Targeting Ewing Sarcoma By Genetically and Epigenetically Modified Ex-Vivo Expanded NK Cells in Combination with NKTR-255 and Dinutuximab

Background Metastatic Ewing sarcoma (ES) has a dismal prognosis, largely due to therapy resistance within the tumor microenvironment (TME). NK cell resistance in solid tumors is mainly due to the small number and lack of specific targeting of NK cells, poor NK cell function and persistence, TGFβ-mediated immune suppression, and inability of NK cells to infiltrate the tumor [1]. IL1RAP is a novel immunotherapy target highly expressed on ES cells [2]. Addition of TGFβ during NK expansion epigenetically reprograms NK cells (TGFβi-NK) and enhances NK-specific lysis of tumor cells [3]. Dinutuximab is an FDA approved mAb against GD2 which is overexpressed in ES. NKTR-255 is a recombinant hIL-15 agonist that boosts NK cells, improves NK persistency and enhances response to ADCC-mediated therapy. CXCR2 is an IL-8 receptor known to facilitate migration and tumor infiltration of NK cells [4]. Here we develop “suppression proof” NK cells by genetic and epigenetic modification in combination with immune modulators to overcome the ES TME resistance. Methods PBMCs were expanded into NK cells using feeder cells or into TGFβi-NK cells by TGFβ selection. The IL1RAP CAR NK cells and CAR CXCR2 NK cells were generated by electroporation of CAR or CAR CXCR2 mRNA into expanded NK cells. NK cell cytotoxicity was evaluated by luciferase based cytotoxicity assay. IFN-γ and perforin secretion were analyzed by ELISA. An orthotopic mouse model by implanting ES cells into the tibia of NSG mice was utilized for evaluating in vivo tumor growth and animal survival. Results We found a significantly increased cytotoxicity of IL1RAP CAR NK cells compared to mock NK cells against ES cells at various effector to target ratios (p<0.01 and p<0.05) (Fig 1A). IL1RAP CAR NK cells secreted significantly higher levels of IFN-γ and perforin than mock NK cells (p<0.01 and p<0.05). The enhanced cytotoxicity of CAR NK cells is due to specific targeting of IL1RAP because no significant increase in cytotoxicity with CAR NK compared to mock NK was observed in the IL1RAP knockout cells as we did in the wildtype cells (Fig 1B). IL1RAP CAR expression in TGFβi-NK cells further enhanced NK cell cytotoxicity in vitro against ES cells (Fig 1C) and TGFβi-NK cells significantly slowed down ES tumor growth in vivo (p<0.05) (Fig 1D). Furthermore, we found that NKTR-255 and dinutuximab synergistically enhanced in vitro cytotoxicity of IL1RAP CAR NK cells against ES cells (Fig 1E). The IL1RAP CAR NK cells armored with CXCR2 showed significantly enhanced migration toward IL-8 conditioned media of ES cells (p<0.01) (Fig 1F). Conclusions Our data demonstrated enhanced anti-tumor efficacy of modified NK (IL1RAP CAR NK, TGFβi-NK, IL1RAP CAR CXCR2 NK) cells alone and combined with NKTR-255 and dinutuximab against ES in vitro. This provided a rationale for a preclinical evaluation of the anti-tumor effects of these modified NK cells combined with NKTR-255 and dinutuximab in vivo.