EXTH-02. AAV MEDIATED BRAIN DELIVERY OF AN ADCC-ENHANCED ANTIBODY OBVIATES XENOGRAFT GROWTH IN MOUSE MODELS OF HER2+ BREAST CANCER BRAIN METASTASIS

Abstract

Abstract BACKGROUND HER2+ tumors constitute approximately 20% of breast cancer patients and are characterized by overexpression of the growth factor receptor HER2 (ERBB2), a cell proliferation driver. Effective anti-HER2 therapies confer prolonged patient survival necessitating the need for transformative treatments targeting brain metastases, a major cause of mortality in ~30-50% of HER2+ metastatic breast cancer patients. HER2-directed antibody immunotherapy, while efficacious for peripheral disease, has limited central nervous system exposure (CNS). To overcome these challenges, we transduced CNS cells with a novel AAV vector carrying an anti-HER2 antibody payload. METHODS We assessed the biochemical equivalence and functional effectiveness of AAV vector-encoded antibodies using in vitro assays. After selecting promising vector-encoded antibody candidates, a novel, blood-brain barrier penetrant AAV capsid was administered via i.v. dosing to an orthotopic xenograft mouse model of HER2+ brain metastases. Bioluminescent imaging provided a longitudinal measure of brain tumor burden. At study termination, we measured antibody biodistribution in cerebrospinal fluid (CSF), serum, and brain homogenates with AlphaLISA assays. RESULTS Using HER2+ breast cancer cell lines, we determined that an antibody-dependent cell cytotoxicity (ADCC) enhanced anti-HER2 antibody was most effective and demonstrated that AAV-vector encoded forms of the antibody performed comparably to recombinant reference antibodies. Following i.v. administration of a HER2 antibody encoding AAV vector, we measured >1 ug/mL of the antibody in CSF. Importantly, AAV-mediated expression of the ADCC-enhanced HER2-directed antibody significantly abrogated tumor growth in orthotopic xenograft models. CONCLUSIONS Peripheral administration of an AAV vector was able to transduce brain tissue such that efficacious levels of HER2-directed antibodies were produced. This strategy was successful at preventing tumor growth in our physiologically relevant model of breast cancer brain metastases. Such a treatment modality should be further evaluated in patient derived PDX models to validate translational efficacy for human patients.