AbstractSalivary duct carcinoma (SDC) is a rare and aggressive salivary gland cancer subtype with poor prognosis. The mutational landscape of SDC has been described rather exhaustively; yet, with respect to functional genomics and tumor microenvironment, little is known. In this study, transcriptomics and proteomics were combined to obtain the first characterization of the pathways deregulated in SDC. The data revealed the importance of Notch, TGB-β, and interferon-γ signaling. After associating computational biology, immunohistochemistry, multiplexed immunofluorescence, and digital imaging the first steps towards charting the cellular network within the microenvironment was initiated. According to immune infiltrate, two well-defined groups of tumors were observed, novel SDC immune checkpoints were discovered, and the key role played by macrophages and potentially NK cells in immunosuppression was shown. Furthermore, a clear trend between recurrence-free survival and M2 macrophage abundance was apparent. Independently, a measure of desmoplastic stromal reaction as determined by α-SMA abundance, was also shown. Altogether, these many findings open new perspectives for understanding and treating SDC. Before applying an immunotherapy, classical patient stratification according to immune infiltrate should be taken into account. Moreover, the microenvironment offers new potential targets including macrophages or NK cells, or even fibroblasts or hyaluronic acid. Related therapies that have been developed against, e.g., pancreatic tumors could inspire equivalent therapy for SDC.Additional informationFinancial support: MA (1 grant, GIRCI SOOM API-K 2016-811-DRC-AC), JC (2 grants, Fondation ARC PJA 20141201975, Labex EpiGenMed ANR 10-LABX-0012), AT (2 grants, Gunma University GIAR Research Program for Omics-Based Medical Science, Labex MabImprove ANR 10-LABX-0053 starting grant), ET (1 grant, SIRIC Montpellier Cancer Grant INCa_Inserm_DGOS_12553).No conflict of interest5408 words, 1 table, and 4 figuresStatement of translational relevanceBased on the presence or absence of an immune infiltrate, two groups of SDC were identified. These have the potential to provide a rationale for therapy and clinical trial enrolment. Two novel immune checkpoints that could be targeted were also identified; namely, CTLA-4/CD86 and TIM-3/galectin-9. Both showed the important contribution that macrophages and NK cells have in immunosuppression. Treatments that induce reprogramming or elimination of these cells could be considered. Moreover, the importance of the desmoplastic stroma was stressed. The stroma acts as a physical barrier against therapy suggesting that strategies developed against pancreatic tumors could inspire SDC treatments. For SDC devoid of immune infiltrate, components of the stroma including fibroblasts or hyaluronic acid could be targeted, e.g., in combination with drugs against immune checkpoints or mutated genes. Finally, evidence that Notch and TGF-β signaling are prevalent in SDC was obtained. This translates into additional therapeutic options.