The Pan-tumor Vasculature Under the Transcriptomic Magnifying Glass

Abstract In the first pan-cancer analysis of the tumor vasculature, Pan and colleagues profile nearly 200,000 endothelial and mural cells (ECs and MCs), identifying novel subclusters and cell states using consensus trajectory inference. They identify differentiation trajectories in vascular and lymphatic endothelial cells, and subtype the pericyte population. During sprouting angiogenesis, venous cells dedifferentiate and transition to capillary and, finally, arterial ECs. Capillary ECs transition via “three angiogenic stages” (SI-SIII) during which APLN+ TipS1 cells were identified as potential modulators of tumor-induced neovascularization and anti-angiogenic therapy response. In lymphatic ECs, differentiation was inversely correlated between the lymphangiogenic (T1) and antigen-presenting (T2) trajectories, with T2 associated with a better prognosis. While several pericyte clusters were identified, BASP1+ matrix-associated pericytes were associated with APLN+ TipS1 cells and had a worse prognosis. These findings present transcriptional validation of previous experimental findings and serve as a resource to examine the tumor vascular microenvironment in detail.