Özdemir Mustafa, Türk Gamze, Bilgili Mustafa, Akay Ebru, Koç Ali
Department of Radiology, Sakarya University Faculty of Medicine, Sakarya, Turkey.
Department of Radiology, Kayseri City Hospital, Kayseri, Turkey.
Exp Clin Endocrinol Diabetes. 2025 Feb;133(2):98-104. doi: 10.1055/a-2498-7952. Epub 2024 Dec 9.
Several ultrasound-based risk stratification systems (RSSs) have been developed and introduced into clinical practice for managing thyroid nodules. However, there are essential differences among these systems. This study aimed to determine and compare the category-based diagnostic performance of four ultrasound-based risk stratification systems in the detection of thyroid cancer: ACR-TIRADS, ATA, K-TIRADS, and EU-TIRADS.This study included 4238 nodules sampled by fine-needle aspiration biopsy between January 2018 and December 2021. Nodules were classified according to ultrasound imaging features and correlated with biopsy results. The diagnostic success of the risk stratification systems was evaluated and compared.Of the 4238 nodules, 3861 (91.1%) were benign and 376 (8.9%) were malignant. Malignancy was significantly higher in hypoechoic and marked hypoechoic nodules (p=0.001), and solid nodules (p=0.002). For detection of malignancy, areas under the receiving operator characteristics curves were 0.862, 0.850, 0.842, and 0.835 for 2017 ACR-TIRADS, EU-TIRADS, for K-TIRADS, and 2015 American Thyroid Association guidelines, respectively. EU-TIRADS showed the highest sensitivity (91%), whereas ACR-TIRADS had the highest specificity (87%). Compared to other risk stratification systems, ACR-TIRADS resulted in significantly fewer unnecessary biopsies (p=0.009). All RSSs show high diagnostic accuracy and have their own advantages and disadvantages. When selecting an appropriate RSS, the population, the prevalence of the disease, and gender distribution should be considered.
已经开发了几种基于超声的风险分层系统（RSSs）并将其引入临床实践以管理甲状腺结节。然而，这些系统之间存在本质差异。本研究旨在确定并比较四种基于超声的风险分层系统在甲状腺癌检测中的类别诊断性能：美国放射学会甲状腺影像报告和数据系统（ACR-TIRADS）、美国甲状腺协会（ATA）、韩国甲状腺影像报告和数据系统（K-TIRADS）以及欧洲甲状腺影像报告和数据系统（EU-TIRADS）。
本研究纳入了2018年1月至2021年12月期间通过细针穿刺活检取样的4238个结节。根据超声成像特征对结节进行分类，并与活检结果相关联。对风险分层系统的诊断成功率进行评估和比较。
在4238个结节中，3861个（91.1%）为良性，376个（8.9%）为恶性。低回声和显著低回声结节（p=0.001）以及实性结节（p=0.002）中的恶性率显著更高。对于恶性肿瘤的检测，2017年ACR-TIRADS、EU-TIRADS、K-TIRADS以及2015年美国甲状腺协会指南的受试者操作特征曲线下面积分别为0.862、0.850、0.842和0.835。EU-TIRADS显示出最高的敏感性（91%），而ACR-TIRADS具有最高的特异性（87%）。与其他风险分层系统相比，ACR-TIRADS导致的不必要活检显著更少（p=0.009）。所有风险分层系统均显示出较高的诊断准确性，且各有优缺点。选择合适的风险分层系统时，应考虑人群、疾病患病率和性别分布。
