Cellular immunotherapies are poised to transform the treatment of cancer and immunological disorders. In the most promising setting to date, genetic modification to the T lymphocytes in the form of chimeric antigen receptors (CAR) has dramatically increased therapeutic efficacy with reported initial complete remission rates in acute lymphoblastic leukemia ranging between 80-100%. However, pressing challenges remain to be solved to ensure that engineered T-cell immunotherapies can be cost-effectively and consistently manufactured, and safely and reliably delivered at the scale necessary to support wide patient base commercialization.

Adoptive cellular therapies using engineered chimeric antigen receptor T cells (CAR-T cells) are rapidly emerging as a highly effective treatment option for a variety of life-threatening hematological malignancies. Small molecule-mediated modulation of T cell differentiation during the in vitro CAR-T manufacturing process has great potential as a method to optimize the therapeutic potential of cellular immunotherapies. In animal models, T cells with a central or stem memory (TCM/SCM) phenotype display enhanced in vivoefficacy and persistence relative to other T cell subpopulations. We sought to identify small molecules that promote skewing towards a TCM/SCM phenotype during the CAR-T manufacturing process, with associated enhanced viability, expansion and metabolic profiles of the engineered cells. To this end, we developed a high-throughput functional screening platform with primary human T cells using a combination of high-content immunophenotyping and gene expression-based readouts to analyze cells following a high-throughput T cell culture platform that represents a scaled-down model of clinical CAR-T cell production. Multicolor flow cytometry was used to measure expansion, cell viability and the expression levels of cell surface proteins that define TCM cells (e.g., CCR7, CD62L and CD27) and markers of T cell exhaustion (e.g., PD1, LAG3, and TIM3). In parallel, a portion of each sample was evaluated using high content RNA-Seq based gene expression analysis of ~100 genes representing key biological pathways of interest. A variety of known positive and negative control compounds were incorporated into the high-throughput screens to validate the functional assays and to assess the robustness of the 384-well-based screening. The ability to simultaneously correlate small molecule-induced changes in protein and gene expression levels with impacts on cell proliferation and viability of various T cell subsets, enabled us to identify multiple classes of small molecules that favorably enhance the therapeutic properties of CAR-T cells. Consistent with results previously presented by Perkins et al. (ASH, 2015), we identified multiple PI3K inhibitors that could modify expansion of T cells while retaining a TCM/SCM phenotype. In addition, we identified small molecules, and small molecule combinations, that have not been described previously in the literature that could improve CAR-T biology. Several of the top hits from the screens have been evaluated across multiple in vitro (e.g., expansion, viability, CAR expression, serial restimulation/killing, metabolic profiling, and evaluation of exhaustion markers) and in vivo (e.g., mouse tumor models for persistence and killing) assays. Results from the initial screening hits have enabled us to further refine the optimal target profile of a pharmacologically-enhanced CAR-T cell. In addition, we are extending this screening approach to identify small molecules that enhance the trafficking and persistence of CAR-T cells for treating solid tumors. In conclusion, the approach described here identifies unique small molecule modulators that can modify CAR-T cells during in vitro expansion, such that improved profiles can be tracked and selected from screening through in vitro and in vivo functional assays. Disclosures Rosen:Fate Therapeutics: Employment, Equity Ownership. Rezner:Fate Therapeutics, Inc: Employment, Equity Ownership. Robbins:Fate Therapeutics: Employment, Equity Ownership. Hardy:Fate Therapeutics: Employment, Equity Ownership. Peralta:Fate Therapeutics: Employment, Equity Ownership. Maine:Fate Therapeutics: Employment, Equity Ownership. Sabouri:Fate Therapeutics: Employment, Equity Ownership. Reynal:Fate Therapeutics: Employment. Truong:Fate Therapeutics: Employment, Equity Ownership. Moreno:Fate Therapeutics, Inc.: Employment, Equity Ownership. Foster:Fate Therapeutics: Employment, Equity Ownership. Borchelt:Fate Therapeutics: Employment, Equity Ownership. Meza:Fate Therapeutics: Employment, Equity Ownership. Thompson:Juno Therapeutics: Employment, Equity Ownership. Fontenot:Juno Therapeutics: Employment, Equity Ownership. Larson:Juno Therapeutics: Employment, Equity Ownership. Mujacic:Juno Therapeutics: Employment, Equity Ownership. Shoemaker:Fate Therapeutics: Employment, Equity Ownership.

While allogeneic hematopoietic cell transplantation (allo-HCT) is a potentially curative option for many hematologic malignancies, complications such as graft-versus-host disease (GVHD) result in significant morbidity and mortality. Conventional approaches to manage GVHD, such as prophylaxis with immunosuppressive agents or T-cell depletion strategies, are limited by increases in graft failure, viral-associated lymphoproliferative disorders, and disease relapse. Here we present a novel strategy to reduce the rates of GVHD by programming mobilized peripheral blood ex vivo with a cocktail of small molecules prior to allo-HCT. An established xenogeneic mouse model was used to examine the potential of this cell programming strategy to reduce rates of GVHD. Sub-lethally irradiated NOD-scid IL2rγnull (NSG) mice were transplanted with human peripheral blood mononuclear cells (PBMC) pulse treated ex vivo with either vehicle or a cocktail of two small molecules (FT1050+FT4145). Recipients of pharmacologically programmed PBMCs had significantly lower GVHD scores, decreased levels of circulating IFN-ɣ and enhanced survival relative to recipients of vehicle PBMCs (p<0.0001, Mantel-Cox log rank). In addition to xenograft-GVHD studies, we explored the impact of this cell programming strategy in a murine model of GVL. Lethally irradiated BALB/c (H-2Kd) recipient mice were transplanted with either control or FT1050+FT4145 programmed C57BL/6 (H-2Kb) CD8+ T cells and T cell-depleted bone marrow. Prior to allo-HCT, recipients were injected with 2x104 luciferase-expressing A20 lymphoma cells (A20-luc). Bioluminescence imaging was used to monitor the tumor burden over a 28 day period post-HCT. Donor cells programmed with this small molecule cocktail significantly improved survival (p<0.001) while retaining GVL effects against the A20 lymphoma cells. Combined, these studies demonstrate that pharmacologic programming of hematopoietic cells with FT1050+FT4145 prior to allo-HCT may offer an innovative therapeutic approach to reduce rates of GVHD without compromising GVL activity.

16,16 dimethyl prostaglandin E2 (FT1050) was previously identified to be a critical regulator of hematopoetic stem cell (HSC) homeostasis and we hypothesized that a brief ex vivo modulation could improve patient outcomes in umbilical cord blood (UCB) transplantation by increasing the “effective dose” of HSCs. Using preclinical models, we have demonstrated that pulse treating human HSCs with FT1050 significantly enhances the homing and engraftment of donor cells to the bone marrow niche, in part by upregulating CXCR4.