Endocrine Abstracts

Objective: To evaluate computed tomography (CT) features for diagnosing metastatic cervical lymph nodes (LNs) in patients with differentiated thyroid cancer (DTC) and propose CT-based risk stratification system modified from the Korean Thyroid Imaging Reporting and Data System (K-TIRADS) guidelines.

Materials and Methods: A total of 463 LNs from 397 patients with DTC who underwent preoperative CT staging and US-guided fine needle aspiration of LNs were included. The CT features of each LN were evaluated: presence of a hilum, cystic change, calcification, strong enhancement, and heterogeneous enhancement. Multivariable logistic regression analysis was performed to identify independent CT features associated with metastatic LNs, and the diagnostic performance of each CT feature was evaluated. LNs were classified into probably benign, indeterminate, and suspicious categories according to K-TIRADS, and the malignancy risk of each category malignancy risk was calculated.

Results: The absence of a hilum, strong enhancement, calcification, and cystic change were independent CT features associated with metastatic LNs in patients with DTC. Strong enhancement, calcification, and cystic change showed moderate to high specificity (70.1–100%) and positive predictive value (91.8–100%). The absence of a hilum showed high sensitivity (97.8%) but low specificity (34.0%). When LNs were classified according to K-TIRADS guidelines, the malignancy rates of the probably benign, indeterminate, and suspicious groups were 6.7%, 13.2%, and 90.9%, respectively. Our proposed modified CT criteria excluding heterogeneous enhancement from the four K-TIRADS suspicious features gave malignancy rates of 6.5%, 17.0%, and 92.5% for probably benign, indeterminate, and suspicious categories, respectively.

Conclusion: CT features of absence of a hilum, strong enhancement, calcification, and cystic change were independently associated with metastatic LNs in patients with DTC. Our modified CT criteria provided better risk stratification of LNs than the current K-TIRADS guidelines. Our study results may provide a basis for revising CT-based LN classification in future guidelines.