Molecular subtyping and prognostic risk characterization of head and neck squamous cell carcinoma based on lysosome-related genes

Abstract

Lysosomes possess a multitude of biological functions and are known to play a crucial role in the proliferation and metastasis of head and neck squamous cell carcinoma (HNSCC). This study aims to systematically investigate the potential role of lysosomes-related genes (LRGs) in the development of heterogeneity and prognosis in HNSCC. Publicly available transcriptome and clinical data of HNSCC were obtained and analyzed using consensus clustering to identify molecular subtypes. A risk model based on LRGs was developed and evaluated, including its correlation with clinical features, immune infiltration, drug sensitivity, and response to immune therapy. Gene set enrichment analysis was conducted to explore relevant pathways, and a prognostic nomogram model for HNSCC was constructed and evaluated. In this study, we identified 542 LRGs that exhibited differential expression in HNSCC, with 116 of these being significantly associated with overall survival. Two LRGs-derived molecular subtypes were identified, which displayed significant differences in prognosis and immune cell infiltration. Additionally, a prognostic risk model was developed, which included 13 LRGs. This model successfully divided HNSCC into low-risk and high-risk groups with different prognoses and immune cell infiltrations. The LRGs-derived risk signature was associated with immune infiltration, clinical features, drug sensitivity and immunotherapy response. The good prognosis of the low-risk group was linked to the activation of immune response-related processes and the inhibition of pathways such as necroptosis and neutrophil extracellular trap formation. Patients in the low-risk group had better immune therapy response, while those in the high-risk group had higher drug sensitivity. Finally, our nomogram, which combines clinical N staging and LRG-derived model, demonstrated excellent prognostic evaluation performance as shown by decision curve analysis and calibration curve. The study provides a comprehensive analysis of the expression and prognostic significance of LRGs in HNSCC, leading to the identification of 2 distinct molecular subtypes and the development of a risk model based on LRGs.