A novel mast cell marker gene-related prognostic signature to predict prognosis and reveal the immune landscape in head and neck squamous cell carcinoma.

BACKGROUND

Head and neck squamous cell carcinoma (HNSCC) is a highly aggressive and heterogeneous malignant tumor. Mast cells are one of the immune cells widely distributed in the tumor microenvironment (TME), and their immune response with various immune cells is essential in promoting or inhibiting tumor growth and metastasis. However, the role played by mast cells in HNSCC has yet to be fully clarified.

METHODS

We identified mast cell marker genes using single-cell RNA sequencing (scRNA-seq) from the GSE103322 of the GEO database. The HNSCC data from the TCGA databases was divided into training and validation groups. Cox regression and LASSO regression analyses were used to screen the prognostically relevant mast cell-related genes (MRGs) to construct a prognostic signature and differentiate risk groups. The receiver operating characteristic (ROC) and calibration curves were used to test the model's accuracy. We revealed the immune landscape of HNSCC by immune infiltration, immune checkpoint levels, ESTIMATE, and TIDE analyses. Drug sensitivity analyses were used to understand the sensitivity of different risk groups to drug therapy.

RESULT

The 14-MRGs prognostic signature classified patients into high- and low-risk groups, and the overall survival (OS) of the low-risk group was significantly higher than that of the high-risk group (p < 0.05). The areas under the ROC curves of the nomogram were 0.740, 0.737 and 0.707 at 1-, 3-, and 5-year, and they also showed better detection efficacy in the validation group than other independent predictors. The low-risk group had richer immune cell infiltration and higher immune scores. The lower TIDE score in the low-risk group demonstrates that patients in this group were less prone to have immune escape and more likely to benefit from immunotherapy. In addition, the low-risk group was more sensitive to a broader range of drugs than the high-risk group.

CONCLUSION

We combined scRNA-seq data and bulk RNA-seq data to construct a 14-MRGs-based prognostic model capable of well predicting the prognosis of HNSCC patients. This model may also help identify patients who can benefit from immunotherapy.