Abstract
BACKGROUND: Thyroid papillary carcinoma (TPC) is the most common type of thyroid cancer (TC). The prognosis of TPC patients with tumor-cell metastasis is poor. Therefore, this study aims to develop a model for predicting TPC patients' recurrence-free survival (RFS). METHODS: We included 546 TPC patients who were clinically and pathologically diagnosed with TPC. The methylation biomarkers that associate with RFS were explored. These 546 samples were divided into training dataset (first 70%) and validation dataset (remaining 30%) randomly. The training dataset was used to identify prognostic biomarkers and construct risk prediction model, in addition, the validation dataset was used to verify the predictive performance of the model. We used Cox proportional hazard analysis and the least absolute shrinkage and selection operator (LASSO) Cox regression analysis to identify the significant predictive biomarkers, and establish the relapse risk prediction model from the identified biomarkers. RESULTS: A 6-DNA methylation signature yielded a high evaluative performance for RFS. The Kaplan-Meier analysis indicated that the 6-DNA methylation signature could significantly distinguish the high- and low-risk patients in training, validation and entire sets. In addition, a nomogram was constructed based on risk score, metastasis status and residual tumor status, and C-index, receiver operating characteristic (ROC) and the calibration plots analysis which demonstrated the good performance and clinical utility of the nomogram. CONCLUSIONS: The results suggested that the 6-DNA methylation signature is the independent prognostic marker for RFS and functioned as a significant tool for guiding the clinical treatment of TPC patients.