Abstract
Although immunotherapy has recently demonstrated a substantial promise in treating advanced thyroid carcinoma (THCA), it is not appropriate for all THCA patients. As a result, this study aims to identify biomarkers for predicting immunotherapy efficacy and prognosis in THCA patients based on a constructed prognostic model. The transcriptomic and corresponding clinical data of THCA patients were obtained from the Cancer Genome Atlas (TCGA) database. We identified differentially expressed genes (DEGs) between THCA and normal samples and performed an intersection analysis of DEGs with immune-related genes (IRGs) downloaded from the ImmPort database. Functional enrichment analysis was performed on the chosen immune-related DEGs. Subsequently, Cox and LASSO regression analyses were conducted to obtain three hub immune-related DEGs, including PPBP, SEMA6B, and GCGR. Following that, a prognostic risk model was established and validated based on PPBP, SEMA6B, and GCGR genes to predict immunotherapy efficacy and THCA prognosis. Finally, we investigated the association between the constructed risk model and tumor mutational burden (TMB), abundance of tumor-infiltrating immune cells (TICs) as well as immunotherapeutic targets (PDL-1, PD-1, and CTLA4) in THCA. THCA patients in the high-risk score (RS) group showed higher TMB levels and worse prognosis than the low RS group. Patients in the high-RS group had higher proportions of monocytes, M2 macrophages, and activated dendritic cells, whereas those in the low-RS group exhibited higher numbers of M1 macrophages and dendritic resting cells. Our data implied that the constructed THCA prognostic model was sound and we concluded that the THCA patients having high TMB and low PD-L1 expression levels might respond poorly to immunotherapy. Taken together, we constructed a novel prognostic model for THCA patients to predict their prognosis and immunotherapy efficacy, providing a viable option for the future management of THCA patients in the clinic.