Ann Surg Oncol. 2026 Aug 8. doi: 10.1245/s10434-026-20381-1. Online ahead of print.
ABSTRACT
BACKGROUND: The 2025 American thyroid association (ATA) guidelines introduced a revised risk stratification system for thyroid cancer incorporating histology-specific models and a four-tier classification framework. However, its clinical performance in patients with papillary thyroid carcinoma (PTC) remains insufficiently validated. This study compared the 2015 and 2025 ATA risk stratification systems for predicting PTC recurrence and evaluated reclassification patterns.
PATIENTS AND METHODS: This retrospective cohort study included 284 patients who underwent surgery for PTC at Samsung Medical Center, Seoul, Korea, between 2019 and 2021. Risk classification was performed according to the 2015 ATA three-tier system (low, intermediate, and high risk) and the 2025 ATA four-tier system (low, low-intermediate, intermediate-high, and high risk). Predictive performance for recurrence was assessed using disease-free survival (DFS) and the integrated area under the receiver operating characteristic curve (iAUC).
RESULTS: During follow-up, structural recurrence occurred in 23 patients (8.1%). Under the 2015 system, recurrence occurred in 0/119 low-risk, 8/106 intermediate-risk, and 15/59 high-risk patients. Under the 2025 system, recurrence occurred in 0/86 low-risk, 1/56 low-intermediate-risk, 2/53 intermediate-high-risk, and 20/89 high-risk patients, with recurrence events predominantly concentrated in the high-risk group. The 2025 ATA system resulted in substantial reclassification, particularly among patients initially classified as intermediate risk. The iAUC for predicting DFS was 0.873 for the 2015 system and 0.852 for the 2025 system.
CONCLUSIONS: The 2025 ATA risk stratification system demonstrated predictive performance comparable to the 2015 system while concentrating recurrence events predominantly within the high-risk group, suggesting improved risk discrimination in patients with PTC.
PMID:42570987 | DOI:10.1245/s10434-026-20381-1
