BACKGROUND: Ferroptosis is an iron-dependent and regulated cell death that has been widely reported in a variety of malignancies. The overall survival of papillary thyroid cancer (PTC) is excellent, but the identification of patients with poor prognosis still faces challenges. Nevertheless, whether ferroptosis-related genes (FRGs) can be used to screen high-risk patients is not clear. METHODS: We obtained the clinical data of patients with PTC and FRGs from the UCSC Xena platform and the FerrDb respectively. Differentially expressed genes (DEGs) of FRGs were obtained from the entire The Cancer Genome Atlas (TCGA). Subsequently, the entire TCGA dataset was randomly split into two subsets: training and test datasets. Based on DEGs, we constructed a predictive model which was tested in the test dataset and the entire TCGA dataset to predict progression-free survival (PFS). Patients were categorized into high- or low-risk groups based on their median risk score. We analyzed differences in some aspects, including pathway enrichment analysis, single-sample Gene Set Enrichment Analysis (ssGSEA), tumor microenvironment (TME), human leukocyte antigen (HLA) genes, and tumor mutation burden (TMB) analyses, between high-risk and low-risk groups. RESULTS: A predictive model with three FRGs (HSPA5, AURKA, and TSC22D3) was constructed. Patients in the high-risk group had worse PFS compared with patients in the low-risk group. Functional analysis results revealed that ssGSEA, immune cell infiltration, TME, HLA, and TMB were closely associated with ferroptosis. CONCLUSION: The prognostic model constructed in this study can effectively predict PFS for patients with PTC.
BACKGROUND: Ferroptosis is an iron-dependent and regulated cell death that has been widely reported in a variety of malignancies. The overall survival of papillary thyroid cancer (PTC) is excellent, but the identification of patients with poor prognosis still faces challenges. Nevertheless, whether ferroptosis-related genes (FRGs) can be used to screen high-risk patients is not clear. METHODS: We obtained the clinical data of patients with PTC and FRGs from the UCSC Xena platform and the FerrDb respectively. Differentially expressed genes (DEGs) of FRGs were obtained from the entire The Cancer Genome Atlas (TCGA). Subsequently, the entire TCGA dataset was randomly split into two subsets: training and test datasets. Based on DEGs, we constructed a predictive model which was tested in the test dataset and the entire TCGA dataset to predict progression-free survival (PFS). Patients were categorized into high- or low-risk groups based on their median risk score. We analyzed differences in some aspects, including pathway enrichment analysis, single-sample Gene Set Enrichment Analysis (ssGSEA), tumor microenvironment (TME), human leukocyte antigen (HLA) genes, and tumor mutation burden (TMB) analyses, between high-risk and low-risk groups. RESULTS: A predictive model with three FRGs (HSPA5, AURKA, and TSC22D3) was constructed. Patients in the high-risk group had worse PFS compared with patients in the low-risk group. Functional analysis results revealed that ssGSEA, immune cell infiltration, TME, HLA, and TMB were closely associated with ferroptosis. CONCLUSION: The prognostic model constructed in this study can effectively predict PFS for patients with PTC.