Preoperative prediction of central lymph node metastasis in clinically lymph node negative papillary thyroid microcarcinoma: a nomogram based on clinical and ultrasound features.

Abstract

Background: Accurate preoperative assessment of central lymph nodes is crucial for determining the extent of surgery for papillary thyroid microcarcinoma (PTMC). Patients who are clinically lymph node negative (cN0) lack clinical evidence of central lymph node metastasis (CLNM) on preoperative ultrasonography or computed tomography. This study aimed to identify clinical factors associated with CLNM based on ultrasonographic features and clinical data, and to develop a nomogram for personalised clinical decision-making.

Methods: A retrospective analysis was conducted on patients diagnosed with cN0 PTMC who were treated at the Vascular Thyroid Surgery Department of the Hospital from December 2020 to February 2022, totaling 834 individuals. The patients were divided into CLNM and non-CLNM groups based on postoperative pathology. The clinical characteristics and ultrasonographic features of the PTMC were collected. The Least Absolute Shrinkage and Selection Operator (LASSO) regression method was applied in R for feature selection. A nomogram was then developed based on multivariable logistic regression using the predictors selected by the LASSO algorithm. The receiver operating characteristic curve and Hosmer-Lemeshow test were used to assess the discrimination and calibration of the nomogram model, respectively. Decision curve analysis (DCA) was performed using the Risk Model Decision Analysis package to evaluate the clinical utility of the model in the validation dataset.

Results: Six variables associated with patients with PTMC were identified through LASSO shrinkage and selection operator regression analysis and used to establish the nomogram. The predictive model showed an area under the receiver operating characteristic curve (AUC) of 0.719 (95% confidence interval (CI) 0.681-0.757), and in internal validation, the AUC was 0.717 (95% CI 0.683-0.754). The calibration curve indicated a good fit for the model, and the Hosmer-Lemeshow test demonstrated a close match between the predicted and observed values (P = 0.437). DCA revealed that applying the nomogram to predict the risk of CLNM would be beneficial for patients with PTMC when the threshold probability was between > 12.5% and < 75%.

Conclusion: The LASSO regression model nomogram based on clinical risk factors and ultrasonographic features is valuable in predicting CLNM in cN0 PTMC, and can assist surgeons in making more personalised clinical decisions.