Construction and evaluation of a prognosis prediction model for thyroid carcinoma based on lipid metabolism-related genes.

Background: Thyroid cancer is one of the most common tumors worldwide, and the molecular studies on lipid metabolism disorders in thyroid cancer remain unclear.

Aim: This study intends to explore the model constructed by lipid metabolism genes to evaluate the prognosis of thyroid cancer.

Methods: The data of thyroid cancer patients were obtained by The Cancer Genome Atlas database. The cancerous tissue from the thyroid gland was used to evaluate specific genes. Besides, Gene Set Enrichment Analysis (GSEA) and Cox proportional hazard regression models were adopted to identify the lipid metabolism genes of thyroid cancer. The survival status of patients with a risk score was analyzed by the Kaplan-Meier method, and the accuracy of the risk score was evaluated by the receiver operating characteristic (ROC) curve.

Findings: Age, tumor node metastasis stage, and risk score were independent prognostic factors for thyroid cancer. FADS1, WNT3A, PCDHA2 and ITGA5 were high-risk genes. The prognostic risk score model was established according to the four lipid metabolism genes. The overall survival of patients with high-risk thyroid cancer was significantly lower than that of low-risk patients in this study.

Discussion: According to the above findings, FADS1, WNT3A, PCDHA2, and ITGA5 are unfavorable factors for the prognosis of thyroid cancer in the pathway of lipid metabolism. A prognostic model composed of the above four genes was constructed, and it was confirmed that the model was not affected by age and sex.

Conclusion: The prognosis prediction model for thyroid cancer based on lipid metabolism related genes was successfully constructed, and the model had good predictive ability for the prognosis of thyroid cancer patients.