Risk-stratification of pediatric thyroid nodules using ATA and ACR TIRADS adult thyroid nodule scoring systems.

OBJECTIVE: To evaluate the performance of American Thyroid Association (ATA) and American College of Radiology Thyroid Imaging and Reporting Data System (ACR TI RADS) in pediatrics. METHODS: Retrospective review, patients aged ≤18 years with thyroid nodules from 2002-2022. Patients with no available ultrasound images or pathology results were excluded. Each neck ultrasound was independently and blindly reviewed and scored using ATA and TIRADS by two study radiologists; discordant scores were arbitrated by a third radiologist. Thyroid nodules that did not undergo surgical resection were considered benign if their size was stable and/or if repeat fine needle aspiration (FNA) was again benign. The performance of each system was calculated. Sensitivity analyses performed to determine the optimal size cut-off for FNA biopsy while maintaining sensitivity equivalent to adults. RESULTS: 77 patients (median age, 15 years [range 2-18]); 71.4% female) with 115 thyroid nodules. Malignancy rate 27.8%. Using adult size cut-offs for FNA, ATA and TIRADS would have failed to recommend biopsy in 11/32 and 15/32 malignant nodules, respectively (sensitivity 66%, specificity 55%, versus 53%, 87% respectively). In our sensitivity analysis the optimal size cut-offs for ATA were: 2 cm for very low risk, 1 cm for low risk, and 0.5 cm for intermediate and high-risk categories (sensitivity 94%; specificity 33%). For TIRADS were: 3 cm for TIRADS-2, 0.8 cm for TIRADS-3 and TIRADS-4, and 0.5 cm for TIRADS-5 (sensitivity 91%; specificity 59%). CONCLUSION: ATA and TIRADS have low sensitivity in pediatrics. Lowering the size cutoffs for FNA increases sensitivity to >90%.