Comparative spatial transcriptomic profiling of canine and human lung metastatic osteosarcoma tumor microenvironments

Abstract Osteosarcoma (OS) is a rare pediatric cancer with poor outcomes due to metastases. These outcomes are mirrored in the spontaneous canine disease, offering a comparative opportunity to understand disease progression and tumor immunology. Studies have identified overlaps in molecular features between species, bolstering the utility of the canine model. With the introduction of cross-species-compatible spatial transcriptomics, there is unprecedented opportunity to dissect the metastatic OS tumor microenvironment through comparative oncology. This study identified conserved cellular signals in metastatic OS through independent and integrated analyses of canine (n = 7) and human (n = 5) GeoMx spatial transcriptomic datasets. Macrophage (CD163 + or CD68+), T cell (CD3+), and negative (CD163-/CD3- or CD68-/CD3-) tumor/stroma compartments were captured from intratumoral and peritumoral regions. Data were analyzed independently using GeomxWorkflows and integratively with Seurat, focusing on orthologous genes. We identified spatially dependent differences in tumor-associated T cell and macrophage infiltrates and conserved regional immune heterogeneity across canine and human metastatic OS, further strengthening the translational capabilities of the canine model. Shared features include upregulation of a JAG1-Notch signaling within macrophages, metabolic alterations in T cells, and increased MIF expression. These conserved pathways provide a foundation for investigating mechanisms contributing to metastatic OS progression and intratumoral immunosuppression.