列线图模型对浸润性乳腺癌伴腋窝淋巴结转移的预测价值

Nomogram for Predicting Invasive Breast Cancer with Axillary Lymph Node Metastasis

目的探讨列线图模型对浸润性乳腺癌伴腋窝淋巴结转移的预测价值。资料与方法回顾性分析2020年9月—2022年3月宁夏医科大学总医院疑诊乳腺癌患者122例,根据有无腋窝淋巴结转移分为转移组57例和非转移组65例。所有病灶经手术病理证实。两组接受合成磁共振、动态对比增强磁共振成像(DCE-MRI)及扩散加权成像扫描,测量乳腺病灶合成磁共振参数[T1、T2、质子密度(PD)]、时间-信号强度(TIC曲线)、表观扩散系数(ADC)。比较两组参数差异,并筛选出浸润性乳腺癌伴腋窝淋巴结转移的独立风险因素。结果Logistic回归结果显示Ki-67(OR=2.971,95%CI 1.306~6.762,P=0.009)、病灶大小(OR=1.652,95%CI 1.067~2.556,P=0.024)、ADCratio(OR=1.685,95%CI 1.014~2.801,P=0.044)、T2ratio(OR=3.015,95%CI 1.433~6.340,P=0.003)、PDratio(OR=2.782,95%CI 1.471~5.262,P=0.002)是浸润性乳腺癌伴腋窝淋巴结转移的独立风险因素。5种模型比较显示逻辑回归模型效能最优,曲线下面积为0.729(95%CI 0.621~0.789),准确度、特异度、敏感度分别为70.65%、62.79%、77.55%。对列线图模型准确性进行检验,得出C-index=0.844,即本次建立的列线图模型准确性良好,其截断风险为0.468,截断分数为143.50,当总分大于143.50时,发生腋窝淋巴结转移的风险将高于46.8%。结论列线图模型对浸润性乳腺癌伴腋窝淋巴结转移具有较好的预测能力。

Purpose To explore the predictive value of nomogram model for invasive breast cancer with axillary lymph node metastasis.Materials and Methods Retrospective analysis was made on 122 patients suspected to be breast cancer in the General Hospital of Ningxia Medical University from September 2020 to March 2022.According to whether there was axillary lymph node metastasis,all subjects were divided into 57 patients in the metastasis group and 65 patients in the non-metastasis group.All lesions were pathologically confirmed by surgery.The two groups received synthesis of magnetic resonance imaging(syMRI),dynamic contrast enhancement magnetic resonance imaging(DCE-MRI)and diffusion weighted imaging(DWI)scans.The syMRI parameters[including T1,T2,proton density(PD)],DCE-MRI time signal intensity curve,apparent diffusion coefficient(ADC)value of breast lesions were measured.Compared the difference of parameters between the two groups,and screened the independent risk factors of invasive breast cancer with axillary lymph node metastasis.Results Logistic regression results showed that Ki-67(OR=2.971,95%CI 1.306-6.762,P=0.009),lesion size(OR=1.652,95%CI 1.067-2.556,P=0.024),ADCratio(OR=1.685,95%CI 1.014-2.801,P=0.044),T2ratio(OR=3.015,95%CI 1.433-6.340,P=0.003),PDratio(OR=2.782,95%CI 1.471-5.262,P=0.002)were independent risk factors for invasive breast cancer with axillary lymph node metastasis.The comparison of the five models showed that the Logistic regression model had the best performance,with the area under curve of 0.729(95%CI 0.621-0.789),the accuracy,specificity and sensitivity were 70.65%,62.79%and 77.55%,respectively.The accuracy of the nomogram model was tested,and C-index=0.844,the accuracy of the nomogram model established was good,cut-off risk was 0.468,and the cut-off score was 143.50,which means that when the total score exceeds 143.50,the risk of axillary lymph node metastasis would be higher than 46.8%.Conclusion Nomogram model has a good predictive ability for invasive breast cancer patients with axillary lymph node metastasis.