SIRPα blockade therapy potentiates immunotherapy by inhibiting PD-L1(+) myeloid cells in hepatocellular carcinoma.

Tumor-infiltrating myeloid cells (TIMs) are pivotal cell populations involved in the immunosuppressive tumor immune microenvironment (TIME). However, there has been little success in large-scale clinical trials of myeloid cell modulators. We aim to investigate potential molecular targets for TIMs and disclose the underlying mechanism. Using mass cytometry by time of flight (CyTOF), we analyzed 24 spontaneous HCC tissues from mouse. Orthotopic and subcutaneous tumor models were established with or without anti-SIRPα antibody treatment. Patient-derived tumor xenografts model (PDX) was used to identify the CD47-SIRPα axis blocked therapy. In 24 murine spontaneous HCC tissues, we observed that the proportion of myeloid-derived suppressor cells (MDSCs) plus macrophages accounts for 40-90% of TIMs and SIRPα was highly expressed in TIMs, especially in macrophages and MDSCs. Through in vivo experiments, we showed that anti-SIRPα therapy inhibited tumor growth, accompanied by increased CD8(+) T cells infiltration and decreased TIMs including MDSCs and macrophages. We found that anti-SIRPα inhibited immunosuppressive function, migration and PD-L1 expression of myeloid cells. In a series of in vivo experiments, we demonstrated the anti-tumor and immune-active effect of SIRPα-blocked therapy. Mechanistically, anti-SIRPα inhibited the immunosuppressive function and PD-L1 expression of TIMs through downregulating PI3K/AKT signaling in myeloid cells. At last, anti-SIRPα enhanced the antitumor effect of anti-PD-L1 therapy in orthotopic and spontaneous murine models. Together, SIRPα blocked therapy reversed the immunosuppressive TIME, which provides a promising therapeutic rationale for increasing the efficacy of anti-PD-L1 therapy in treating HCC.