Abstract
BACKGROUND: Laryngeal squamous cell carcinoma (LSCC) is one of the world's most common head and neck cancer. However, the immune infiltration phenotypes of LSCC have not been well investigated. METHODS: The multi-omics data of LSCC were obtained from the TCGA (n=111) and GEO (n=57) datasets. The infiltrations of the 24 immune cell populations were calculated using the GSVA method. Then LSCC samples with different immune cell infiltrating patterns were clustered, and the multi-omics differences were investigated. RESULTS: Patients were clustered into the high-infiltration and low-infiltration groups. The infiltration scores of most immune cells were higher in the high-infiltration group. Patients with high-infiltration phenotype have high N and TNM stages but better survival, as well as less mutated COL11A1 and MUC17. Common targets of immunotherapies such as PD1, PDL1, LAG3, and CTLA4 were significantly up-regulated in the high-infiltration group. The differentially expressed genes were mainly enriched in several immune-related GOs and KEGG pathways. Based on the genes, miRNAs, and lncRNAs differentially expressed in both the TCGA and GEO cohorts, we built a ceRNA network, in which BTN3A1, CCR1, miR-149-5p, and so on, located at the center. A predictive model was also constructed to calculate a patient's immune infiltration phenotype using 16 genes' expression values, showing excellent accuracy and specificity in the TCGA and GEO cohorts. CONCLUSIONS: In this study, the immune infiltration phenotypes of LSCC and the corresponding multi-omics differences were explored. Our model might be valuable to predicting immunotherapy's outcome.