乳腺癌多模态超声评估新辅助化疗后腋窝淋巴结病理状态的研究

Evaluation of the pathological status of axillary lymph nodes after neoadjuvant chemotherapy based on multimodal ultrasound characteristics for breast cancer

目的应用多模态超声特征构建诊断模型,预测乳腺癌患者新辅助化疗(NAC)后腋窝淋巴结(ALN)的病理状态。方法选取2017年7月至2021年5月于哈尔滨医科大学附属第二医院及哈尔滨医科大学附属肿瘤医院就诊且经穿刺病理证实为ALN转移的女性乳腺癌患者204例,根据NAC后ALN手术病理结果分为两组:完全缓解(pCR)组和非完全缓解(npCR)组。回顾性分析其超声图像特征、免疫组化指标及血常规指标,进行两组间比较,筛选出两组间P<0.20的指标,基于逻辑回归算法构建乳腺癌患者NAC后ALN病理状态的预测模型,并绘制ROC曲线评估模型诊断效能。结果通过t检验、Mann-WhitneyU检验、χ2检验分析发现乳腺肿瘤大小、乳腺肿块血流阻力指数(RI)、弹性评分、ALN门结构、皮质最大厚度、淋巴结内血流分布、淋巴结血流RI及免疫组化检测指标雌激素受体(ER)、孕激素受体(PR)、人表皮生长因子受体2(HER-2)、Ki67分子表达水平在两组间比较P<0.20。基于多元逻辑回归分析得出乳腺肿瘤大小、ALN门结构、皮质最大厚度、淋巴结内血流分布、淋巴结血流RI、PR、HER-2分子表达水平是预测乳腺癌患者NAC后ALN病理状态的独立因素(P<0.05),并获得该预测模型的ROC曲线下面积为0.870(95%CI=0.819~0.922,P<0.05),敏感性为86.82%,特异性为70.67%。结论应用多模态超声特征及免疫组化指标构建预测乳腺癌患者NAC后ALN病理状态的模型,具有较好的诊断效能,可为临床治疗方案的制定以及预后的评估提供更多的客观依据。

Objective To construct a diagnostic model based on the multimodal ultrasound imaging,and to predict the axillary lymph node(ALN)status of breast cancer patients after neoadjuvant chemotherapy(NAC).Methods A total of 204 female breast cancer patients with ALN metastasis who had undergone puncture biopsy of aspiration in the Second Affiliated Hospital and Harbin Medical University Cancer Hospital between July 2017 to May 2021 were included.According to the pathological results of ALN surgery after NAC,the cases were divided into pathologically complete response(pCR)group and non-pCR group.The ultrasound images,immunohistochemistry and blood routine index were collected and compared between the two groups,the indexes whose P<0.02 were selected.In terms of logistic regression algorithm,a predictive model for the pathological state of axillary lymph nodes in breast cancer patients was established after NAC,and ROC curve was plotted to evaluate the performance of the model.Results The P values for comparison between the two groups of the breast tumor size,blood flow resistance index(RI),elasticity score,lymph hilum structure,maximum cortical thickness,blood flow distribution,blood flow RI,and immunohistochemical detection indicators including estrogen receptor(ER),progesterone receptor(PR),human epidermal growth factor 2(HER-2),Ki67 molecular expressions were<0.20 by t test,Mann-Whitney U test,andχ2 test analysis;in the multiple logistic regression analysis,tumor size,lymphatic hilum structure,maximum cortical thickness,lymph node blood flow distribution and blood flow resistance index,PR and HER-2 molecular expressions were the independent factors predicting the pathological status of axillary lymph nodes in breast cancer patients after NAC(P<0.05).The performance of the predictive model was 0.870(95%confidence interval:0.819-0.922,P<0.05),with sensitivity of 86.82%and specificity of 70.67%.Conclusions The model for predicting the pathological state of ALN in breast cancer patients after NAC using multi-modal ultrasound characteristic and immunohistochemical indexes achieves good diagnostic performance providing more objective evidence for the formulation of clinical treatment plans and prognostic evaluation.