Yunhe Liu

Context: The 2025 American Thyroid Association (ATA) guidelines updated recurrence risk stratification for papillary thyroid cancer (PTC). Whether age modifies the absolute structural recurrence risk represented by these strata is unclear. Objective: To determine whether age modifies 5-year structural recurrence risk across ATA 2025 stratification in PTC and improves risk prediction. Methods: This retrospective cohort study included 14 397 patients with PTC who underwent initial surgery at a tertiary cancer center in China from 2000 to 2018. The primary outcome was 5-year structural recurrence-free survival. Cox models with restricted cubic splines evaluated age effects and ATA2025-by-age interaction. ATA-based and age-interaction models were compared by temporal validation and decision-curve analysis. Results: The median age was 43.0 years, 10 867 patients (75.5%) were female, and 758 (5.3%) experienced structural recurrence during a median follow-up of 60.0 months. Observed 5-year risk rose from 1.7% to 12.5% across ATA2025 stratification. Age modified risk across stratifications (global interaction P < .001). After treatment adjustment, 5-year risk fell from 2.37% at age 35 years to 0.69% at age 65 years in the low-risk group but rose from 8.21% to 19.28% in the high-risk group. In temporal validation, the C-index was 0.745 for both models; 5-year areas under the curve were 0.743 and 0.747, and Brier scores were 0.043 and 0.042. Decision curves showed little overall difference. Conclusion: Age identified meaningful heterogeneity in absolute structural recurrence risk, particularly in high-risk PTC, but added little to predictive performance or decision value. External validation is required before age is used to guide treatment or surveillance.