Diagnosis and Prognosis of Thyroid Cancer by Immune-related Genes

Abstract

Background: Thyroid carcinoma (THCA) is the most common malignant endocrine tumor with low mortality and a relatively good prognosis. Immune genes have attracted much attention as molecular markers of THCA prognosis and potential targets of immunotherapy.

Methods: Our study analyzed the transcriptome and clinical data of immune-related genes (IRGs) of THCA in gene expression omnibus, the cancer genome atlas-THCA, and ImmPort databases. By univariate Cox regression analysis, 15 genes were significantly correlated with the survival of patients with THCA. Five IRGs ( NMU, UBE2C, CDKN2A, COL19A1, and GPM6A ) were selected by LASSO regression analysis as independent prognostic factors to construct a disease-free survival-related prognostic risk model.

Results: Kaplan-Meier survival analysis showed that there was a significant difference in disease-free survival between high and low-risk groups. The higher the risk score, the worse the survival of patients. Clinical correlation analysis showed that age and Stage stage of patients were correlated with risk score ( P < 0.05). Quantitative real-time polymerase chain reaction confirmed that there were differences in the expression of 5 IRGs between tumor tissues and normal thyroid tissues. Spearman correlation analysis indicated that the relative expression levels of NMU, CDKN2A, UBE2C, COL19A1 , and GPM6A were positively correlated with programmed death-ligand 1 and recombinant a disintegrin and metalloproteinase with thrombospondin 1.

Conclusion: Based on the bioinformatics method, we constructed a prognosis evaluation model and risk score system of IRGs in THCA, which provided a reference for predicting the prognosis of patients with THCA.