Application Of Related Genes Of Cancer Associated Fibroblast In Predicting Prognosis And Treatment Of Patients With High-grade Serous Ovarian Cancer

Background:High-grade serous ovarian cancer(HGSOC)was characterized by rapid progress,extensive disease expansion and low survival rate.It was urgent to find biomarkers related to the prognosis of HGSOC,and to carry out disease risk stratification and individualized management for HGSOC patients.Previous studies have shown that cancer associated fibroblast(CAF),was significantly associated with the poor prognosis of various solid tumors.This study aimed to explore the relationship between CAF and the prognosis of HGSOC,identify CAF related biomarkers and tried to develop a CAF prognostic model for disease risk stratification,as well as better select individualized treatment for patients.Methods:The transcriptome information and corresponding clinical information of 375 HGSOC patients were downloaded from the Cancer Genome Atlas(TCGA)database.MCP-counter deconvolution algorithm was used to evaluate the CAF infiltration level,and ESTIMATE algorithm was used to assess the stromal score of tumor microenvironment CAF related genes were screened,and the CAF prognosis model was constructed by weighted gene co-expression network analysis(WGCNA)and least absolute contraction and selection operator(LASSO)algorithm.Gene Onotology Consortium(GO)and Kyoto Encyclopedia of Genes and Genomes(KEGG)enrichment analysis were used to elucidate its potential molecular mechanism.Pearson test was used to analyze the correlation between CAF infiltration level and CAF related genes and CAF risk score.Tumor Immune Dysfunction and Rejection(TIDE)algorithm was used to evaluate the immune invasion potential and to predict the response of immunotherapy.Semi-inhibitory concentration(IC50)was used to evaluate the sensitivity of HGSOC patients to common chemotherapy drugs.Results:Compared with the low CAF infiltration group,the high CAF infiltration group showed significantly poor clinical prognosis(p=0.023).Six CAF related genes,included TGFBI,TIMP3,COL16A1,FSTL3,SORCS2,and RASSF2,were screened to establish the CAF prognosis model.The predictive effect of this model was verified in the GSE49997 dataset.Pearson correlation analysis showed that CAF risk score was highly positively correlated with CAF infiltration level and 6 related genes.Univariate and multivariate Cox analysis showed that CAF risk score was a risk factor for the prognosis of HGSOC patients,affecting the prognosis along with the age.IC50 analysis showed that patients lower CAF risk score were more sensitive to paclitaxel,cisplatin.In addition,TIDE analysis showed that patients with high CAF risk score had higher immune escape potential,while patients with low CAF risk scores were more likely to benefit from immunotherapy.Conclusion:CAF played an unfavorable role in the prognosis of HGSOC disease.The CAF prognosis model constructed by six CAF related genes TGFBI,TIMP3,COL16A1,FSTL3,SORCS2 and RASSF2 had certain prediction value.This model could not only be used to stratify the prognosis risk of HGSOC patients,but also to a certain extent to provide a plan for individualized treatment of patients,including common chemotherapy and immunotherapy.