Feed

Objective: To characterize the longitudinal evolution of thyroid nodules across ultrasound-based American College of Radiology Thyroid Imaging Reporting and Data System (ACR-TIRADS) categories with respect to nodule size, ultrasonographic features, and the presence of abnormal cervical lymph nodes. Methods: A single-center retrospective cohort study. Clinical and follow-up data were collected for patients with thyroid nodules who initially attended the Endocrinology Outpatient Clinic of the First Medical Center of Chinese PLA General Hospital from January 2018 to March 2022 (the first month of each quarter served as the enrollment period). Collected variables included gender, age, nodule size, number of nodules, and ultrasonographic features. Nodules were classified according to the ACR-TIRADS criteria. Generalized estimating equations were used to compare changes in indicators such as nodule size and ultrasonographic features during follow-up. Results: In total, 5 467 patients with 12 668 nodules were included. The mean age of the enrolled patients was (49.6±13.0) years, and 76.2% (4 164/5 467) were female. Multiple nodules were present in 83.7% (4 575/5 467) of patients. Among the 12 668 nodules at baseline, the proportions of nodules in ACR-TIRADS categories 1-2, 3, 4, and 5 were 70.7% (8 957/12 668), 19.6% (2 485/12 668), 6.1% (769/12 668), and 3.6% (457/12 668), respectively. The mean follow-up duration was (48.2±3.8) months. During follow-up, regardless of the ACR-TIRADS category, only 25.6% (3 246/12 668) of nodules showed significant enlargement, and most remained stable in size. The difference in nodule size changes across different ACR-TIRADS categories was statistically significant (Wald χ²=126.37, P<0.001). ACR-TIRADS category 5 nodules had the highest proportion of stable size, reaching 75.7% (346/457), while ACR-TIRADS category 3 nodules had the highest proportion of enlargement (29.2%, 726/2 485). Overall, 12.7% (1 605/12 668) and 3.9% (490/12 668) of nodules were upgraded or downgraded, respectively, due to changes in ultrasonographic features. The difference in ultrasonographic feature changes across different categories was statistically significant (Wald χ²=1 872.45, P<0.001). The highest rates of classification upgrade and downgrade were observed in ACR-TIRADS category 4 nodules (25.9%, 199/769) and category 5 nodules (21.4%, 98/457), respectively; ACR-TIRADS categories 1-2 nodules had the highest proportion of stable classification (87.4%, 7 826/8 957). During follow-up, the papillary thyroid carcinoma (PTC) detection rates for nodules that underwent fine-needle aspiration biopsy due to simple enlargement, classification upgrade alone, or enlargement combined with classification upgrade were 21.8% (47/216), 35.7% (71/199), and 41.8% (41/98), respectively. The overall difference in PTC detection rates among the three biopsy indications was statistically significant (Wald χ²=18.62, P<0.001), and the detection risk showed a significant increasing trend with the elevation of ACR-TIRADS category, with the highest risk observed in the enlargement combined with classification upgrade (OR=2.78, 95%CI 2.15-3.59, P<0.001). Among the 5 467 patients with thyroid nodules, 102 had abnormal cervical lymph nodes on ultrasound at baseline. During follow-up, abnormal lymph nodes resolved spontaneously in 42 cases, while 53 patients developed new abnormal lymph nodes. Conclusions: For thyroid nodules with different ultrasound-based ACR-TIRADS categories, most clinical characteristics including nodule size and ultrasonographic features remain stable during approximately 4 years of follow-up. The overall incidence of abnormal cervical lymph nodes is low and some can resolve spontaneously. Both nodule enlargement and classification upgrade indicate an increased probability of malignancy of thyroid nodules, and the malignant risk is higher when ACR-TIRADS classification upgrade occurs alone or in combination with nodule enlargement.