EGFR deletion in myeloid cells reprograms the immunosuppressive landscape of colorectal cancer.

Despite advances in the treatment of metastatic colorectal cancer (mCRC), it remains the second leading cause of cancer-related mortality, with limited effective therapeutic options. While EGFR inhibition is a standard first-line therapy for mCRC patients lacking KRAS mutations, resistance frequently develops, limiting its clinical benefit. Murine CRC models have shown that EGFR deletion in myeloid cells reduces tumor burden, and the presence of EGFR-positive myeloid cells is associated with poor prognosis in mCRC patients. However, the role of these cells in the therapeutic response to anti-EGFR therapies remains poorly understood. In this study, we integrated mouse models, single-cell RNA sequencing (scRNAseq), proteomics, and patient-derived mCRC datasets to investigate how EGFR signaling in myeloid cells shapes the tumor microenvironment (TME). In a preclinical therapeutic trial, we demonstrate that EGFR deletion in myeloid cells of tumor-bearing CRC mice reduced tumor growth, whereas EGFR loss in intestinal epithelial tumor cells alone had no therapeutic impact. EGFR deletion also decreased the abundance of F4/80hi macrophages, particularly the Spp1+ and C1qc+ subsets, and reduced inflammatory pathways like TGFβ, IFNγ, and JAK/STAT signaling in myeloid cells. These changes altered myeloid-T cell interactions, resulting in a less immunosuppressive TME characterized by reduced immune checkpoint expression. Furthermore, we identified thrombospondin-1 (THBS1) as a myeloid-derived ligand interacting with T cells, and confirmed its regulation by EGFR signaling through proteomic analysis. Analysis of human CRC datasets revealed that high EGFR and THBS1 expression correlates with poor patient outcomes. Collectively, these findings establish EGFR signaling in myeloid cells as a critical regulator of the immunosuppressive TME and suggest that targeting EGFR within specific myeloid subsets could represent a promising therapeutic strategy in mCRC.