Bioinformatic analysis of cancer-associated fibroblast related gene signature as a predictive model in clinical outcomes and immune characteristics of gastric cancer

BACKGROUND: Gastric cancer (GC) has a high incidence and high mortality rate among Asian countries, and distinguishing predictive prognosis biomarkers for GC are essential. Cancer-associated fibroblasts (CAFs) play a significant role in the progression, immune evasion, and therapeutic resistance of GC. Therefore, CAF-associated genes might have huge potential as prognostic biomarkers for predicting tumor progression and survival rate in GC pateints. METHODS: A sum of 1,134 GC patients from the The Cancer Genome Atlas Stomach Adenocarcinoma (TCGA-STAD), GSE62254, and GSE84437 datasets as well as GC cohorts from Xijing hospital were included. Firstly, we performed univariate Cox regression analysis to identify CAF-associated prognostic genes. Subsequently, the Least Absolute Shrinkage and Selection Operator (LASSO) regression analysis was used to develop a CAF gene signature (CAFGS) in the TCGA-STAD training cohort. CAFGS’s predictive performance was examined in both the training and validation cohorts, and the relationship between CAFGS and the tumor microenvironment (TME) was investigated by ssGSEA, CIBERSORT, TIMER, and ESTIMATE. Finally, a nomogram of CAFGS was established. RESULTS: Ten CAF-associated genes (ANGPTL4, CPNE8, CST2, HTR1F, IL1RAP, NR1D1, NTAN1, OLFML2B, TMEM259, and VTN) were identified to develop CAFGS. A high CAFGS score represented a worse outcome for GC patients in four cohorts, and a strong correlation was found between CAFGS and the infiltration of immune cells. We showed that CAFs contribute to immune evasion and unfavorable prognoses of GC patients by promoting the formation of an immunosuppressive microenvironment, and a high level of CAF infiltration may attenuate the efficacy of immunotherapy. The nomogram based on CAFGS showed reasonable predictive ability and may deliver great clinical net benefits. CONCLUSIONS: We established a CAFGS model with 10 CAF-associated genes that had a great predictive value for GC prognosis and survival rate evaluation. This study could provide a novel insight for investigating the role of CAFs in GC.