甲状腺良恶性结节超声成像特征分析

Ultrasound imaging characteristics analysis of benign and malignant thyroid nodules

目的分析甲状腺结节的超声影像特征及其在预测恶性肿瘤风险中的重要作用。方法回顾性分析武汉大学人民医院2012年6月~2014年9月接受甲状腺超声检查并行手术切除治疗的甲状腺结节患者933例,术中共切除结节1362个,依据组织病理学检查结果将切除结节分为良性结节组(n=1022)和甲状腺癌组(n=340),分析各组各结节的超声检查结果并比较组间差异,Logistic回归分析筛查甲状腺癌的危险因素。结果良性结节组与甲状腺癌组在肿瘤大小、边界、钙化类型、纵横比、血流信号、回声声像及颈部淋巴结是否肿大等影像特征方面比较差异有统计学意义(P<0.05)。对低回声结节有以下4个超声影像特征:微小钙化(OR=27.954,95%CI=18.119~43.128,P<0.01),中央区血管(OR=5.841,95%CI=4.162~8.197,P<0.01),边界不清或有毛刺(OR=3.034,95%CI=2.227~4.135,P<0.01)和颈部淋巴结肿大(OR=22.981,95%CI=13.045~40.483,P<0.01)均为甲状腺癌的独立危险因素。如果没有上述异常的超声影像特征癌症的分级为Ⅰ级(风险<20%);如果有微小钙化之外的任何一个影像特征则分级为Ⅱ级(风险20%~<50%);如果具有两个上述影像特征或只有微小钙化则分级为Ⅲ级(风险为50%~80%);如果具备3个及以上的典型超声影像特征,则分级为Ⅳ级(风险>80%)。结论甲状腺超声影像可以用来明确患者是否具有高危风险,从而减少低危患者的过度诊疗。

Objective To analyze ultrasound imaging characteristics of thyroid nodules with the relative importance of these features in predicting risk for malignancy. Methods 933 patients underwent thyroid ultrasound imaging and thyroid nodule resection from June 2012 to September 2014 in Renmin Hosptial of Wuhan University were retrospectively analyzed. 1362 resected nodules were divided into benign group （n=1022） and thyroid cancer group （n=340） according to the histopathological examination results. Ultrasound results of each nodule were analyzed, the difference between two groups was compared. Risk factors of thyroid cancer screening were analyzed by Logistic regression. Results Ultrasound nodule characteristics of tumor size, boundary, calcification type, aspect ratio, blood flow signal, echo image and cervical lymph node enlargement and other imaging characteristics between benign group and thyroid cancer group were compared, with statistical differences （P 〈 0.05）. For hypoechoic nodules following four ultrasound nodule characteristics were considered to be the only findings associated with the risk of thyroid cancer： microcalcifications （OR= 27.954, 95%CI=18.119-43.128, P 〈 0.01）, eentral vascularity （OR=5.841, 95%CI=4.162-8.197, P 〈 0.01）, ill defined or spiculated margins （OR=3.034, 95%CI=2.227-4.135, P 〈 0.01） and cervical lymphadenopathy （OR=22.981, 95%CI= 13.045-40.483, P 〈 0.01）. Grade Ⅰ was defined as none of abnormal ultrasound imaging characteristic and the risk of cancer was less than 20%. Grade Ⅱ was defined as 1 characteristic was used as an indication for ultrasound except microcaleifications, the risk of caneer was 20%-〈50%.Grade Ⅲ was defined as 2 characteristics were required for ultrasound imaging included only had mierocalcifications, the risk of cancer was 50%-80%. And grade Ⅳ was defined as 3 or more suggestive ultrasound characteristics were contained, the risk of cancer was more than 80%. Conclusion Thyroid ultrasound imaging can be used to identify patients who have a low risk of cancer to reduce unnecessary and excessive treatment and biopsy.