Single-cell RNA sequencing identifies cancer-associated fibroblast marker genes for determining cancer subtypes in oral squamous cell carcinoma and predicting patient prognosis

Background: Oral squamous cell carcinoma (OSCC) is an extremely aggressive cancer of which its progression and prognosis are strongly affected by cancer-associated fibroblast (CAF) and immunosuppression procedure. This study aimed to identify CAF marker genes through single-cell RNA sequencing (scRNA-seq) data and clarify the roles of CAF-related genes in tumor typing and disease prognosis prediction for OSCC patients. Methods: Gene expression data of OSCC patients were downloaded from The Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO) database. The differential expression of CAF genes was identified using the limma R package. Univariate Cox regression, least absolute shrinkage and selection operator (LASSO) Cox analysis, and multivariate Cox analysis were performed to construct a CAF-related prognostic signature. The accuracy of prediction performance was assessed by the time-dependent receiver operating characteristic (ROC) curve, Kaplan-Meier analysis, principal component analysis (PCA), decision curve analysis, calibration curve, and concordance index. Differences of tumor mutational burden, tumor-infiltrating immune cells, and response to immunotherapy between the high- or low-risk groups were investigated. Results: Our results showed that the CAF-high subtype exhibited poorer survival outcomes than the CAF-low subtype. The CAF-high subtype exhibited higher immune infiltration, stromal components, and immune escape potential than the CAF-low subtype. The CAF-high subtype OSCC patients demonstrated reduced sensitivity to immunotherapy than the CAF-low subtype. Conclusions: Our prognostic model could predict overall survival (OS) and the immunotherapy response of OSCC patients. These findings highlight the importance of targeting CAF in OSCC treatment and provide valuable insights into the tumor microenvironment (TME) heterogeneity and immune responses.