Prospective Validation of ATA Risk Score for Papillary Thyroid Microcarcinoma: An ITCO Real-World Study

Context: The risk of recurrence of papillary thyroid carcinoma (PTC) smaller than 1 cm (microPTC) is low. Predictors of disease persistence in microPTC are still unclear. Objective: To compare the clinical and pathological characteristics of microPTCs with macrocarcinomas (PTC > 1 cm), identifying the predictors of biochemical and structural incomplete response 1 year after initial treatment in microPTC. Methods: We included patients consecutively enrolled in the Italian Thyroid Cancer Observatory (NCT04031339), and selected patients with a histological diagnosis of PTC for whom complete pathological, clinical, treatment information, and results at the 1-year follow-up visits were available. Results: Among 5038 patients in the cohort, 2345 (46.5%) had a microPTC. Patients with microPTCs had tumors with more indolent pathological features: only 3% of patients were classified as high risk according to the American Thyroid Association (ATA) risk stratification system for persistent or recurrent disease and 1% had distant metastases at diagnosis. MicroPTCs had a significantly better outcome: only 5% had a biochemical response and 2.3% a structural incomplete (SIR) response. Distant metastases at diagnosis were the best predictor of SIR in microPTCs (OR 5.13, 95% CI 1.11-23.73, P = .04). In a subgroup of 925 patients treated by total thyroidectomy and radioiodine treatment, the best predictor of SIR was the ATA high risk (OR 5.47, 95% CI 1.42-21.04, P = .01). Conclusion: Our study confirms the favorable initial outcome of microPTC in a large series. We demonstrate that the ATA risk classification is reliable in predicting biochemical and structural persistence in patients with microPTC. Distant metastases, although rare, remain the best predictor of structural persistence at 1-year follow-up. These findings underscore the importance of tailored management strategies based on comprehensive risk stratification, rather than solely on tumor size.