| Obstract:To explore the value of magnetic resonance dynamic enhancement semi-quantitative parameters combined with diffusion-weighted imaging ADC values in predicting molecular typing of breast cancer,in order to use non-invasive imaging to determine molecular typing and provide early evaluation for the treatment and prognosis of the disease.Methods:A total of 148 patients with breast cancer from July 2019 to July 2022 in the First Affiliated Hospital of Dali University were included in this study.All patients underwent preoperative dynamic contrast enhanced magnetic resonance imaging(DCE-MRI)and diffusion weighted imaging(DWI)to obtain semi-quantitative parameters[time to peak(Tmax),maximum enhancement(SImax%),fastest enhancement rate(Rmax),elution score(washout)]and ADC values.All patients underwent pathological and immunohistochemical examination after puncture or operation to obtain the pathological type of tumor,axillary lymph node metastasis and the expression of biological markers(ER,PR,HER2 and Ki-67).According to the biological markers,breast cancer was divided into Luminal A type(49 cases),Luminal B type(64 cases),HER2 overexpression type(12 cases)and triple negative type(23 cases).Pathological types include eczematoid carcinoma,intraductal papilloma,intraductal papilloma with ductal carcinoma in situ,fibroadenoma,malignant phyllodes tumor,mucinous carcinoma,invasive lobular carcinoma,invasive carcinoma,invasive ductal carcinoma.The lesion size,lesion morphology and TIC type were obtained from the imaging data.Fisher’s exact probability method was used to compare the age,lesion size,lesion morphology,axillary lymph node metastasis and TIC type of the four molecular types;The semi-quantitative parameters of four groups of molecular typing were calculated and compared by K independent sample nonparametric test,and the apparent diffusion coefficient(ADC)was compared by scanning with b value=800s/mm~2;The receiver operating characteristic curve(ROC)was used to analyze the diagnostic value of DCE-MRI,DWI and their combination in molecular typing of breast cancer.Results:There were statistically significant differences in lesion size,lesion morphology and TIC among the four molecular subtypes of breast cancer patients(P<0.05);There were significant differences in Tmax,SImax%,Rmax and ADC among the four molecular subtypes(P<0.05).The sensitivity of Tmax,SImax%,Rmax,ADC and their combination in the diagnosis of Luminal A breast cancer was 81.63%,83.67%,93.88%,93.88%and 81.63%,respectively,and the specificity was 43.43%,49.49%,31.31%,18.18%and 49.49%,respectively;The sensitivity of Tmax,SImax%,Rmax,ADC and their combination in the diagnosis of Luminal B breast cancer was 75.00%,48.44%,25.00%,42.19%and 50.00%,respectively,and the specificity was 36.90%,76.19%,91.67%,85.71%and 82.14%,respectively;The sensitivity of Tmax,SImax%,Rmax,ADC and their combination in the diagnosis of HER2overexpressing breast cancer was 41.67%,50.00%,83.33%,100.00%and 83.33%,respectively,and the specificity was 79.41%,32.35%,43.38%,53.68%and 58.82%,respectively;The sensitivity of Tmax,SImax%,Rmax,ADC and their combination in the diagnosis of TNBC was 30.43%,78.26%,56.52%,69.57%and 56.52%,respectively,and the specificity was 97.60%,36.80%,67.20%,63.20%and 86.40%,respectively.For Luminal A and non-Luminal A breast cancer,the area under the curve(AUC)of Tmax,,SImax%,Rmax,ADC and their combination were 0.624,0.655,0.635,0.505 and 0.663,respectively,and the combination was higher than the single parameter ADC(P<0.05);For Luminal B and non-Luminal B breast cancer,the AUC of Tmax,SImax%,Rmax,ADC and their combination were 0.524,0.616,0.536,0.660 and 0.664,respectively,which were higher than those of Tmax and Rmax(P<0.05);For HER2 overexpressed and non-HER2overexpressed breast cancer,Tmax,SImax%,Rmax,ADC and their combined AUC were0.559,0.509,0.609,0.697 and 0.710,respectively,and the combined AUC was higher than SImax%(P<0.05);For triple negative and non-triple negative breast cancer,Tmax,SImax%,Rmax,ADC and the combined AUC were 0.630,0.539,0.597,0.679 and 0.753,respectively,and the combined AUC was higher than the single parameter SImax%(P<0.05).Conclusion:There were differences in lesion size,lesion morphology and TIC type among the four molecular subtypes of breast cancer patients.The morphology of Luminal B was mostly lobulated,and the morphology of the other three molecular subtypes was more irregular.The TIC type of Luminal B was mostly type I,and the TIC type of the other three molecular subtypes was more common in type III;The semi-quantitative parameters(Tmax,SImax%and Rmax)of DCE-MRI were different among the four molecular types of breast cancer,the SImax%value of Luminal A type was higher than that of Luminal B type;The ADC values of four molecular subtypes of breast cancer were different,the ADC value of Luminal B subtype was lower than that of HER2 overexpression and triple negative subtype;DCE-MRI and DWI have certain value in predicting breast cancer,but the combined diagnosis has better diagnostic efficiency than single diagnosis. |