canonical CAR-T cells are engineered using constructs encoding a single chain variable fragment derived from a murine or humanized antibody, a costimulatory domain connected by a hinge region, and the signaling domain of CD3 zeta. This restricts T cells transduced with these constructs to recognizing one tumor antigen, such as CD19, and a limited set of cancers. Prior work in murine models demonstrated preclinical efficacy of T cells transduced with a novel activating CAR construct that exploits the ability of NK cells to recognize a variety of tumor antigens. Specifically, a CAR was generated by fusion of native full-length human Natural Killer Group 2D (NKG2D) gene with the human CD3 zeta cytoplasmic signaling domain. The NKG2D receptor in association with natural costimulatory molecule DAP10 creates a unique 6 protein receptor complex enabling NK and CD8+ T cells to kill many cells types via recognition of ligands including MIC-A, MIC-B, and UL-16 binding proteins 1-6. NKG2D-ligands are upregulated in malignancy but protein expression is absent or minor in healthy tissues. Complete remissions and durable CD4+ and CD8+ T-cell memory have been demonstrated in murine models of lymphoma, myeloma and ovarian cancer following adoptive therapy with NKG2D CAR T cells. At the same time, NKG2D CAR T cells significantly alter the tumor microenvironment through cytokine secretion to further promote anti-tumor immunity. This technology is being evaluated clinically in a phase I dose-escalation study (ClinicalTrials. gov NCT02203825) arising through collaboration between academic investigators and commercial sponsors. GMP manufacturing procedures for NKG2D CAR T cells were developed to support this first-in-human trial of NKG2D CAR T cells to assess safety and feasibility in acute myeloid leukemia/myelodysplastic syndrome and multiple myeloma. Following isolation of mononuclear cells, T cells were activated with anti-CD3 mAb and IL-2, subjected to 2 rounds of transduction with SFG retroviral vector containing the NKG2D CAR construct (CM-CS1), and expanded in media containing IL-2. GMP manufacturing procedures were validated using T cells from healthy donors as well as patients with AML and myeloma. Validation studies and initial clinical manufacturing demonstrated consistent viability, robust cell expansion, vector-mediated surface expression of NKG2D on a median of 53.6% CD8+ and 90.3% of CD4+ T cells, consistent viral copy number/cell less than 5, and the absence of replication-competent retrovirus in CM-CS1 T cells. Functionally, CM-CS1 T cells exhibit potent IFN-γ production following exposure to NKG2D ligand-expressing tumor cells. We also demonstrated minimal carry-through of malignant myeloid or plasma cells after deriving autologous CM-CS1 T cells from patients with malignancies. The Phase I trial using CM-CS1 T cells is ongoing, with dose-escalation proceeding as planned. Future goals include application of this novel cell therapy to additional malignant diseases.