1、临床超声医学杂志2023年3月第25卷第3期J Clin Ultrasound in Med,March 2023,Vol.25,No.3 临床研究 不同TI-RADS指南预测甲状腺囊实性结节恶性风险分层价值的对比研究刘淑刘晏均马卓群赵艳茹魏炜施秉银摘要目的比较 ACR TI-RADS、ATA 指南及 KSThR 指南预测甲状腺囊实性结节恶性风险分层的价值。方法回顾性分析经病理证实的293个甲状腺囊实性结节(良性210个,恶性83个)的超声表现特征,应用不同TI-RADS指南对甲状腺囊实性结节进行恶性风险分层,多因素Logistic回归筛选甲状腺囊实性恶性结节的危险因素;绘制受试者工作特征(R
2、OC)曲线分析不同TI-RADS对甲状腺囊实性结节良恶性的预测效能。结果甲状腺囊实性良恶性结节大小、结构成分及实性成分为低回声、实性成分为偏心位、偏心位实性成分与囊壁成锐角、微钙化、边界不规则占比比较差异均有统计学意义(均P0.05)。ACR TI-RADS将良性结节分为极低危135个、低危46个、中危27个、高危2个,恶性结节分为极低危9个、低危13个、中危47个、高危14个;ATA指南将良性结节分为极低危158个、低危38个、高危14个,恶性结节分为极低危22个、低危8个、高危53个;KSThR指南将良性结节分为中危14个、低危187个、良性9个,恶性结节分为中危53个、低危28个、良性2
3、个。多因素Logistic回归分析显示,实性成分为低回声、微钙化、边界不规则均为甲状腺囊实性恶性结节的危险因素(OR=6.684、2.459、2.587,均P0.05)。ROC曲线分析显示,ACR TI-RADS、ATA指南、KSThR指南预测甲状腺囊实性结节良恶性的曲线下面积分别为0.843(95%可信区间:0.7910.895)、0.797(95%可信区间:0.7330.862)、0.785(95%可信区间:0.7170.852)。结论ACR TI-RADS预测甲状腺囊实性结节恶性风险分层的价值优于ATA指南及KSThR指南。关键词超声检查;TI-RADS;ATA指南;KSThR指南;甲状
4、腺囊实性结节,良恶性中图法分类号R445.1;R736.1文献标识码 AValue of different TI-RADS guidelines in predicting the risk stratification ofpartially cystic thyroid nodules:a comparative studyLIU Shu,LIU Yanjun,MA Zhuoqun,ZHAO Yanru,WEI Wei,SHI BingyinDepartment of Endocrinology,the First Affiliated Hospital of Xi an Jiaoton
5、g University,Xi an 710061,ChinaABSTRACTObjectiveTo compare the value of ACR TI-RADS,ATA guideline and KSThR guideline in predictingthe risk stratification of partially cystic thyroid nodules.MethodsThe sonographic features of 293 partially cystic thyroidnodules(210 benign and 83 malignant)confirmed
6、by pathology were retrospectively analyzed,different TI-RADS were used inthe risk stratification of partially cystic thyroid nodules,and the risk factors of malignant partially cystic thyroid nodules werescreened by multivariate Logistic regression.Receiver operating characteristic(ROC)curve was dra
7、wn to analyze the predictiveefficacy of different TI-RADS for benign and malignant partially cystic thyroid nodules.ResultsThere were statisticallysignificant differences between benign and malignant nodules in size,solid structure and the proportion of solid part presentedhypoecho,solid part presen
8、ted eccentric growth,acute angle of cystic wall,microcalcification and irregular boundary(all P0.05).According to ACR TI-RADS,benign nodules were divided into 135 very low-risk,46 low-risk,27 intermediate-risk,and 2 high-risk,while the malignant nodules were divided into 9 very low-risk,13 low-risk,
9、47 intermediate-risk,and 14 high-risk.Accordingto ATA guideline,benign nodules were divided into 158 very low-risk,38 low-risk,and 14 high-risk,while the malignant noduleswere divided into 22 very low-risk,8 low-risk,and 53 high-risk.According to KSThR guideline,benign nodules were divided into14 in
10、termediate-risk,187 low-risk,and 9 benign,while the malignant nodules were divided into 53 intermediate-risk,28 low-riskand 2 benign.Multivariate Logistic regression analysis showed that solid part presented hypoecho,microcalcification and irregularboundary were all risk factors for malignant partia
11、lly cystic thyroid nodules(OR=6.684,2.459,2.587,all P1、边缘不规则、微钙化、分叶状及甲状腺外侵犯等已广泛用于预测高危结节8-10。本研究旨在探讨不同TI-RADS指南在预测甲状腺囊实性结节恶性风险分层中的应用价值。资料与方法一、研究对象选取2016年1月至2018年12月于我院经手术病理或经超声引导下甲状腺细针穿刺活检(fine needleaspiration,FNA)证实的甲状腺囊实性结节患者288例,男76例,女212例,年龄1080岁,平均(46.313.8)岁;283例患者为单发结节,5例患者均为2个结节,共纳入293个结节;包括
12、恶性结节83个(均为甲状腺乳头状癌),良性结节210个。排除FNA仅为囊液或细胞学病理诊断为不确定性且未行手术治疗者。本研究经我院医学伦理委员会批准,所有患者均知情同意。二、仪器与方法1.仪器:使用 GE Logiq E 9、GE Logiq E 8 及东芝NEMZO 17彩色多普勒超声诊断仪,线阵探头,频率513 MHz。2.方法:患者取仰卧位,使用仪器预设的甲状腺检查模式多切面扫查甲状腺及颈部淋巴结。由两名经验丰富的甲状腺超声医师对超声图像进行分析,分别描述以下超声特征:结节大小(最大径)、结构成分(囊性成分50%定义为囊性为主;囊性成分1、边缘钙化中断、软组织突出钙化外、甲状腺外侵犯)、
13、低危(结节实性成分为偏心位,不伴有恶性特征)和极低危(结节实性成分为非偏心位,不伴有恶性特征),其中极低危和低危结节判为良性,高危结节判为恶性7;KSThR指南将囊实性结节分为中危(具有以下任一恶性特征:结节伴微钙化、纵横比1、边缘不规则)、低危(结节不伴任一恶性特征)和良性(囊实性结节伴彗星尾征),其中低危和良性结节判为良性,中危结节判为恶性8。三、统计学处理应用SPSS 20.0统计软件,计量资料以xs表示,采用 t 检验;计数资料以频数或率表示,采用 2检验或Fisher精确检验。应用多因素Logistic回归分析甲状腺囊实性结节良恶性的影响因素;绘制受试者工作特征(ROC)曲线分析不同
14、TI-RADS指南对甲状腺囊实性结节良恶性的预测价值。P0.05为差异有统计学意义。结果一、良恶性结节患者一般资料比较恶性结节患者平均年龄为(40.411.7)岁,良性结节患者平均年龄为(48.613.9)岁,差异有统计学意义(t=4.784,P0.001);恶性结节患者男性占比较高(45/83,54.2%),良性结节患者女性占比较高(149/210,71.0%),差异有统计学意义(2=16.321,P0.001);两组其余一般资料比较差异均无统计学意义。二、良恶性结节超声图像特征比较良恶性结节大小、结构成分及实性成分为低回声、实性成分为偏心位、偏心位实性成分与囊壁成锐角、微钙化、边界不规则占
15、比比较差异均有统计学意义(均 P1 占比比较差异无统计学意义。见表1和图1,2。malignant partially cystic thyroid nodules were 0.843(95%confidence interval:0.7910.895),0.797(95%confidence interval:0.7330.862)and 0.785(95%confidence interval:0.7170.852),respectively.ConclusionACR TI-RADS has higher value inpredicting the risk stratificati
16、on of partially cystic thyroid nodules than ATA guideline and KSThR guideline.KEY WORDSUltrasonography;TI-RADS;ATA guideline;KSThR guideline;Partially cystic thyroid nodule,benignand malignant 162临床超声医学杂志2023年3月第25卷第3期J Clin Ultrasound in Med,March 2023,Vol.25,No.3表1良恶性结节超声图像特征比较病理结果恶性(83)良性(210)t/2值P值大小(cm)2.61.13.11.23.3110.001结构成分(个)实性为主7013710.4650.001囊性为主1373实性成分为偏心位(个)是29446.2210.013否54166偏心位实性成分与囊壁成锐角(个)是16125.7540.016否1332实性成分回声类型(个)高/等回声2616659.9770.001低/极低回声5744病理结果恶性(83)良性(210)t/2值P值边