Spatial transcriptomics advances the use of canine patients in cancer research: analysis of osteosarcoma-bearing pet dogs enrolled in a clinical trial.

Abstract

PURPOSE Pet dogs spontaneously develop many of the same tumor types as humans including osteosarcoma. Recent advances in spatial transcriptomics have improved our ability to utilize formalin-fixed paraffin-embedded tissues collected at canine autopsy. These techniques allow the canine model to be investigated alongside murine models to inform human cancer research. Herein we present the first application of the GeoMx Canine Cancer Atlas to outcome-linked samples from canine patients enrolled in an osteosarcoma clinical trial. EXPERIMENTAL DESIGN A tissue microarray (TMA) of primary osteosarcoma samples was assayed using the GeoMx Digital Spatial Profiler. Samples were stratified by disease-free interval (DFI). Patients within the upper and lower tertiles were assigned to the High (n = 8) and Low DFI (n = 8) groups, respectively. Analyses included the identification of differentially expressed genes, pathway enrichment, and cell deconvolution. RESULTS Genes enriched in High DFI tumors included PTEN and CDKN1B. Low DFI tumors were enriched for NCAM1. Pathways enriched in the Low DFI group included MYC, MTORC1, and oxidative phosphorylation. High DFI tumors were enriched for interferon alpha response and allograft rejection pathways among others. Macrophages and CD8+ T cells were elevated in High DFI tumors. CONCLUSIONS Gene and pathway enrichment found to be associated with disease-free interval in this study overlap with that described in human osteosarcoma patients underscoring the value of the canine model in metastatic osteosarcoma research.