多参数MRI联合免疫炎症指标预测乳腺癌腋窝淋巴结转移的价值研究

Value of multi-parameter MRI combined with immune inflammatory markers in predicting axillary lymph node metastasis of breast cancer

目的探讨多参数MRI联合免疫炎症指标预测乳腺癌腋窝淋巴结转移(axillary lymph node metastasis,ALNM)的价值。材料与方法回顾性分析52例经手术病理证实为乳腺癌的患者临床资料,根据病理结果将其分为淋巴结转移组与无转移组,单因素分析评估两组患者的临床病理特征、免疫炎症指标及多参数MRI特征与ALNM的关系。采用多因素logistic回归筛选临床和MRI危险因素建立临床预测模型、MRI预测模型及联合模型。Spearman等级相关方法分析免疫炎症指标与免疫组化因子、多参数MRI特征之间的相关性。通过绘制受试者工作特征(receiver operating characteristic,ROC)曲线和校准曲线评估模型效能,Delong检验和决策曲线(decision curve analysis,DCA)比较和验证不同模型的预测性能。结果logistic回归分析显示乳腺癌淋巴结转移组与无转移组间的Ki-67表达、血小板/淋巴细胞比值(platelet-lymphocyte ratio,PLR)、肿瘤长径、瘤周最大表观扩散系数(peritumoral maximum apparent diffusion coefficient,ADCpmax)、瘤周-肿瘤ADC比值(ratio of peritumoral-tumor ADC,ADCratio)及MRI淋巴结特征具有统计学意义(P<0.05);PLR与ADCpmax、ADCratio呈正相关(P<0.05);临床预测模型(Ki-67+PLR)曲线下面积(area under the curve,AUC)为0.722,多参数MRI预测模型(肿瘤长径+ADCpmax+ADCratio+MRI淋巴结特征)AUC为0.898,联合预测模型AUC为0.914。联合预测模型的临床价值高于临床预测模型。结论多参数MRI联合免疫炎症指标PLR可用于术前无创性预测乳腺癌患者腋窝淋巴结状态,为临床诊断及预后评估提供参考。

Objective:To investigate the value of multi-parameter MRI combined with immune inflammatory markers in axillary lymph node metastasis(ALNM)of breast cancer.Materials and Methods:In this retrospective analysis,52 breast cancer patients were divided into lymph node metastasis group and non-metastasis group according to the pathological results.The relationship between clinical,pathological,immune inflammatory markers,multi-parameter MRI features and axillary lymph node metastasis was evaluated by univariate analysis.Multivariate logistic regression was used to screen clinical and MRI risk factors to establish clinical prediction model,MRI prediction model and combined model.The correlation between immune inflammatory markers,immunohistochemical factor expression and multi-parameter MRI features were analyzed by spearman rank correlation analysis.Evaluate model effectiveness by drawing receiver operating characteristic(ROC)curve and calibration curve.The predictive performance of different models was compared and verified by the Delong test and decision curve analysis(DCA).Results:Logistic regression analysis showed that Ki-67 expression,platelet-lymphocyte ratio(PLR),tumor size,the peritumoral maximum apparent diffusion coefficient(ADCpmax),the ratio of peritumoral tumor ADC(ADCratio)and MRI lymph node characteristics were statistically significant(P<0.05).PLR was positively correlated with ADCpmax and ADCratio(P<0.05).The area under the curve(AUC)of the clinical prediction model(Ki-67+PLR)was 0.722,the AUC of the multi-parameter MRI prediction model(tumor length+ADCpmax+ADCratio+MRI lymph node characteristics)was 0.898,and the AUC of the combined prediction model was 0.914.DCA showed that the clinical value of the combined model was higher than that of the clinical prediction model.Conclusions:Multi-parameter MRI combined with immune inflammatory index PLR can be used to predict the status of axillary lymph nodes in breast cancer patients non-invasively before surgery,and provide a reference for clinical diagnosis and prognosis evaluation.