Molecular Subtype Identification and Prognosis Stratification Based on Golgi Apparatus-Related Genes in Head and Neck Squamous Cell Carcinoma

Abstract

Abstract Background: Abnormal dynamics of the Golgi apparatus reshape the tumor microenvironment and immune landscape, playing a crucial role in the prognosis and treatment response of cancer. This study aims to investigate the potential role of Golgi apparatus-related genes (GARGs) in the heterogeneity and prognosis of head and neck squamous cell carcinoma (HNSCC). Methods: The transcriptional data and corresponding clinical information of HNSCC were obtained from public databases. GARGs with differential expression were identified, and those associated with prognosis were identified through univariate Cox regression analysis. Consensus clustering was used to classify HNSCC into molecular subtypes. Lasso Cox regression analysis was used to screen and construct prognostic risk features for HNSCC, and receiver operating characteristic curves were used to evaluate the prognostic performance of the risk model. Differences in immune infiltration, clinical and pathological characteristics, and treatment response were compared between different subtypes and risk groups. Results: A total of 321 GARGs that were differentially expressed were identified, out of which 69 were associated with the prognosis of HNSCC. Based on these prognostic genes, two molecular subtypes of HNSCC were identified, which showed significant differences in prognosis. Additionally, a risk feature consisting of 28 genes was constructed for assessing the prognosis of HNSCC. This feature could divide HNSCC into high-risk and low-risk groups with significant differences in prognosis, and the ROC curve indicated excellent prognostic evaluation performance. This prognostic feature was related to survival outcome, grade, T stage, chemotherapy, and other features. Immune response-related pathways were significantly activated in the low-risk group with better prognosis. There were significant differences in chemotherapy drug sensitivity and immune therapy response between the high-risk and low-risk groups, with the low-risk group being more suitable for receiving immunotherapy. Riskscore, N stage, and radiotherapy were independent prognostic factors for HNSCC and were used to construct a nomogram, which had good clinical applicability. Conclusions: We have identified two molecular subtypes of HNSCC that are derived from GARGs. A GARGs-derived risgnature was also developed, which can be used in combination with other clinical features for the assessment of HNSCC prognosis.