Targeting Suppressive Myeloid Cells Potentiates Checkpoint Inhibitors to Control Spontaneous Neuroblastoma

Purpose: Neuroblastoma is the most common extracranial solid cancer type in childhood, and high-risk patients have poor prognosis despite aggressive multimodal treatment. Neuroblastoma-driven inflammation contributes to the induction of suppressive myeloid cells that hamper efficient antitumor immune responses. Therefore, we sought to enhance antitumor immunity by removing immunosuppression mediated by myeloid cells. Experimental Design: The prognostic values of myeloid cells are demonstrated by analyzing genomic datasets of neuroblastoma patients. The impact of tumor-derived factors on myelopoiesis and local induction of suppressive myeloid cells is dissected by in vitro culture models using freshly isolated human CD34+ hematopoietic stem cells, primary human monocytes, and murine bone marrow cells. To test the therapeutic efficacy of BLZ945 as a monotherapy or in combination with checkpoint inhibitors, we used a transgenic murine model (TH-MYCN) that develops aggressive spontaneous neuroblastoma. Results: We report that infiltrating CSF-1R+ myeloid cells predict poor clinical outcome in patients with neuroblastoma. In vitro, neuroblastoma-derived factors interfere with early development of myeloid cells and enable suppressive functions on human monocytes through M-CSF/CSF-1R interaction. In a transgenic mouse model (TH-MYCN) resembling high-risk human neuroblastoma, antagonizing CSF-1R with a selective inhibitor (BLZ945) modulates the induction of human and murine suppressive myeloid cells and efficiently limit tumor progression. While checkpoint inhibitors are insufficient in controlling tumor growth, combining BLZ945 with PD-1/PD-L1 blocking antibodies results in superior tumor control. Conclusions: Our results demonstrate the essential role of CSF-1R signaling during the induction of suppressive myeloid cells and emphasize its clinical potential as an immunotherapy for human cancers. Clin Cancer Res; 22(15); 3849–59. ©2016 AACR.