A vasculogenic mimicry prognostic signature associated with immune signature in human gastric cancer.

Background: Gastric cancer (GC) is one of the most lethal malignant tumors worldwide with poor outcomes. Vascular mimicry (VM) is an alternative blood supply to tumors that is independent of endothelial cells or angiogenesis. Previous studies have shown that VM was associated with poor prognosis in patients with GC, but the underlying mechanisms and the relationship between VM and immune infiltration of GC have not been well studied.

Methods: In this study, expression profiles from VM-related genes were retrieved from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. Cox regression was performed to identify key VM-related genes for survival. Subsequently, a novel risk score model in GC named VM index and a nomogram was constructed. In addition, the expression of one key VM-related gene (serpin family F member 1, SERPINF1) was validated in 33 GC tissues and 23 paracancer tissues using immunohistochemistry staining.

Results: Univariate and multivariate Cox regression suggested that SERPINF1 and tissue factor pathway inhibitor 2 (TFPI2) were independent risk factors for the prognosis of patients with GC. The AUC (> 0.7) indicated the satisfactory discriminative ability of the nomogram. SsGESA and ESTIMATE showed that higher expression of SERPINF1 and TFPI2 is associated with immune infiltration of GC. Immunohistochemistry staining confirmed that the expression of SERPINF1 protein was significantly higher in GC tissues than that in paracancer tissues.

Conclusion: A VM index and a nomogram were constructed and showed satisfactory predictive performance. In addition, VM was confirmed to be widely involved in immune infiltration, suggesting that VM could be a promising target in guiding immunotherapy. Taken together, we identified SERPINF1 and TFPI2 as immunologic and prognostic biomarkers related to VM in GC.