The Value Of Magnetic Resonance Imaging In Evaluating The Neoadjuvant Chemotherapy Of Breast Cancer Based On Tumor Heterogeneity

Objective:To explore the 3.0T magnetic resonance imaging（MRI）technology in evaluating and predicting the efficacy of neoadjuvant chemotherapy（NAC）for breast cancer an different stages of NAC:（1）evaluation value of MRI in different subtypes of breast cancer after NAC/preoperative（NAC₄）;（2）prediction value of NAC in Luminal type breast cancer at mid-NAC（NAC₂）;（3）The predicting value of diffusion kurtosis imaging（DKI）and diffusion-weighted imaging（DWI）for the efficacy of NAC in different subtypes of breast cancer before neoadjuvant chemotherapy（NAC₀）.Materials and Methods:（1）From January 2014 to January 2017,121 women with invasive malignant breast tumors were included in this study.All patients underwent 4cycles of NAC,and underwent DCE-MRI and DWI detection before and after NAC.The image was measured to obtain the parameter of tumor diameter（D）and the apparent diffusion coefficient（ADC）value.First to explore the evaluation value of tumor D base on the RECIST standard,and then combine tumor D and ADC value,to discuss whether the evaluation is better than the single indicator of tumor D.（2）Sixty-one patients who underwent NAC for pathologically confirmed breast cancer were enrolled in this retrospective study.The ADC values of different shrinkage patterns（concentric shrinkage,nest or dendritic shrinkage,and mixed shrinkage）for tumors shown by DWI were measured independently using three ROI placement methods（single-round,three-round,and freehand）.Multivariate logistic regression analysis was performed to identify the independent predictors of PCR.（3）Seventy-six patients who underwent NAC for pathologically confirmed breast cancer and 3.0T Philips multi-b-value diffusion kurtosis imaging（DKI）examination were enrolled in this prospective study.Statistical analyses were performed by using software（SPSS version 21.0）.T-test,Mann–Whitney U-test were used for statistical analysis,and the AUCs were applied to assess the ability of MD and MK to predict the PCR among different type of breast cancer.Multivariate logistic regression analysis was performed to identify the independent predictors of PCR.Result:（1）Evaluation results obtained by only utilizingΔD%were tested in conformity with the pathology results,and the Kappa values were 0.594,0.617,0.401,0.690 and0.800 in all patients,Luminal A type,Luminal B type,Her-2 type and triple negative type.When combined parameterΔD%withΔADC%,the Kappa values were 0.844,0.792,0.854,0.847 and 0.800.The sensitivity of the combined parameter to evaluate PCR was significantly increased in Luminal A and Luminal B type,with a slightly elevated in Her-2 type,and invariant in triple negative breast cancer.（2）The single-round method was optimal for concentrically shrinking tumors（AUC=0.970）;the freehand method was optimal for nest or dendritically shrinking tumors（AUC=0.714）;and the three-round method was optimal for mixed shrinking tumors（AUC=0.975）.Multivariate logistic analysis showed that estrogen receptor（ER）,ΔADC%and tumor diameter reduction（ΔD%）were independent factors in evaluating the PCR.（3）In the overall patient,the parameters MD and MK were statistically different between the PCR group and the non-PCR group（P=0.044,P=0.010）.The area under the ROC curves（AUC）for predicting the PCR of the two parameters are 0.634 and 0.675respectively.However,when the tumors were evaluated separately according to their subtypes,the effectiveness results differed and the parameters MD and MK were only different between the two groups in the Luminal B type of breast cancer（MD:P<0.001,AUC=0.877;MK:P=0.022,AUC=0.754）.According to the multivariate logistic analysis,ER expression（P<0.001;OR=1.395,95%CI:1.064-2.028）and MD（P<0.001;OR=1.021,95%CI:0.919-1.443）were the independent factors most predictive of the PCR for this type of breast cancer.Conclusion:（1）At NAC₄,ΔD%andΔADC%differed in assessing the efficacy of NAC among different types of breast cancer.ΔD%combined withΔADC%has the most significant increase in the evaluation of the Luminal cancers.Therefore,DWI plays an important role in the evaluation of the efficacy of this subtype of breast cancer.（2）At NAC₂,the methods for measuring ADC values vary across different shrinkage patterns of the luminal tumors.ΔADC%,ER andΔD%were independent factors for evaluating the PCR.（3）At NAC₀,the ADC value which is based on DWIhas certain limitations on the predicting of efficacy in any type of breast cancer.While the parameters MD and MK based on DKI have some predicting,but are limited to Luminal B type.