Prediction model of lateral cervical lymph node metastasis in papillary thyroid carcinoma based on SEER database

Papillary thyroid carcinoma (PTC) is the predominant form of thyroid cancer and lymph node metastasis (LNM) significantly impacts patient prognosis. Preoperatively identifying lateral lymph node metastasis (LLNM) presents significant challenges, as current diagnostic techniques, such as ultrasonography, have limited sensitivity and precision. This study aimed to develop and validate a predictive model for LLNM in patients with PTC, using data from the Surveillance, Epidemiology, and End Results (SEER) database and external validation cohorts.

Data from 18,342 patients with PTC diagnosed from 2016 to 2020 were retrieved from the SEER database. The patients were arbitrarily categorized into training (n=12,839) and validation (n=5,503) cohorts. Both univariate and multivariate logistic regression analyses were conducted to identify the independent risk factors for LLNM. A predictive nomogram was developed based on these factors, and its accuracy was assessed using receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA).

Five independent predictors of LLNM were identified: sex, age, race, tumor (T) stage, and metastasis (M) stage. The nomogram demonstrated strong predictive performance, with an area under the curve (AUC) of 0.715 [95% confidence interval (CI): 0.706-0.724] in the training cohort and 0.707 (95% CI: 0.693-0.720) in the validation cohort. The calibration curves indicated good agreement between the predicted and actual outcomes, while DCA confirmed the clinical applicability of the model across various risk thresholds.

This study successfully developed a predictive model for LLNM in patients with PTC by integrating demographic and clinicopathological indicators. This model demonstrates significant predictive precision and practical clinical use, assisting medical professionals in identifying high-risk patients and optimizing surgical choices. Further studies incorporating more variables are warranted to improve the diagnostic accuracy of the model.