Extramedullary disease (EMD) in multiple myeloma (MM) is associated with poor outcomes driven by aggressive kinetics and resistance to therapy. While the bone marrow tumor microenvironment (TME) is recognized to support survival and drug resistance in marrow-confined MM, the tissue architecture and cellular interactions that define EMD sites have been less well characterized. The study applied spatial transcriptomics to tumor biopsies from EMD to define cellular composition, spatial organization, and intercellular signaling within these lesions.
The investigators used spatial transcriptomics on EMD tumor biopsies to capture gene expression with preserved tissue context. This approach allowed mapping of tumor cells and nonmalignant TME components across tissue sections, enabling identification of spatially distinct niches and modeling of cell–cell interactions based on co-located transcriptional programs. Details of patient numbers, sample preparation, sequencing depth, and computational pipelines were not reported in the abstract.
Across samples, tumor plasma cells (PCs) maintained a characteristic PC transcriptome but displayed substantial interpatient heterogeneity in expression profiles, a pattern similar to marrow-restricted MM. This finding indicates that, although EMD PCs retain core plasma cell identity, there is variability between patients in tumor cell transcriptional states. The spatial approach preserved the localization of these varied PC phenotypes within their native tissue microenvironments.
The EMD TME was dominated by macrophages with a suppressive M2-type phenotype. CD8+ T cells were present in many tumors but frequently expressed exhaustion markers LAG3 and TIGIT, consistent with a functionally impaired cytotoxic compartment in situ. Tumor-associated fibroblasts were abundant and associated with a fibrotic, suppressive extracellular matrix (ECM) in certain spatial niches. Together, these components depict an immunosuppressive milieu in EMD characterized by macrophage dominance, T-cell dysfunction, and stromal remodeling.
Three recurrent TME niches were defined based on immune cell composition and spatial relationships:
Immune-excluded: This niche comprised the bulk of tumor regions and was characterized by spatial exclusion of immune cells from tumor-dense zones. Immune cells, including CD8+ T cells, were spatially separated from tumor PCs.
Immune-suppressed: Regions in this niche contained abundant tumor-associated fibroblasts producing a suppressive ECM. The fibrotic environment in these areas is proposed to further contribute to T-cell exclusion and dysfunction.
Immune-permissive: A minority of regions exhibited a microenvironment more permissive to immune infiltration and activity.
The predominance of immune-excluded niches suggests that spatial barriers and local suppressive cell populations are central features of EMD lesions.
Modeling of putative cell–cell interactions revealed a complex, bidirectional network between PCs and TME cells. Tumor PCs were predicted to modulate the microenvironment through secreted mediators such as PGE2 and VEGFB. In turn, PCs were predicted to receive canonical prosurvival signals through pathways involving CD38, CXCR4, and BCMA. Notably, TME cells—particularly fibroblasts and macrophages—were predicted to act collectively to promote the suppressive macrophage phenotype and to deposit a fibrotic ECM, reinforcing an immunosuppressive niche that supports local disease persistence and progression.
The spatially resolved data support a model in which EMD adapts to, or substitutes into, local tissue niches rather than existing as entirely niche-independent tumors. The identification of dominant suppressive elements—M2 macrophages, exhausted CD8+ T cells, and ECM-producing fibroblasts—highlights microenvironmental contributors to immune evasion and therapy resistance in EMD. The predicted signaling axes (for example, tumor-derived PGE2 and VEGFB, and prosurvival inputs via CD38, CXCR4, and BCMA) point to potentially targetable interactions between PCs and the TME. The study identifies clinically tractable microenvironmental niches and signaling networks that could be considered in the design of therapeutic strategies for this high-risk myeloma subset.
Spatial transcriptomics of extramedullary myeloma biopsies reveals a predominantly immunosuppressive, T-cell–excluded tumor microenvironment dominated by M2-like macrophages and abundant fibroblasts producing a fibrotic ECM. Three recurrent niches—immune-excluded, immune-suppressed, and immune-permissive—were identified, with immune exclusion being the most common. Bidirectional signaling between plasma cells and TME components was predicted, implicating mediators such as PGE2, VEGFB, and canonical prosurvival receptors (CD38, CXCR4, BCMA) as contributors to local tumor survival. These spatial insights support niche substitution as a model of EMD behavior and nominate microenvironmental features and signaling pathways that may be amenable to therapeutic intervention.
Note: The abstract summarizes these key findings; the original article should be consulted for full methods, cohort details, statistical analyses, and complete data supporting the reported results.