P08.23.A CAR-ENGINEERED PRIMARY NK CELLS SHOW PROMISING EFFECTS AS TARGETED THERAPY AGAINST EGFR-POSITIVE GLIOBLASTOMA

Abstract BACKGROUND Glioblastoma (GB) is the most frequent primary brain tumor with a dismal prognosis despite multimodal treatment regimens. Effector cell therapies with T cells or natural killer cells (NK) carrying tumor-specific chimeric antigen receptors (CAR) hold promise for the development of more effective treatment strategies. CAR-engineered derivatives of the human NK cell line NK-92 targeting EGFR or EGFRvIII showed potent antitumor activity against glioblastoma cell lines in vitro and in vivo. Nevertheless, NK-92 cells have limitations when compared to primary NK cells, such as the required irradiation prior to application in humans. METHODS To overcome this restriction, we evaluated the suitability of peripheral blood derived human primary NK cells (pNK) as carriers for a CAR targeting the common GB antigen epidermal growth factor receptor (EGFR) and the tumor-specific EGFR variant EGFRvIII. pNK cells from healthy donors were transduced with a lentiviral vector encoding the cetuximab-derived 225.28.z CAR (CAR-pNK), which recognizes an epitope common to both receptors. Specific tumor cell lysis by the CAR-pNK cells was then investigated in FACS-based cytotoxicity assays with GB cells expressing EGFR, EGFRvIII, or both target antigens. In addition, we tested in vivo CAR-pNK cell therapy in subcutanous and intracranial LNT-229/EGFRvIII glioblastoma xenograft models. RESULTS When compared to non-transduced pNK cells, CAR-pNK cells achieved enhanced lysis of LNT-229 GB cells expressing EGFR and EGFRvIII. Likewise, markedly increased cytotoxicity of CAR-pNK was observed against the murine EGFR transduced GB cell line GL261 and the human cell lines SKMG-3 and BS153 with intrinsic EGFR gene amplification or expression of EGFRvIII, respectively. Furthermore, we observed delayed tumor growth following a single intravenous injection of CAR-pNK cells in the subcutaneous LNT-229/EGFRvIII glioblastoma xenograft model. In the intracranial glioblastoma xenograft model, repetitive local CAR-pNK cell injections significantly increased survival compared to untreated mice, with a meaningful reduction of tumor volume. Moreover, we detected CAR-pNK cell persistence in the tumor for at least 14 days after the last intracranial effector cell injection. CONCLUSION In summary, these results demonstrate promising efficacy of primary CAR-NK cells targeting EGFR and EGFRvIII both in vitro and in vivo, and may provide the rationale for early clinical testing. Further in-depth in vivo and in vitro analysis including in an in vitro patient-derived organoid model is ongoing.