Background GPC3-directed CAR-T had shown promising safety but limited efficacy in the treatment of hepatocellular carcinoma (HCC). We develop an armored GPC3-directed CAR-T construct by co-expressing GPC3-CAR and a secreted CD70 molecule. CD70 interacts with CD27 on T cells and serves as a co-stimulatory signal. We hypothesize co-engineering CAR-T cells with a secreted CD70 construct that mediates enhanced CAR-T cell expansion, memory and long-term persistence can improve outcomes of GPC3 CAR-T cell therapy.

Methods A construct encoding the selected anti-GPC3 scFv CAR and an armored CD70 was generated and the effector functions of the engineered CAR-T cells were evaluated in vitro with different liver cancer cell lines. The antitumor efficacies were evaluated in vivo against multiple xenograft models.

Results To determine the impact of CD70 armor construct on CAR-T cell function in response to target tumor cells, we stimulated CD70 armored CAR-T and unarmored CAR-T with irradiated liver cancer cell line Hep3B and HepG2 in vitro, the CD70-GPC3 CAR-T cells promoted durable cytotoxicity in a low E:T ratio and persisted significantly better compared to CAR-T cells without armoring construct. in vivo, treatment with CD70-GPC3 CAR-T cells significantly controlled GPC3 positive Hep3B subcutaneous tumors in mice and enhanced their overall survival compared to unarmored CAR-T cells. Furthermore, CD70-GPC3 CAR-T had superior in vivo T cell expansion in peripheral blood and spleen with a higher CD4+ T cell population in contrast to unarmored CAR day 125 post treatment. To compare the efficacy and toxicity of CD70-GPC3 CAR-T to other clinical-testing armor designs in vivo, a GPC3 low expression xenogeneic mice model was used. All armored CAR-T treatments mediate potent antitumor activity with a low dose. However, only CD70-GPC3 CAR-T treated mice maintained a good safety profile for a long-term observation. Other armored CAR-T treated mice experienced different level body weight loss and off-target toxicity.

Conclusions Different armored GPC3 CAR-T therapies have demonstrated early clinical advances in treating heavily pretreated HCC patients. However, potential challenges with respect to dose-limited toxicities exist. We believe our novel armor design could addresse safety issues without compromising anti-tumor efficacy. CD70 armored GPC3 CAR-T will have the potential to improve clinical outcome and provide a broader therapeutic window as compared to other GPC3 CAR-T therapies currently in clinical trials.