甲状腺影像报告和数据系统在甲状腺结节分类中的前瞻性验证

Prospective validation of the thyroid imaging reporting and data system on thyroid nodules

目的：前瞻性验证甲状腺影像报告和数据系统（TI-RADS）对甲状腺结节分类的应用价值。方法根据2011年Kwak提出的TI-RADS分类标准对415例甲状腺结节患者的810个结节进行分类，并与病理结果对照，应用受试者操作特性（ROC）曲线评价TI-RADS鉴别甲状腺良恶性的价值，计算各分类的实际恶性百分率。结果经病理证实810个结节中恶性结节161个、良性结节649个；TI-RADS 2类11个（全部为良性结节）、TI-RADS 3类370个（良性结节368个，恶性结节2个）、TI-RADS 4a类150个（良性结节143个、恶性结节7个）、TI-RADS 4b类116个（良性结节87个、恶性结节29个）、TI-RADS 4c类146个（良性结节39个、恶性结节107个）；TI-RADS 5类17个（良性结节1个、恶性结节16个）。以病理诊断结果作为金标准，TI-RADS 2~5类ROC曲线下面积为0.81， TI-RADS 2、3、4a、4b、4c、5类的实际恶性百分率为0、0.5%、4.6%、25.0%、73.0%、94.0%。结论 Kwak等提出的TI-RADS经过验证，实际恶性百分率与理论恶性百分率有很高的符合度，在甲状腺结节性疾病超声检查中具有重要的诊断价值。

Objective To prospectively validate the feasibility and usefulness of the thyroid imaging reporting and data system （TI-RADS） suggested by Kwak. Methods According to the TI-RADS published in Radiology by Kwak et al in 2011, the TI-RADS score of 810 thyroid nodules in 415 patients whose pathologic diagnoses were available were categorized . According to TI-RADS and pathologic results, receiver operating characteristic （ROC） curve was plotted, and the probability of malignancy in each category was calculated. Results Eight hundred and ten pathologically proven thyroid nodules included 649 benign and 161 malignant lesions. Among them, 11 nodules were categorized as TI-RADS 2 （all benign nodules）;370 nodules were categorized as TI-RADS 3 （368 benign nodules and 2 malignant nodules）;150 nodules were categorized as TI-RADS 4a （143 benign nodules and 7 malignant nodules）;116 nodules were categorized as TI-RADS 4b （87 benign nodules and 29 malignant nodules）; 146 nodules were categorized as TI-RADS 4c （39 benign nodules and 107 malignant nodules）;17 nodules were categorized as TI-RADS 5 （1 benign nodules and 16 malignant nodules）. The area under the curve of TI-RADS was 0.89, and the probability of malignancy in nodules with a classiifcation of TI-RADS 2, 3, 4a, 4b, 4c and 5 was 0, 0.5%, 4.6%, 25.0%, 73.0%and 94.0%, respectively. Conclusions The TI-RADS suggested by Kwak has great diagnostic value in diagnosing thyroid nodules. The actual probability of malignancy conforms with the theoretical risk of malignancy.