T Cell-associated Co-expression Module Genes as Prognostic Biomarkers and Predictors of Immunotherapy Response in...

The important roles of T cells in tumor progression support the development of a T cell-associated prognostic model for Head and Neck Squamous Cell Carcinoma (HNSCC).

The single-cell RNA-seq (scRNS-seq) data from GSE181919 were processed via the Seurat package for quality control and cell annotation. hdWGCNA was used to identify T cell-associated co-expression modules. Key genes were screened by univariate and LASSO Cox regression in TCGA to build a RiskScore model, followed by validation in the GSE41613 and GSE117973 datasets. The correlations of the RiskScore with the immune microenvironment, predicted immunotherapy response, and drug sensitivity were analyzed, and preliminary in vitro assays were conducted to explore PIM2 function in HNSCC cells.

Ten cell populations were identified in HNSCC tissues, with T cells representing one of the major cell populations. An eight-gene RiskScore model (TNFRSF4, TNFRSF18, PIM2, CORO1B, CUL9, SOD1, ZC3H12D, and TUBA1B) was established, showing moderate but significant prognostic value. High-risk patients exhibited significantly poorer survival. A nomogram integrating the RiskScore and clinical features was further constructed. Additional analyses supported the prognostic value of the model in HPV-negative patients and showed that the RiskScore was independently associated with overall survival after adjustment for tumor purity. C-index comparisons further revealed higher and relatively consistent prognostic discrimination compared with two previously published HNSCC models. The high-risk group exhibited lower ImmuneScore, StromalScore, ESTIMATEScore, higher TIDE scores, and reduced immune infiltration. The RiskScore was significantly correlated with 11 candidate compounds, and preliminary in vitro assays showed that PIM2 knockdown suppressed malignant phenotypes of HNSCC cells.

The scRNS-seq and hdWGCNA analyses identified eight T cell-associated module genes related to HNSCC prognosis. High- and low-risk patients stratified by the RiskScore exhibited distinct immune infiltration and angiogenesis in their Tumor Microenvironment (TME). Drug sensitivity analysis predicted candidate compounds for HNSCC treatment, but further experimental and clinical validation is required.

The T cell-associated prognostic model was associated with immune microenvironment features and predicted treatment response and may help predict HNSCC prognosis, providing candidate biomarkers for further investigation.