BACKGROUND
Thyroid nodules are very common. The concern about any thyroid nodules is whether it is a cancer. Fortunately, only ~5% are cancerous. Ultrasound is the best initial imaging test to evaluate thyroid nodules. An ultrasound-guided thyroid biopsy can then be performed to determine it the nodule is cancer or benogn (non-cancerous). In the last 10 years, several professional societies have developed cancer risk stratification systems (RSSs) for thyroid nodules based on their ultrasound appearance. These RSSs then can be used to guide the selection of nodules with a higher risk of cancer that then require biopsy. The currently used systems include the American Thyroid Association (ATA) system and the Thyroid Imaging Reporting and Data Systems (TIRADS): European (EU-TIRADS), Korean (K-TIRADS), American College of Radiology (ACRTIRADS), and Chinese (C-TIRADS).

The multiple RSSs for thyroid nodules has created confusion and limited their consistent clinical application. To simplify this process, based on an international expert consensus, the International Thyroid Nodule Ultrasound Working Group recently published a standardized ultrasound vocabulary to be used when assessing the cancer risk of thyroid nodules. The aim of this study was to use the recent standardized ultrasound vocabulary proposed by the international expert consensus to compare the diagnostic performance of the 5 currently used RSSs for thyroid nodules.

THE FULL ARTICLE TITLE
Na DG et al et al. Diagnostic performance of five societies’ ultrasound risk stratification systems for thyroid malignancy according to nodule size: a comparison using a standardized ultrasound lexicon. Thyroid 2026;36(4):408–419; doi: 10.1177/10507256261429117. PMID: 41789443.

SUMMARY OF THE STUDY
The study evaluated 3143 consecutive patients with 3774 thyroid nodules >1 cm who underwent US-guided biopsy at GangNeung Asan Hospital, Korea between March 2017 and October 2024. Most patients were women with an average age of 58. The ultrasound features of thyroid nodules were described using the new standardized ultrasound vocabulary. The cancer risk of the nodules was then calculated based on their ultrasound features and size for each system. The thyroid nodules were grouped into the same 4 categories for each system: high-suspicion, intermediate-suspicion, low-suspicion for cancer, and no-biopsy-indicated. The diagnostic performance was compared between the five RSSs according to nodule size (≤2 cm versus >2 cm).

Significant differences were found among the five RSSs regarding the thyroid nodule classification into the four risk categories, cancer risk prediction, and correct cancer diagnosis across categories. The agreement across these systems for the ultrasound classification of nodules varied significantly. The highest agreement was noted between the K-TIRADS and ATA system for the high-suspicion (41%), intermediate-suspicion (45%), and low-suspicion (49%) categories. The proportion of nodules classified as being highly suspicious for cancer was highest in the EU-TIRADS (46%) with the highest risk of cancer noted in the K-TIRADS (71%) for this category. The proportion of low-suspicion nodules was highest in the ATA system (46%) with the highest risk of malignancy in the C-TIRADS (11%). The proportion of nodules diagnosed as cancerous was highest in the EU-TIRADS (82%) for the high-suspicion group, highest in the K-TIRADS (32.5%) for the intermediate-suspicion group, and highest in the C-TIRADS (20%) in the low-suspicion group.

The ATA system and EU-TIRADS showed higher rate of identifying cancer but also had higher benign biopsy rate for both small (≤2 cm) and large nodules (>2 cm), whereas ACR-TIRADS had both a lower rate of identifying cancer and a lower benign biopsy rate across both size groups. K-TIRADS demonstrated the lowest rate of identifying cancer and rate of benign biopsies for small nodules, but a high benign rate for large nodules. C-TIRADS showed a low rate of identifying cancer for both size groups, but a significantly higher rate of benign biopsies than ACR-TIRADS in the large nodule group.

Most missed cancers were from the intermediatesuspicion group in the EU-TIRADS, K-TIRADS, and ACR-TIRADS, from the low-suspicion group in the ATA system and from the no-biopsy-indicated group in the C-TIRADS. For small nodules, benign biopsies were most frequent in the high-suspicion group for the EU-TIRADS, and intermediate-suspicion group in the ATA system, K-TIRADS, ACR-TIRADS, and C-TIRADS. For large nodules, benign biopsies were most frequent in the low-suspicion group for the ATA system, EU-TIRADS, K-TIRADS, and C-TIRADS, and in the intermediate-suspicion group in the ACR-TIRADS.

WHAT ARE THE IMPLICATIONS OF THIS STUDY?
Comparison using the same standardized ultrasound vocabulary showed that the 5 currently used ultrasoundbased cancer risk stratification systems for thyroid nodules had significant differences in nodule risk classification and diagnostic performance, depending on the nodule size. These differences were mainly driven by variations in ultrasound criteria and nodule size thresholds used for biopsy by each system. The study results emphasize the importance of developing a unified, standardized ultrasound-based risk stratification system for cancer in thyroid nodules.

— Alina Gavrila, MD, MMSC
Elie Naous, MD

Table of Contents | PDF File for Saving and Printing