The composition of the tumor microenvironment modulates cancer progression. This microenvironment possesses cellular and physicochemical features that highly influence each other and that determine the tumor malignancy. During this thesis, we studied the influence of cycling hypoxia, which is a physicochemical characteristic of most solid tumors, on the phenotype of two cell types present in the tumor microenvironment namely endothelial cells and macrophages. We particularly investigated their pro-inflammatory phenotypes. Indeed, on one hand, tumor-promoting inflammation is now well recognized as an enabling characteristic for cancer progression. On the other hand, cycling hypoxia, which refers to a phenomenon of hypoxia/reoxygenation cycles caused by an irregular erythrocyte flux within the tumor blood network, is more and more described to favor angiogenesis, metastasis and resistance to anti- cancer treatments. As cycling hypoxia and inflammation are two processes that support cancer progression, we sought to examine the role of cycling hypoxia on tumor-promoting inflammation. We assessed the impact of cycling hypoxia on the inflammatory phenotype of endothelial cells stimulated or not with TNFα in vitro. Results showed that cycling hypoxia enhanced the endothelial inflammatory response induced by TNFα, characterized by an increased expression of two pro-inflammatory cytokines IL-6 and IL-8; by an increased expression of adhesion molecules in particular ICAM-1 protein expression and VCAM-1 and E-selectin mRNA expression; and consequently by an increase in THP-1 monocyte adhesion. The increased expression of these inflammatory proteins by endothelial cells has been shown to occur through an overactivation of NF-κB. Thereafter, we examined the influence of cycling hypoxia on the overall tumor inflammation in vivo in LLc tumor-bearing mice. Results showed that cycling hypoxia, which was artificially amplified, led to an increase in the whole tumor mRNA expression of PTGS2, IL-6, KC and MIP-2 (two murine functional homologs of IL-8) as well as to an enhanced leukocyte infiltration in tumors, meaning that cycling hypoxia amplified the overall tumor inflammation. Furthermore, we investigated the potential use of the high expression of this inflammatory signature combined to the low expression of BIRC5, which is a cycling hypoxia-specific downregulated gene, as a phenotypic biomarker of poor prognosis in human colon cancer. Results demonstrated that patients presenting the simultaneous gene expression phenotype BIRC5low/PTGS2high/ICAM- 1high/IL-6high/IL-8high in their tumors had a much lower probability of cancer survival than the other patients. This new and innovative phenotype could be used in the clinic as a prognostic marker in colon cancer. Moreover, this result means that the inflammation amplified at least in part by cycling hypoxia is pro-tumoral. Thereafter, we assessed the impact of cycling hypoxia on the inflammatory phenotype of human and murine macrophages that were unpolarized (M0) or polarized in M1 (LPS + IFNγ) or M2 (IL-4 + IL-13) state. Results showed that cycling hypoxia induced a pro-inflammatory phenotype in human and murine M0 macrophages and amplified the pro-inflammatory phenotype of human and murine M1 macrophages. This inflammatory phenotype is notably characterized by an increase in TNFα secretion. Altogether, these results demonstrated that cycling hypoxia participates in the amplification of a tumor-promoting inflammation by altering the inflammatory phenotype of endothelial cells and macrophages. However, cycling hypoxia probably also changes the phenotype of cancer cells and other cell types of the tumor microenvironment, as we demonstrated that cycling hypoxia increased the overall tumor inflammation in vivo. Moreover, as macrophages exposed to cycling hypoxia secrete more TNFα and as TNFα- stimulated endothelial cells exposed to cycling hypoxia favor a higher monocyte adhesion that is probably in part responsible for the increase in leukocyte intratumor infiltration, we suggest that cycling hypoxia initiates a positive feedback loop between endothelial cells and macrophages that could account for the amplification of tumor-promoting inflammation.