Angiogenesis-Related Gene Expression Signatures Predicting Prognosis in Gastric Cancer Patients

SIMPLE SUMMARY: To elucidate the role of angiogenesis as a prognostic signature in gastric cancer, we analyzed the expression level of 36 angiogenesis-related genes (ARGs) from Stomach Adenocarcinoma (STAD) from The Cancer Genome Atlas (TCGA). Consensus clustering analysis showed two major angiogenesis-related types: one related to more aggressive clinicopathological characteristics and worse survival, and the other related to lower tumor, lymph node, metastasis (TNM) stage and better outcomes. Our analysis of TCGA with a least absolute shrinkage and selection operator (LASSO) regression model identified 10 genes associated with overall survival in gastric cancer patients. With this gene signature, we computed angiogenesis-related gene signature risk scores for individual cancer patients that predicted overall and disease-free survival, which were further validated in the independent dataset Asian Cancer Research Group (ACRG). Moreover, an overall survival (OS)-related nomogram was established and had better performance in prognosis prediction than TNM stage. Our analysis provides a comprehensive map of ARGs that can be serve as useful biomarkers for gastric cancer. ABSTRACT: Increasing evidence indicates that angiogenesis is crucial in the development and progression of gastric cancer (GC). This study aimed to develop a prognostic relevant angiogenesis-related gene (ARG) signature and a nomogram. The expression profile of the 36 ARGs and clinical information of 372 GC patients were extracted from The Cancer Genome Atlas (TCGA). Consensus clustering was applied to divide patients into clusters 1 and 2. Least absolute shrinkage and selection operator (LASSO) Cox regression analyses were used to identify the survival related ARGs and establish prognostic gene signatures, respectively. The Asian Cancer Research Group (ACRG) (n = 300) was used for external validation. Risk score of ARG signatures was calculated, and a prognostic nomogram was developed. Gene set enrichment analysis of the ARG model risk score was performed. Cluster 2 patients had more advanced clinical stage and shorter survival rates. ARG signatures carried prognostic relevance in both cohorts. Moreover, ARG-risk score was proved as an independent prognostic factor. The predictive value of the nomogram incorporating the risk score and clinicopathological features was superior to tumor, lymph node, metastasis (TNM) staging. The high-risk score group was associated with several cancer and metastasis-related pathways. The present study suggests that ARG-based nomogram could serve as effective prognostic biomarkers and allow a more precise risk stratification.