Comparison and Calibration of the 2025 and 2015 ATA Recurrence-Risk Classifications in Differentiated Thyroid Cancer

Abstract Context The 2025 American Thyroid Association (ATA) guidelines maintained the low- and high-risk categories and subdivided the intermediate-risk category of differentiated thyroid cancer (DTC) into low-intermediate and intermediate-high tiers. Objective To determine whether observed rates of structural disease match the risks the 2025 and 2015 ATA classifications assign, and to compare the 2 systems. Patients and Methods All 258 consecutive adults with DTC diagnosed in 2017 were classified under both systems using initial postoperative data. The main outcome was structural disease at any assessment, compared against the guideline-assigned ranges. Results Sixty-three patients (24.4%) were reclassified, including 52 of 121 (43.0%) low-risk patients upstaged. Thirty-two (12.4%) developed structural disease. Observed 2025 rates were 0.0% (95% CI, 0.0–5.3), 1.5% (0.3–8.0), 15.5% (9.3–24.7), and 47.4% (32.5–62.7) for low, low-intermediate, intermediate-high, and high risk. The low-intermediate interval fell entirely below its assigned 10–15% range (P = .013 vs the 10% lower bound); the other categories of both systems were consistent with theirs. No patient upstaged from the 2015 low-risk category developed structural disease. Structural disease later developed in 7 of 14 patients with a biochemically incomplete first response, 2 of 79 indeterminate, and 0 of 129 excellent responses. Discrimination did not differ between the two systems (ΔAUC 0.002; 95% CI, −0.031 to 0.036). Conclusions The 2025 low-intermediate category overestimated recurrence risk in this cohort, and upstaging of low-risk patients identified nobody who developed structural disease. Confirmation in larger multicenter cohorts is needed before these tiers guide treatment intensity.