Developing a Radiotherapy and Immune‐Related Genes‐Based Prognostic Model to Predict Prognosis and Immune Microenvironment in Thyroid Cancer

Purpose : This study sought to identify radiotherapy‐ and immune‐related genes linked to prognosis, as well as develop a predictive model for thyroid cancer (THCA). Methods : The THCA transcriptomic dataset was obtained from the UCSC Xena Database, followed by screening for differentially expressed immune‐related genes (DE‐IRGs) linked to radiotherapy sensitivity. Then, a risk score (RS) predictive model for patient stratification was developed by selecting DE‐IRGs associated with prognosis, and its efficacy was evaluated using a nomogram and receiver operating characteristic curves. Next, KEGG pathways and immune cell infiltration patterns in different risk categories were estimated. Results : Our study identified 40 DE‐IRGs in THCA, and nine genes, including ULBP2, PDIA3, PIK3R3, PGF, CAMP, CLDN4, GRP, SFTPA1 , and NR1D1 , were identified as characteristic genes. Subjects with low RS have better survival outcomes. The prediction nomogram was constructed by using the RS model, radiotherapy, and age. Patients across different risk groups showed distinct involvement in 11 KEGG pathways, including tyrosine metabolism, glycine serine, and threonine metabolism, as well as infiltration of eight immune cells, such as T‐cell follicular helper, B‐cell plasma, T‐cell regulatory (Tregs), and T‐cell CD4 + memory resting. Among the selected DE‐IRGs, PDIA3, PGF, and GRP were related to more immune‐related items than other genes. Conclusion : A prognostic model utilizing the RS value was successfully proposed. Meanwhile, we identified PDIA3, PGF, and GRP as novel treatment biomarkers for THCA.