Ultrasound Pattern-Based Malignancy Risk Stratification and Subcategorization of K-TIRADS Intermediate-Suspicion Thyroid Nodules to Improve Biopsy Decision-Making

Why clinicians should know about it

• Picked for Radiology, Radiation Oncology, Nuclear Medicine, Medical Physics and Imaging (paper of the day, 9 November 2026): Ultrasound malignancy stratification informs thyroid nodule biopsy decisions

Abstract

OBJECTIVE: This study aimed to develop an ultrasound (US) pattern-based subcategorization system for Korean Thyroid Imaging Reporting and Data System (K-TIRADS) 4 nodules and to compare it with US feature-based subcategorization system and 2021 K-TIRADS as biopsy criteria. MATERIALS AND METHODS: Overall, 1,409 K-TIRADS 4 nodules (>1 cm) with final diagnoses, including 190 malignant lesions, were included. A US pattern-based subcategorization system for K-TIRADS 4 nodules was developed using US predictors of malignancy in each of the following three subgroups defined according to the 2021 K-TIRADS: solid hypoechoic nodules without suspicious features (subgroup 1), partially cystic or iso-/hyperechoic nodules with suspicious features (subgroup 2), and diffuse infiltrative lesions (subgroup 3). The biopsy criterion modified according to US pattern-based subcategorization (>1 cm for K-TIRADS 4B and >1.5 cm for K-TIRADS 4A) was compared with those based on the US feature-based subcategorization system and the 2021 K-TIRADS in terms of theoretical sensitivity (cancers detected at biopsy/all cancers), diagnostic yield (cancers detected at biopsy/all nodules), and unnecessary biopsy rate (benign lesions undergoing biopsy/all nodules). RESULTS: The US pattern-based subcategorization criteria included macrocalcification in subgroup 1, rim calcification or at least two suspicious features in subgroup 2, and a diffuse infiltrative pattern in subgroup 3; these criteria significantly increased malignancy risk (P < 0.001). Nodules were subcategorized as K-TIRADS 4B (higher malignancy risk, 38.6%) or K-TIRADS 4A (lower malignancy risk, 9.3%; P < 0.001). Compared with the >1.5-cm criterion and the US feature-based criterion, the modified biopsy criterion based on US pattern-based subcategorization significantly increased sensitivity by 17.4 percentage points (%p) and 10.0%p, diagnostic yield by 2.3%p and 1.3%p, and the unnecessary biopsy rate by 4.4%p and 2.9%p, respectively (all P < 0.001). CONCLUSION: As a biopsy criterion, US pattern-based subcategorization significantly improved sensitivity and diagnostic yield, with only small increases in the unnecessary biopsy rate compared with the other two criteria. US pattern-based subcategorization may provide an intuitive and clinically effective method for risk stratification and management of K-TIRADS 4 thyroid nodules.

Abstract as published, via PubMed.