Spatial architecture of tenascin-c–driven ECM–immune niches in sarcoma.

Abstract

e23562 Background: Tenascin-C (TNC) is an extracellular matrix glycoprotein highly expressed during embryogenesis but largely silent in adult tissues and re-activated in cancer and inflammation. We hypothesized that TNC shapes an immunosuppressive sarcoma microenvironment by organizing ECM–macrophage spatial relationships. Methods: We profiled a commercial soft-tissue sarcoma TMA (64 cases/192 FFPE cores) including liposarcoma, angiosarcoma, leiomyosarcoma, MPNST, and synovial sarcoma. Multiplex OPAL IF assessed TNC, Fibronectin, NRP1, CD68, and CD163. HALO analytics included segmentation, heatmaps, nearest-neighbor (NN) distances, proximity metrics, and neighborhood analysis. Heatmaps localized ECM/immune hotspots; area quantification measured TNC-ECM burden; proximity tested per-cell macrophage distances to TNC; NN captured shortest intercellular coupling. We also established 18 custom TMAs from 180 samples (90 first diagnosis and 90 paired metastasis/progression). A 37-plex COMET panel was optimized for these TMAs, covering ECM macrophage phenotypes, EMT/plasticity, stemness, angiogenesis, TLS/immune architecture, and proliferation. COMET image analysis and survival correlations are ongoing. The targeting peptide PL1 (binding to TNC/fibronectin) was tested in JBT-19 cells. JBT-19 xenografts were examined by multiplex staining. Results: In the commercial TMA, multiplex imaging revealed distinct TNC-rich ECM architectures across sarcoma types. HALO heatmaps demonstrated that TNC-dense ECM micro-domains frequently co-localized with macrophage-rich clusters. Spatial metrics showed preferential accumulation of CD68⁺ macrophages along TNC-positive boundaries, whereas CD163⁺ and NRP1⁺ TAMs were enriched within the core of TNC-rich niches. NN and proximity analyses revealed significantly shorter macrophage–ECM distances (~10–20 µm) in these regions, suggesting directed positioning rather than random infiltration. Neighborhood mapping highlighted stable CD163⁺/NRP1⁺ TAM communities associated with TNC-structured ECM. JBT-19 xenografts recapitulated strong TNC and Fibronectin expression. PL1 peptide bound robustly to JBT-19 cells. COMET staining across 18 TMAs is fully established; quantitative spatial metrics and clinical-outcome mapping will be finalized before the congress. Conclusions: TNC-structured, TAM-enriched ECM niches constitute an immune-exclusion architecture that likely impedes T-cell trafficking and cytotoxicity in sarcoma. This stromal configuration may underpin resistance to immune-checkpoint therapies and contribute to suboptimal responses to cell therapies such as TCR-T. Xenograft staining corroborates the presence of TNC-rich matrix domains in vivo, and robust PL1 binding in JBT-19 cells underscores their tractability as therapeutic targets. Final COMET-based spatial metrics and outcome correlations will be presented at the meeting.