Derivation, Comprehensive Analysis of Cancer-associated Fibroblasts and Their Correlation to Prognosis, Immune Status and Drug Sensitivity in Patients of Soft Tissue Sarcoma

Background: Soft tissue sarcoma is among the most challenging malignancies for clinicians and associated with poor prognosis. Former studies have revealed that cancer-associated fibroblasts (CAFs) are involved in tumor development, migration, invasion and metastasis.Methods: DEGs and WGCNA were employed to obtain significant CAFs-related genes for constructing the signature. Somatic mutation, immune status, immune checkpoints and chemotherapy sensitivity were applied to define the characteristics of the identified signature. Differentially expressed genes between the distinct groups and WGCNA were employed to obtain CAFs score of each patient. Results: 35 genes were identified as CAFs-related genes in soft tissue sarcoma samples. Then, 12 CAFs-related genes were associated with overall survival in univariate Cox regression analysis and K-M method. The five genes signature was constructed based on the result from the LASSO regression analysis in train cohort. K-M curve showed that patients in the low CAFs score group had better survival, with immune infiltration and function significantly increased (P<0.05). The 1-, 3- and 5- year survival AUC values were 0.773, 0.724, and 0.726, respectively. The analysis of tumor mutations revealed that TP53, ATRX, TTN, RB1 and MUC16 were enriched in both clusters. The analysis of chemotherapy response demonstrated that most of chemotherapeutic drugs in GDSC database had lower IC50 in the high-risk group. All of the results were validated in test and whole cohorts. Conclusion: This study strongly revealed the indispensable role of CAFs in soft tissue sarcoma and constructed the prognostic signature including five CAFs-related genes.