Colorectal cancer (CRC) is the third cause of cancer-related mortality in industrialized countries. Local invasion and metastasis formation are events associated with poor prognosis for which today there are no effective therapeutic options. Invasion and metastasis are strongly modulated by cells of the tumor microenvironment (TME), in particular fibroblasts and endothelial cells. Unraveling interactions between tumor cells and cells of the TME may identify novel mechanisms and therapeutic targets to prevent or treat metastasis. We report here the development and in vivo validation of a 3D tumor spheroid model to study the interactions between CRC cells, fibroblasts and endothelial cells in vitro. Co-cultured fibroblasts promoted SW620 and HCT116 CRC spheroid invasion, and this was prevented by the SRC and FGFR kinase inhibitors Dasatinib and Erdafitinib, respectively. To validate these findings in vivo, we injected SW620 cells alone or together with fibroblasts orthotopically in the caecum of mice. Co-injection with fibroblasts promoted lung metastasis growth, which was fully reversed by treatment with Dasatinib or Erdafitinib. Co-culture of SW620 or HCT116 CRC spheroids with endothelial cells suppressed spheroid growth while it had no effect on cancer cell migration or invasion. Consistent with this in vitro effect, co-injected endothelial cells significantly inhibited primary tumor growth in vivo. From these experiments we conclude that effects on cancer cell invasion and growth induced by co-cultured TME cells and drug treatment in the 3D spheroid model in vitro, are predictive of in vivo effects. The 3D spheroid model may be considered as an attractive model to study the effect of heterotypic cellular interactions and drug activities on cancer cells, as animal testing alternative. This model may be adapted and further developed to include different types of cancer and host cells and to investigate additional functions and drugs.

The complex interaction between tumor-associated macrophages (TAMs) and tumor cells through soluble factors provides essential cues for breast cancer progression. TAMs-targeted therapies have shown promising clinical therapeutical potential against cancer progression. The molecular mechanisms underlying the response to TAMs-targeted therapies depends on complex dynamics of immune cross-talk and its understanding is still incomplete. In vitro models are helpful to decipher complex responses to combined immunotherapies. In this study, we established and characterized a 3D human macrophage-ER+ PR+ HER2+ breast cancer model, referred to as macrophage-tumor spheroid (MTS). Macrophages integrated within the MTS had a mixed M2/M1 phenotype, abrogated the anti-proliferative effect of trastuzumab on tumor cells, and responded to IFNγ with increased M1-like polarization. The targeted treatment of MTS with a combined CSF1R kinase inhibitor and an activating anti-CD40 antibody increased M2 over M1 phenotype (CD163+/CD86+ and CD206+/CD86+ ratio) in time, abrogated G2/M cell cycle phase transition of cancer cells, promoted the secretion of TNF-α and reduced cancer cell viability. In comparison, combined treatment in a 2D macrophage-cancer cell co-culture model reduced M2 over M1 phenotype and decreased cancer cell viability. Our work shows that this MTS model is responsive to TAMs-targeted therapies, and may be used to study the response of ER+ PR+ HER2+ breast cancer lines to novel TAM-targeting therapies.