Liver metastasis is the leading cause of mortality in colorectal cancer (CRC), yet the phenotypic diversity and spatial organization of myeloid cells in this setting remain unclear. Publicly available single-cell RNA sequencing, spatial transcriptomic, and bulk transcriptomic datasets from CRC and liver metastasis were integrated. Single-cell analyses were performed to characterize myeloid cell heterogeneity, differentiation trajectories, and inferred intercellular communication, while spatial transcriptomic analyses were conducted to assess cell-type enrichment and inferred spatial relationships. Differential expression and survival analyses were combined to construct an HLA-DRB5⁺ macrophage-associated prognostic signature, which was validated across multiple independent cohorts. Eleven myeloid cell subsets were identified, among which SPP1⁺ and HLA-DRB5⁺ macrophages were markedly enriched in liver metastases and occupied distinct positions along myeloid differentiation trajectories. Cell–cell communication analysis revealed distinct predicted signaling patterns, with SPP1⁺ macrophages interacting with B cells and regulatory T cells through MIF–(CD74 + CXCR4) signaling, and HLA-DRB5⁺ macrophages communicating with T and NK cells via LGALS9–CD45 signaling. Spatial transcriptomic analyses demonstrated distinct localization patterns of these macrophage subsets across tumor, stromal, and hepatocyte regions, and MISTy-based analysis identified dynamic spatial associations with immune and stromal populations. In addition, a six-gene prognostic signature derived from HLA-DRB5⁺ macrophage-associated genes exhibited robust and independent predictive performance for overall survival across multiple CRC cohorts. This integrative multi-omics analysis reveals distinct transcriptional and spatial features of SPP1⁺ and HLA-DRB5⁺ macrophages in CRC liver metastasis and highlights their potential roles in shaping the metastatic immune microenvironment.
Xie et al. (2026) studied this question.