Transcriptional heterogeneity of tumor-associated high endothelial venules defines inflammatory and stress-metabolic states with distinct prognostic associations.

Tumor-associated high endothelial venules (TA-HEVs) mediate lymphocyte trafficking into tumors and modulate the tumor microenvironment, with reported effects on clinical outcomes. However, reports have described discordant associations across cancers and microenvironmental contexts. Studies on state-specific, pan-cancer analyses of TA-HEV function remain limited. We integrated publicly available single-cell RNA sequencing datasets from 11 cancer types. Functional features of TA-HEVs were inferred by pathway enrichment and single-cell gene-set scoring for pathway gene sets. State-specific programs were applied to The Cancer Genome Atlas dataset to assess their clinical impact. We constructed a comprehensive atlas of tumor-associated endothelial cells and identified TA-HEV subclusters. Five TA-HEV subclusters were grouped into two functional states: inflammatory and stress-metabolic. The inflammatory TA-HEVs were enriched for innate immune stimulation, cytokine/chemokine signaling, and MHC class II antigen presentation, whereas the stress-metabolic TA-HEVs were characterized by the unfolded protein response, heat shock pathways, oxidative phosphorylation, and ATP biosynthesis. Across cancers, the stress-metabolic TA-HEV state was generally associated with worse prognosis, while the inflammatory TA-HEV state showed context-dependent associations. Together, these findings define TA-HEVs as a heterogeneous endothelial population comprising distinct functional states with divergent clinical associations, providing a pan-cancer framework for interpreting TA-HEV signals in tumor biology.