超声对具有神经内分泌特征乳腺癌的诊断价值

目的探讨超声图像对具有神经内分泌特征乳腺癌的诊断价值。方法选取2011-2021年间本院接受治疗的36例具有神经内分泌特征的乳腺癌患者为研究对象,回顾性分析其手术前的超声图像、临床及病理结果资料各项指标与神经内分泌特征乳腺癌的相关性。结果36例患者中,高分化神经内分泌瘤占13.9%,低分化神经内分泌癌占5.5%,伴神经内分泌特征的浸润性乳腺癌占80.6%。超声特征包括形态不规则(91.7%)、平行生长(83.3%)、低回声(94.4%)、边缘不光整(83.3%)、后方回声无变化(69.4%)、无钙化(77.8%)、血流信号Ⅱ-Ⅲ级(61.1%)。BI-RADS分类为4A类41.7%、4B类22.2%、4C类25%和5类11.1%。72.2%的患者已绝经,淋巴转移阳性率为25%。免疫组织化学检测显示ER、PR、HER-2、CgA和Syn的阳性率分别为94.4%、94.4%、11.1%、86.1%和100%,Ki-67增殖指数在2%~90%之间。ROC曲线显示变量的诊断效能较低(AUC:0.365~0.640),多因素Logistic回归分析未能筛选出独立判断因素。结论超声检查对具有神经内分泌特征乳腺癌有较高的检出率,但超声图像特征不能早期预测不同病理类型。

Objective To explore the diagnostic value of ultrasound imaging for breast cancer with neuroendocrine features.Methods A retrospective analysis was conducted on 36 patients with neuroendocrine-featured breast cancer treated at our hospital from 2011 to 2021.Preoperative ultrasound images,clinical,and pathological outcome data were analyzed for their correlation with neuroendocrine-featured breast cancer.Results Among the 36 patients,well-differentiated neuroendocrine tumors accounted for 13.9%,poorly differentiated neuroendocrine carcinomas for 5.5%,and invasive breast cancers with neuroendocrine features for 80.6%.Ultrasound characteristics included irregular shape(91.7%),parallel growth(83.3%),hypoechoic(94.4%),irregular margins(83.3%),no change in posterior echo(69.4%),no calcification(77.8%),and blood flow signals of grade II-III(61.1%).BI-RADS classifications were 4A(41.7%),4B(22.2%),4C(25%),and 5(11.1%).Postmenopausal status was observed in 72.2%of patients,and the positive rate of lymph node metastasis was 25%.Immunohistochemical testing showed positive rates of 94.4%for ER and PR,11.1%for HER-2,86.1%for CgA,and 100%for Syn.The Ki-67 proliferation index ranged from 2%to 90%.Receiver operating characteristic(ROC)curves indicated low diagnostic efficacy of the variables(AUC:0.365~0.640),and multivariate logistic regression analysis did not identify independent predictive factors.Conclusion Ultrasound examination has a high detection rate for breast cancer with neuroendocrine features,but ultrasound image characteristics cannot predict different pathological types at an early stage.