Evaluation of SSEA-4 as a CAR T cell therapeutic target for the treatment of chemoresistant triple negative breast cancers

Abstract

Triple negative breast cancer (TNBC) is an aggressive subtype of breast cancer that accounts for 15-20% of breast cancer incidences. Since it lacks estrogen receptor, progesterone receptor and HER2 overexpression, therapy options are limited to surgery, radiation and chemotherapy, which are curative only to a small fraction of TNBC patients. Accordingly, TNBC has the highest rates of metastatic disease and the poorest overall survival of all breast cancer subtypes. Recently, the sialoglycolipid stage-specific embryonic antigen 4 (SSEA-4) has been identified as an epitope whose overexpression in TNBC strongly correlates with metastasis and chemoresistance. As TNBC patients are typically exposed to multiple rounds of chemotherapy and SSEA-4-positive cells have been found to be enriched in residual tumors surviving this treatment, a sequential therapeutic approach using chemotherapy followed by SSEA-4-directed chimeric antigen receptor (CAR) T cell administration holds great promise to improve the therapeutic outcome of TNBC patients. Consequently, the aim of this study was to develop and evaluate an SSEA-4-directed CAR T cell-based treatment modality for TNBCs. In order to explore the optimal CAR configuration, five second generation chimeric receptors that differed in type and length of the spacer region were generated. All receptors contained a human/mouse cross-reactive single chain variable fragment which allowed a preclinical on target/off tumor toxicity profiling due to the correlated antigen expression between the mouse and the human organism. The three lead candidates, that showed good surface expression, activated T cells in an antigen-specific manner as characterized by T cell degranulation, secretion of inflammatory cytokines and effective killing of SSEA-4-expressing target cells in vitro. The efficacy of T cell activation was not uniform between the constructs and no direct correlation between CAR potency and CAR spacer length was observed. When the SSEA-4-directed CAR T cells were adoptively transferred into mice with subcutaneous TNBC cell line xenografts (MDA-MB-231), only the CAR variant that showed the strongest in vitro bioactivity could control tumor growth. No anti-tumor effect was detectable in mice cohorts receiving the other CAR T cell variants. In parallel to the tumor burden control, mice receiving T cells incorporating the most potent CAR variant exhibited toxicity symptoms and cachexia. It was found that the administered CAR T cells proliferated primarily in the lungs and in bone marrow with the latter demonstrating hypocellularity and a strong decrease of the CD45+Sca-1bright population upon therapy. In bone marrow, hematopoietic multipotent progenitor cells were identified to express SSEA-4 which were likely co-targeted by the CAR T cells. In the lungs, the SSEA-4-positive subset was found to co-express CD44, CD81, CD98, and Prominin-1 suggesting an epithelial pluripotent population. The severe side-effects observed in this study raise safety concerns for therapies targeting SSEA-4-positive tumor cells, as a high risk exists to collaterally damage vital pluripotent cells within the organism. Therefore, great care has to be invested in SSEA-4-directed immunotherapies to avoid life threatening side effects.