Magnetic Resonance Imaging In Predicting The Pathological Features Of Intraductal Malignant Lesions Of Breast

BackgroundBreast cancer is the most common malignant tumor in women worldwide,and the number of new cases is increasing year by year.The common pathological types of breast cancer are invasive ductal carcinoma,ductal carcinoma in situ,and invasive carcinoma with ductal carcinoma in situ.The component of ductal carcinoma in situ is associated with marginal infiltration in breast conserving surgery and is an independent risk factor for local recurrence.According to the 2013 Breast Imaging Report and Data System Atlas,breast lesions were divided into punctate,mass and non-mass enhanced lesions.Most of the previous studies focused on breast cancer with mass enhancement.There were few studies on breast cancer with non-mass enhancement and the conclusion was controversial.Invasive ductal carcinoma was the most common mass lesion,and ductal carcinoma in situ was the most common non-mass lesion.Therefore,in this paper,the gold standard of pathology was used as a control to analyze the image features of MRI mass and non-mass enhancement lesions by conventional MRI sequence and functional imaging sequence of diffusion imaging,and to explore the predictive value of MRI for pathological information of three different pathological types.The paper is divided into two parts:1.A retrospective analysis was conducted on 110 cases of breast cancer diagnosed by surgery and puncture pathology in our hospital from September 2020 to December 2021,including 76 cases of mass lesions and 34 cases of non-mass lesions,and the correlation between their routine sequence MRI imaging features and pathology was analyzed.2.Patients with breast diseases meeting the exclusion criteria were prospectively and continuously collected.All patients received the preoperative routine MRI,DWI and Diffusion Kurtosis Imaging(DKI)sequences,and ADC,Mean Kurtosis(MK)value,Mean Diffusivity(MD)value;Pathological and immunohistochemical results of all patients were recorded,including histological grade,histological pathological type,expression of ER,PR,Her-2 and Ki-67,presence or not of vascular invasion and lymph node metastasis,and correlation between functional imaging parameters and pathology was analyzed.Materials and MethodsThe retrospective and prospective studies were combined.All patients received preoperative MRI scans,and the imaging features,pathology and immunohistochemical results of all patients were recorded.Data of patients meeting the exclusion and inclusion criteria were used in this study.SPSS software was used for statistical analysis,and Medcalc software was used for Region of interest(ROC)curve.P<0.05 was considered as statistically significant difference.The Bland-Altman method was used to test the repeatability of the parameters between different observers.T-test was used for measurement data consistent with normal distribution,Wilcoxon rank sum test was used for measurement data inconsistent with normal distribution.The count data were tested for homogeneity of variance and compared by Chi-square test.With age and MRI signs of patients as independent variables and pathological type,pathological grade,molecular classification,lymph node metastasis and vascular infiltration of breast cancer as dependent variables,logistic multivariate regression analysis was performed respectively.Then ROC curves of each pathological feature were made,then,ROC curves of each pathological feature were made to calculate the Area Under Curve(AUC),sensitivity and specificity..Spearman correlation analysis was performed between imaging parameters and pathological results to observe the predictive value of functional image parameters on pathological results.ResultsThe MRI features and pathological types of mass and non-mass enhanced lesi ons were compared,and it was found that there were significant differences in the number of lesions,TIC curve,lesion size,ADC value,pathological type and expres sion of ER,PR and Her-2 between them(P<0.05).There were no significant diff erences in fibroglandular histological classification,background enhancement charact eristics,histological classification,Ki-67 expression,molecular classification,lymph n ode and vascular metastasis between the two groups(P>0.05).In mass lesions,age and MRI features had good efficacy in predicting pathological type,histological grade,molecular typing,ER,PR,Her-2,Ki-67 expression,lymph node metastasis and vascular metastasis,with an AUC above 0.700.Among them,the prediction of lymph node metastasis,PR expression and vascular metastasis was the best,and the AUC was above 0.800.In non-mass enhanced lesions,age,MRI features in the diagnosis of pathological type,histological grade,molecular analysis,PR,Her-2,Ki-67 expression,lymph node metastasis and vascular metastasis were all good,and the AUC was greater than 0.700.Among them,the prediction efficiency of Her-2 expression was the best,with an AUC of 0.763(0.591-0.935).The MD value was statistically different between IDC and EIC(P=0.005).The MK value was statistically different between EIC and DCIS(P=0.012).The ADC,MK,and MD values were statistically different between DCIS and IDC(P<0.001).Comparing BL with IDC,ADC,MK and MD values were statistically different(P<0.001).The ADC and MD values were statistically different between DCIS and BL(P<0.001,P=0.020).There were significant differences in PR and Her-2 expression between IDC and DCIS(P<0.05).ADC values of the three lesions were positively correlated with pathological grade and Her-2 expression,and negatively correlated with PR expression.MK value was negatively correlated with pathological grade and Her-2 expression,and positively correlated with PR expression.MD was positively correlated with pathological grade and negatively correlated with PR expression.All P values were less than 0.05.ADC values showed high efficacy in predicting pathological classification,histological grading and molecular typing(AUC>0.800).MD value was effective in predicting pathological classification and histological grading(AUC>0.800).The value of MK was effective in predicting pathological classification and molecular typing(AUC>0.800).Conclusion1.The enhancement features of mass enhanced breast cancer and non-mass enhanced breast cancer can provide certain clinical value for predicting the pathological types and prognostic factors of breast cancer.2.DWI and DKI can help in the differential diagnosis of benign and malignant breast lesions,and provide clinical value for further prediction of three different pathologic types of IDC,DCIS and EIC,as well as immune prognostic factors,but their value in predicting histological grade,lymph node metastasis and vascular invasion of breast cancer is limited.