Comparative Predictive Value of TI-RADS,Elastography,Contrast-Enhanced Ultrasound,and Fine-Needle Aspiration Cytology in Assessing Cervical Lymph Node Metastasis of Papillary Thyroid Carcinoma

Abstract

Background

This study compared the diagnostic efficacy of Thyroid Imaging Reporting and Data System (TI-RADS), elastography, enhanced ultrasound, and fine-needle aspiration cytology (FNAC) in predicting cervical lymph node metastasis (LNM) in papillary thyroid carcinoma (PTC).

Objective

A retrospective analysis of 153 PTC patients (68 with LNM, 85 without) was conducted. All patients underwent preoperative TI-RADS scoring, elastography, enhanced ultrasound, and FNAC.

Methods

Clinical data from 153 patients with PTC admitted to our department between 2020 and 2024 were retrospectively analyzed. The patients were divided into the LNM group (n = 68) and the non-metastasis (non-LNM) group (n = 85) based on postoperative pathology. All patients underwent TI-RADS scoring, elastography, enhanced ultrasound, and FNAC examinations. The differences in the TI-RADS score of the primary lesion, elastography score, enhanced ultrasound features (enhancement intensity, uniformity, ring enhancement, etc.), and FNAC Bethesda classification between the two groups were compared. Statistical analyses were employed to identify differences in detection modalities. The diagnostic performance of each method and combinations was then evaluated by analyzing sensitivity, specificity, and area under the curve (AUC) from receiver operating characteristic (ROC) curves.

Result

Significant differences were observed between the groups with respect to TI-RADS category (≥5), elastography score (4–5 points), contrast-enhanced ultrasound (CEUS) features (low or heterogeneous enhancement, absence of ring enhancement), and FNAC results (Bethesda categories V–VI) (all P < 0.05). When used alone, elastography yielded the highest specificity (94.2%) but a relatively low sensitivity (48.5%). The combined model incorporating all four modalities achieved the highest overall predictive performance, with a sensitivity of 79.4%, specificity of 79.1%, and an AUC of 0.826.

Conclusion

While each modality has its own advantages, combining multimodal ultrasound techniques (TI-RADS, elastography, CEUS) with FNAC offers superior preoperative accuracy for predicting LNM in papillary thyroid cancer, demonstrating high clinical value.