Construction and validation of a nomogram for predicting distant metastasis in Hurthle cell carcinoma of the thyroid

To analyze independent risk factors for distant metastasis of Hurthle cell carcinoma (HCC)of the thyroid and to construct and validate a nomogram, thereby assisting clinicians in making more individualized clinical decisions. Patient data were retrieved from the Surveillance, Epidemiology, and End Results (SEER) database. Chi-square tests and Logistic regression analysis were used to determine independent risk factors for distant metastasis of HCC. Logistic regression models based on significant independent risk factors were established using the rms package of R software (version 4.3.3) and visualized as a nomogram. The performance of the nomogram was evaluated using subject operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). A total of 1312 HCC patients were included, among whom 48 patients developed distant metastasis, accounting for 3.66% of the entire cohort. Results of chi-square test and univariate logistic regression analysis indicated that age, gender, marital status, tumor size, and lymph node metastasis were significantly associated with distant metastasis of HCC. Multifactorial logistic regression analysis revealed that age, gender, marital status and lymph node metastasis were independent risk factors for distant metastasis of HCC, which were used to construct the nomogram. Results of ROC curve analysis, calibration curves, and DCA demonstrated that the nomogram had good predictive performance. We successfully constructed and validated a nomogram for predicting distant metastasis of HCC. This nomogram is important for clinicians to promptly identify patients at high risk of distant metastasis and formulate more individualized clinical decisions.