探讨美国放射学会甲状腺影像报告与数据系统（ACR TI-RADS）与2011版Kwak甲状腺影像报告与数据系统（Kwak TI-RADS）对甲状腺峡部结节的诊断价值。

回顾性选取2015年1月至2020年1月于浙江大学附属第二医院超声科检出的甲状腺峡部结节患者308例，共308个结节。所有患者均经手术病理或细针穿刺活检（FNAB）证实确诊，且均行超声检查。分别按照ACR TI-RADS和Kwak TI-RADS分类标准对308个峡部结节进行评分分类。应用多因素Logistic回归分析甲状腺峡部恶性结节的独立危险因素；将ACR TI-RADS、Kwak TI-RADS分类结果与病理结果进行比较分析；以FNAB和手术病理结果为金标准，构建ACR TI-RADS、Kwak TI-RADS诊断甲状腺峡部恶性结节的ROC曲线，分析两种分类方法的诊断效能。

308个甲状腺峡部结节中，恶性结节197个；良性结节111个。与良性结节相比，甲状腺峡部恶性结节的以下超声特征占比更高：实性（99.5% vs 77.5%）、低/极低回声（94.9% vs 46.8%）、纵横比＞1（21.3% vs 5.4%）、边缘分叶/不规则或甲状腺外侵犯（41.1% vs 7.2%）、点状强回声（43.2% vs 11.7%），差异均有统计学意义（P均＜0.05）。其中，实性（P=0.047）、低/极低回声（P＜0.001）、边缘分叶/不规则或甲状腺外侵犯（P=0.002）、点状强回声（P=0.007）是甲状腺峡部恶性结节的独立危险因素，而纵横比＞1并非恶性结节的独立危险因素（OR=2.683，P=0.050）。ACR TI-RADS 1~2类结节均为良性结节，4类结节中70.1%为恶性结节，5类结节中87.0%为恶性结节；Kwak TI-RADS 2~3类结节均为良性结节，4b、4c、5类结节中分别有70.0%、83.2%、100%为恶性结节。ACR TI-RADS与Kwak TI-RADS分类诊断甲状腺峡部恶性结节的ROC曲线下面积分别为0.823、0.820。ACR TI-RADS分类以TI-RADS 4类为截断值，其诊断敏感度、特异度、准确性、阳性预测值、阴性预测值分别为99.5%、53.1%、82.8%、79.0%、98.3%，Youden指数为0.526。Kwak TI-RADS分类以≥4b类为标准，其诊断恶性结节的敏感度、特异度、准确性、阳性预测值、阴性预测值分别为99.0%、53.1%、82.5%、78.9%、96.7%，Youden指数为0.521。

ACR TI-RADS与Kwak TI-RADS对甲状腺峡部结节的良恶性诊断均具有较高的价值，对于甲状腺峡部结节部分级别的恶性风险阈值以及管理建议，仍需进一步研究证实。

To evaluate the diagnostic value of American College of Radiology thyroid imaging report and data system (ACR TI-RADS) and 2011 Kwak TI-RADS in thyroid isthmic nodules.

This retrospective study included 308 patients with 308 thyroid isthmic nodules diagnosed pathologically by fine needle aspiration biopsy or surgery at department of ultrasound, the Second Affiliated Hospital of Zhejiang University School of Medicine from January 2015 to January 2020. All nodules were classified according to the ACR TI-RADS and Kwak TI-RADS classifications. Multivariate logistic regression analysis was performed to analyze the independent risk factors formalignant thyroid isthmic nodules.The ACR TI-RADS and Kwak TI-RADS classifications results for the nodules were compared with the pathological results, and the receiver operating characteristic (ROC) curves of the two methods were plotted to assess their diagnostic efficiency.

Of the 308 thyroid isthmic nodules, 197 were malignant and 111 were benign. Compared with benign thyroid nodules, malignant nodules located in the isthmus were more likely to have the following ultrasonic features: solid type (99.5% vs 77.5%), hypo-/very hypo-echogenicity (94.9% vs 46.8%), taller-than-wide shape (21.3% vs 5.4%), lobulated or irregular or extra-thyroidal extension (41.1% vs 7.2%), and punctate echogenic foci (43.2% vs 11.7%); the differences of these features were statistically significant between the two groups (P＜0.05). Solid type (P=0.047), hypo-/very hypo-echogenicity (P＜0.001), lobulated or irregular or extra-thyroidal extension (P=0.002), and punctate echogenic foci (P=0.007) were independent risk factors for malignancy, while taller-than-wide shape was not an independent risk factor for malignancy (OR=2.683, P=0.050). All the ACR TR 1-2 nodules were benign. The malignancy rates of ACR TR 4 and 5 were 70.1%and 87.0%, respectively. All the Kwak TR 2-3 nodules were benign. The malignancy rates of Kwak TR 4b, 4c, and 5 were 70.0%, 83.2%, and 100%, respectively. The areas under the ROC curves of the ACR TI-RADS and Kwak TI-RADS classifications were 0.823 and 0.820, respectively. The optimal cut-off values for diagnosing malignancy by ACR TI-RADS and Kwak TI-RADS were ≥TR4 (total score ≥4 points) and ≥TR4b, and at these points, the sensitivity, specificity, accuracy, positive predictive value, negative predictive value, and Youden index were 99.5% vs 99.0%, 53.1% vs 53.1%, 82.8% vs 82.5%, 79.0% vs 78.9%, 98.3% vs 96.7%, and 0.526 vs 0.523, respectively.

Both of the ACR TI-RADS and Kwak TI-RADS have high diagnostic valuein diagnosing thyroid isthmic nodules. Further studies of TI-RADS are needed to assess the threshold of malignant risk and management in isthmic nodules.