Back

Immune repertoire profiling reveals extensive clonal expansion of T and B cells in CXCL9-rich niches in sinonasal cancer

Alvarez, J.; Gracia Villacampa, E.; Labuz, D.; Guryleva, M.; Urgard, E.; Kagedal, A.; Solmell, O.; Nasman, A.; Escudero Morlanes, J.; Ma, J.; Odum, N.; Achour, A.; Georen, S.; Chambers, B.; Murrell, B.; Cardell, L. O.; Engblom, C.; Thrane, K.; Coquet, J. M.

2026-06-08 immunology
10.64898/2026.06.04.730068 bioRxiv
Show abstract

Sinonasal cancer (SNC) is a rare and aggressive head and neck cancer with limited therapeutic options and an incompletely defined immune microenvironment. Although histopathology remains central to clinical tumor evaluation, immune profiling typically requires tissue dissociation and therefore loses the spatial relationships that shape antitumor immunity. Here, we combined single-cell RNA-sequencing approaches, spatial transcriptomics and Spatial V(D)J, a technique we recently pioneered, to analyse the cellular, spatial, and clonal architecture of SNC. We found that SNC tumors were marked by a robust T cell infiltration, including substantial proportions of classical tissue resident memory (TRM) and exhausted (TExh) CD8+ T cells. This inflamed phenotype was accompanied by an immense infiltration of suppressive regulatory T (Treg) cells, which differentiated towards a Tbet+CXCR3+ phenotype. Spatially, expanded Treg, CD8 T and B cell clones concentrated within specialized peritumoral immune niches enriched for cancer-associated fibroblasts, CXCL9 tumor-associated macrophages and LAMP3IDO1 dendritic cells. T cell clones occupying these niches were clonally related to those infiltrating epithelial tumor regions, linking lymphoid hubs to the broader tumor immune response. Together, these data identify SNC as an inflamed but highly immunoregulatory tumor type and reveal a spatially organized clonal architecture in which suppressive and cytotoxic lymphocyte states coexist within CXCL9 myeloid niches.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.