目的： 在中国人群中行单中心试验来验证国际卵巢肿瘤分析组织（International Ovarian Tumour Analysis,IOTA）简单法（simple rules,SR）在术前鉴别卵巢肿瘤良恶性的价值。目的： 选取2017年5月至12月上海交通大学医学院附属瑞金医院妇产科收治的卵巢肿块患者278例,共278个肿块,术前予行超声检查,并统计SR给出的拟诊。拟诊结果包括良性（175例）、恶性（交界性归为恶性,75例）、不确定（37例）。以病理学结果为金标准,计算SR鉴别卵巢肿瘤良恶性灵敏度、特异度和受试者操作特征（receiver operator characteristic,ROC）曲线下面积。结果： 术后病理提示,本组良性病例为203例,恶性病例为75例。SR能正确检出86.2%（175/203）的良性肿瘤,正确检出68.0%（51/75）的恶性肿瘤,15例为漏诊或误诊,37例未能给出诊断。在SR的超声特征中,M4（不规则多房囊实性肿块,最大径≥100 mm）预测恶性肿瘤的准确率最高（90.5%）;B3（声影）对良性肿瘤的预测价值最高（准确率为100.0%）。去除不确定的37例病例,在241例患者中,SR的诊断恶性卵巢肿瘤的灵敏度为86.4%、特异度为96.2%。在278例患者中,如将获得不确定SR结论的肿块归类为恶性,SR诊断恶性卵巢肿瘤的灵敏度为89.3%,特异度为86.2%,ROC曲线下面积0.88;将不确定肿块归类为良性时,诊断恶性卵巢肿瘤的灵敏度为68.0%,特异度为96.6%,ROC曲线下面积0.82,2种方法间差异有统计学（P=0.04）。8例病理检查结果为恶性而超声检查提示良性,回顾总结超声图像,发现当囊性肿块内壁不光滑,附壁见乳头状突起,或囊性肿块直径大于10 cm时,需考虑肿块有恶性的可能性。7例病理检查结果为良性而超声检查提示恶性,回顾超声图像,发现肿块实性部分存在血供,血流分级2~3级时,需考虑肿块有良性的可能性。结论： SR使用简单、便捷,适于中国人群应用;将SR不确定分类结果计入恶性,此时诊断恶性卵巢肿瘤灵敏度较高。

Objective: To validate the value of IOTA simple rules (SR) in differential diagnosis of malignant and benign ovarian tumors preoperatively in a single center in Chinese population. Methods: A total of 278 patients admitted for ova-rian tumors surgery in Department of Obstetrics and Gynecology, Ruijin Hospital were enrolled. All patients underwent ultrasound examination preoperatively, and 278 ovarian tumors were detected and classified as malignant (including 75 cases of borderline), benign lesions (175 cases) and uncertain lesions (37cases) using SR. Compared with the results of postoperative pathology examination, the sensitivity, specificity of SR in diagnosing malignant tumor as well as the area under receiver operating characteristic (ROC) curve were calculated. Results: The postoperative pathology examination showed that there were 203 benign tumors and 75 malignant ones. It indicated that SR identified 86.2% benign tumors (175 out of 203) and 68.0% malignant tumors (51 cases out of 75), 15 cases were missed or misdiagnosed and 37 cases got uncertain results. Among ultrasonic features of SR, M4 (irregular multilocular-solid tumor with largest diameter ≥100 mm) had the highest predictive value (90.5%) for malignant tumors, while B3 (acoustic shadow) had the highest predictive value (100.0%) for benign tumors. In 241 patients (excluding uncertain cases), the sensitivity and specificity of SR in diagnosing ovarian tumors was 86.4%, 96.2%, respectively. In 278 patients, as the uncertain cases diagnosed with SR were classified to be malignant tumors, the sensitivity was 89.3%, specificity was 86.2%, and AUC was 0.88. However, as uncertain cases were classified to be benign ovarian, the sensitivity and specificity of SR in diagnosing tumors was 68.0% and 96.6%, and the AUC was 0.82. The efficacy between the two methods was different(P=0.04). Eight benign cases were misdiagnosed as malignancy, and the review showed that the possibility of malignancy should be considered when the inner wall of the cystic mass was not smooth, papillary processes attached to the wall, or the size of cystic mass was more than 10 cm. Seven malignant cases were misdiagnosed as benign ones. It indicated that benign mass should be considered when the solid part had insufficient blood supply or even no blood flow on ultrasound imaging. Conclusions: SR is suitable to be applied to Chinese female in diagnosing ovarian tumor. IOTA SR may have better efficacy when the uncertain cases diagnosed with SR are classified as malignant tumors