MRI检查联合钼钯检查诊断非肿块型乳腺癌的临床价值研究

Clinical Value of MRI Combined with Mammography in the Diagnosis of Non-tumor Breast Cancer

背景非肿块型乳腺癌不具有典型的乳腺癌特征,在早期筛查中的检出率低于肿块型乳腺癌。目的探索MRI检查联合钼钯检查诊断非肿块型乳腺癌的价值。目的探索MRI检查联合钼钯检查诊断非肿块型乳腺癌的价值。方法选择2015年1月-2016年7月于河北医科大学第四医院行临床触诊、钼靶检查初诊为乳腺非肿块型病变的符合研究标准的女性患者82例作为研究对象。以术后病理检查结果为金标准分为恶性组和良性组。比较两组患者MRI检查和钼钯检查结果;分析基于BI-RADS分类的MRI检查和钼钯检查结果诊断非肿块型乳腺癌的灵敏度、特异度、符合率,MRI检查和钼钯检查各征象及参数单独诊断非肿块型乳腺癌的灵敏度、特异度、符合率。采用Logistic回归分析建立MRI检查和钼靶检查各征象及参数联合诊断非肿块型乳腺癌的拟合方程,并绘制其诊断非肿块型乳腺癌的受试者工作特征曲线(ROC曲线),计算ROC曲线下面积(AUC)、灵敏度、特异度。结果根据术后病理检查结果,53例恶性患者被纳入恶性组,29例良性患者被纳入良性组。MRI检查中,恶性组表观扩散系数(ADC)值低于良性组,强化方式中不均匀强化比例、动态增强曲线(TIC)类型中流出型高于良性组(P<0.05)。恶性组钼靶检查为集簇状微钙化比例高于良性组(P<0.05)。基于BI-RADS分类标准,MRI检查诊断非肿块型乳腺癌的灵敏度为81.13%(43/53),特异度为65.52%(19/29),符合率为75.61%(62/82);钼钯检查诊断非肿块型乳腺癌的灵敏度为66.04%(35/53),特异度为72.41%(21/29),符合率为68.29%(56/82)。MRI检查中,ADC值、TIC类型、强化方式的符合率均>70.0%(ADC值为76.30%,TIC类型为77.80%,强化方式为74.39%);钼钯检查征象中(局灶性不对称致密影、集簇状微钙化、腺体结构扭曲或紊乱、混合征象)仅微钙化征象的符合率>60.0%(70.73%)。联合诊断模型的拟合方程为Logit(P)=-6.369×ADC值+2.580×TIC类型+1.856×强化方式+1.927×集簇状微钙化征象。联合诊断模型诊断非肿块型乳腺癌的最佳截断值为0.708,AUC为0.919,灵敏度为84.91%,特异度为93.10%。结论MRI检查和钼靶检查诊断非肿块型乳腺癌均各有不足,而利用Logistic回归分析将MRI检查结果中的ADC值、TIC类型、强化方式及钼靶检查结果中的集簇状微钙化征象进行拟合,其联合应用可显著提高非肿块型乳腺癌的诊断价值。

Background Non-tumor breast cancer does not have typical characteristics of breast cancer,and the detection rate in its early screening is lower than that in mass-type breast cancer.Objective To explore the value of MRI combined with mammography in the diagnosis of non-tumor breast cancer.Methods From January 2015 to July 2016,82 female patients who met the clinical criteria for palpation and mammography in the Fourth Hospital of Hebei Medical University were diagnosed as non-tumor breast lesions,and they were recruited as research subjects and divided into the malignant group and the benign group according to postoperative pathological results as the gold standard.The results of MRI and mammography were compared between the two groups.The sensitivity,specificity and accuracy of MRI and mammography based on BIRADS classification were analyzed.Then,individual analysis was performed based on the signs and parameters of MRI and mammography.Logistic regression analysis was used to establish a fitted equation for the signs and parameters of MRI combined with mammography in the diagnosis of non-tumor breast cancer.The ROC curve for the diagnosis of non-tumor breast cancer was plotted and the area under the ROC curve(AUC),sensitivity,and specificity were calculated.Results According to the postoperative pathological results,53 patients were included in the malignant group and 29 patients were included in the benign group.For the MRI examination,the ADC value of the malignant group was lower than that of the benign group,while the outflow type in the time-to-intensity curve(TIC) type and the uneven enhancement ratio in the strengthening mode were higher than those in the benign group(P<0.05).The malignant group had a higher proportion of clustered microcalcification than the benign group(P<0.05).Based on the BI-RADS classification criteria,the sensitivity of MRI for the diagnosis of non-tumor breast cancer was 81.13%(43/53),the specificity was 65.52%(19/29),and the accuracy was 75.61%(62/82).The sensitivity of mammography was 66.04%(35/53),the specificity was 72.41%(21/29),and the accuracy 68.29%(56/82).In MRI examination,the accuracy of the ADC value,TIC type,and enhancement mode were greater than 70.0%(76.30%,77.80%,and 74.39%,respectively).In mammography examination signs(focal asymmetric density,clustered microcalcification,gland architectural distortion or disorder,and mixed signs),only the accuracy of clustered microcalcification was greater than 60.0%(70.73%).The fitted equation of the combined diagnosis model was Logit(P)=-6.369×ADC value+2.580×TIC type+ 1.856×enhancement mode +1.927×clustered microcalcification.The best cut-off value for the combined diagnosis model was 0.708,the AUC value was 0.919,the sensitivity was 84.91%,and the specificity was 93.10%.Conclusion Both MRI and mammography have different deficiencies in the diagnosis of non-tumor breast cancer.Logistic regression analysis is used to fit the ADC value,enhancement mode and TIC type in MRI and clustered microcalcification in mammography.This combined diagnosis model can significantly improve the diagnostic value of non-tumor breast cancer.