M2-like tumor-associated macrophage-associated signatures from bulk and single-cell RNA-seq data may predict head and neck squamous cell carcinoma prognosis and immunotherapy response

Abstract

Abstract Purpose Tumor-associated macrophages (TAMs) are immunosuppressive and crucial in the invasion, development, and metastasis of head and neck squamous cell carcinoma (HNSCC). Despite the prognostic importance of TAMs in HNSCC, their immunological landscape remains unknown. This study used bulk and single-cell ribonucleic acid RNA sequencing (scRNA-seq) to assess TAM prognostic value and the immunological landscape and drug sensitivity of HNSCC. Methods The abundance of M1/M2 macrophages in the transcriptome was calculated using CIBERSORT software. M2-like TAM-related genes were identified by integrating M2-like TAM marker genes from scRNA-seq data and M2 modularity genes from Weighted Correlation Network Analysis (WGCNA) of bulk data. A seven-gene (plasminogen activator, urokinase (PLAU), actinin alpha 1 (ACTN1), thioredoxin (TXN), integrin subunit alpha 5 (ITGA5), solute carrier family 2 member 1 (SLC2A), prolyl 4-hydroxylase subunit alpha 1 (P4HA1) and transforming growth factor beta-1 (TGFB1)) signature model was developed using least absolute shrinkage and selection operator (LASSO) regression analysis and univariate Cox regression. Immune cells, immunological function, and immune escape scores were used to assess the immune landscape of HNSCC patients. Results M2-like TAMs correlated with a poor prognosis in HNSCC patients. The risk score was validated as a separate prognostic factor with strong accuracy. We predicted risk group based anti-cancer drugs and selected nine drugs with significant sensitivity in the high-risk category. Conclusion We created an M2-like TAM-related gene set with good performance in predicting patient prognoses and directing therapy modalities. This gene set can potentially be used to personalize treatment for HNSCC patients and improve clinical outcomes.