Value of dynamic contrast-enhanced ultrasound in predicting cervical lymph node metastasis in papillary thyroid carcinoma patients with Hashimoto's thyroiditis.

BACKGROUND

Hashimoto's thyroiditis (HT) and papillary thyroid carcinoma (PTC) commonly coexist. An accurate assessment of cervical lymph node metastasis (CLNM) is crucial for determining treatment options and predicting prognosis. However, traditional examination methods have certain limitations. This study aimed to investigate the potential utility of dynamic contrast-enhanced ultrasound (DCE-US) using VueBox software in assessing CLNM in patients with PTC coexisting with HT.

METHODS

A retrospective analysis was performed on the clinicopathological data and ultrasound characteristics of 180 thyroid cancer patients who underwent either biopsy or surgery from January 2022 to November 2023. The dataset was partitioned into training and validation sets with a 6:4 ratio. Statistical analyses, including -tests, chi-squared tests, and rank-sum tests, were conducted to evaluate the data. Univariate analysis, least absolute shrinkage and selection operator (LASSO) regression, and multivariate logistic regression were employed to identify the predictive factors for CLNM. Based on these analyses, three predictive models were developed: Model 1, incorporating clinical factors; Model 2, integrating clinical and ultrasound factors; and Model 3, a combined model that included both clinical and ultrasound factors using DCE-US. The diagnostic performance of each model was assessed using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA).

RESULTS

The Model 3 exhibited superior performance compared to Models 1 and 2. Specifically, in the training set, Model 3 achieved an area under the curve (AUC) value of 0.924 [95% confidence interval (CI): 0.857-0.966], and in the validation set, the AUC value was 0.905 (95% CI: 0.813-0.962). These values were significantly higher (P<0.05) than those of Model 1, which had a training AUC of 0.724 (95% CI: 0.630-0.806) and a validation AUC of 0.677 (95% CI: 0.557-0.783), as well as Model 2, with a training AUC of 0.854 (95% CI: 0.773-0.915) and a validation AUC of 0.797 (95% CI: 0.683-0.883). Moreover, DCA indicated that Model 3 provided a greater net benefit than Models 1 and 2 in both the training and validation cohorts.

CONCLUSIONS

The use of VueBox perfusion analysis in DCE-US provides additional value in predicting CLNM in PTC patients with HT. The integrated model, which combines clinical and ultrasound factors, is a valuable diagnostic tool.