Abstract
OBJECTIVE: This study aims to develop a comprehensive nomogram to accurately predict differentiated thyroid cancer (DTC) involved lymph node, enabling the personalization of treatment plans and optimizing patient outcomes. METHODS: The nomogram was developed through an analysis of data obtained from derivation cohorts. To identify independent predictors of involved lymph node, we utilized the Least Absolute Shrinkage and Selection Operator (LASSO) method in conjunction with multivariable logistic regression. These predictive factors were subsequently integrated into the nomogram's design. To ensure the tool's robustness and clinical applicability, we assessed its discriminative ability, calibration, and clinical utility in both the derivation cohort and an independent validation cohort. RESULTS: The identification of four key parameters - Thyroglobulin, Lymphocyte count, BRAF(V600E), and the number of tumor foci - emerged as robust predictors of involved lymph node. The nomogram demonstrated outstanding performance across calibration, discrimination, and clinical applicability, with predictive accuracies of 0.992 in the derivation cohort and 0.920 in the validation cohort. CONCLUSION: Peripheral lymphocyte counts have been found to correlate with involved lymph node in DTC patients. A nomogram tailored specifically for predicting involved lymph node in DTC guides personalized postoperative radioactive iodine (RAI) and plays a critical role in the preoperative assessment.