The CSF-1R inhibitor pexidartinib affects FLT3-dependent DC differentiation and may antagonize durvalumab effect in patients with advanced cancers

Tumor-associated macrophages (TAMs) are a critical determinant of resistance to PD-1/PD-L1 blockade. This phase 1 study (MEDIPLEX, NCT02777710) investigated the safety and efficacy of pexidartinib, a CSF-1R–directed tyrosine kinase inhibitor (TKI), and durvalumab (anti–PD-L1) in patients with advanced colorectal and pancreatic carcinoma with the aim to enhance responses to PD-L1 blockade by eliminating CSF-1–dependent suppressive TAM. Forty-seven patients were enrolled. No unexpected toxicities were observed, one (2%) high microsatellite instability CRC patient had a partial response, and seven (15%) patients experienced stable disease as their best response. Increase of CSF-1 concentrations and decrease of CD14 low CD16 high monocytes in peripheral blood mononuclear cells (PBMCs) confirmed CSF-1R engagement. Treatment decreased blood dendritic cell (DC) subsets and impaired IFN-λ/IL-29 production by type 1 conventional DCs in ex vivo TLR3-stimulated PBMCs. Pexidartinib also targets c-KIT and FLT3, both key growth factor receptors of DC development and maturation. In patients, FLT3-L concentrations increased with pexidartinib treatment, and AKT phosphorylation induced by FLT3-L ex vivo stimulation was abrogated by pexidartinib in human blood DC subsets. In addition, pexidartinib impaired the FLT3-L– but not GM-CSF–dependent generation of DC subsets from murine bone marrow (BM) progenitors in vitro and decreased DC frequency in BM and tumor-draining lymph node in vivo. Our results demonstrate that pexidartinib, through the inhibition of FLT3 signaling, has a deleterious effect on DC differentiation, which may explain the limited antitumor clinical activity observed in this study. This work suggests that inhibition of FLT3 should be considered when combining TKIs with immune checkpoint inhibitors.