Ultrasonic Elastography Predict The Pathological Response To Neoadjuvant Chemotherapy In Patients With Breast Cancer And Preliminary Study On The Mechanism

Objective:Breast cancer is one of the malignant tumors that seriously threaten the health of women all over the world.Preoperative neoadjuvant chemotherapy?NAC?is the established breast-conserving therapeutic strategy for patients with breast cancers.The specific pathological response after NAC is a crucial factor for predicting the clinical outcome.However,clinical evidence does not clearly demonstrate an improvement in event-free survival and overall survival after NAC compared with conventional adjuvant therapy.A reliable predictive modality can be used to predict early response to NAC and to efficiently distinguish NAC responders from non-responders,which is helpful for selecting the ideal patient candidates for NAC before surgery for reducing unnecessary treatme nt and reduce the side reaction of chemotherapy drugs for patients with NAC resistance.Ultrasonic elastography has been widely used to qualitatively and quantitatively evaluate differences in lesion stiffness or elasticity.The present study was performed to compare the diagnostic performances of shear-wave elastography?SWE?and strain elastography?SE?for early prediction of the pathological response to NAC in breast cancer.Moreover,this study evaluated SWE combined with the K i-67 index as novel predictive modalities for the pathological response of invasive breast cancer to NAC.Finally,this study explored the initial mechanism that ultrasonic elastography predict the pathological response to NAC.Methods:The prospective study recruited 71eligible patients with core needle biopsy confirmation of invasive breast cancer,who underwent six cycles of NAC and subsequent surgical excision between June 2014 and December 2016.Tumor stiffness was assessed by the SE strain ratio?R?,the SWE maximum elastic ity(E_max),and the mean elasticity(E_mean),time points were one day before biopsy?time point t0,SWE elasticity E0,SE elasticity R0?,one day before next N AC cycle?time point t1-t5,SWE elasticity E1-E5,SE elasticity R1-R5?,and one day before surgery?t6,E6?.The relative changes in SWE parameters after the first and second NAC cycles were considered as the variables[?E?t1?,?E?t2?].The pathological response was classified according to the residual cancer burden?RCB?protocol.In order to screen out which factors would have predictive value,univariate analyses were used,statistical analysis was performed to compare the differences between Pathological complete response and non-pathological complete response using independent T-test for continuous variables including pathological and stiffness variables,and Kruskal-Wallis test would be chosen to analyze the differences if the continuous variables departured from normality.Pearson Chi-squared test or Fisher's exact test were performed for the comparison of categorical variables.The receiver operating characteristics?ROC?curve,the area under the ROC?AUC?with the 95%confidence intervals,and optimal cutoff were obtained of elasticity variables,Z test using the De Long method was used to analyze the difference among AUCs of each ultrasonic elastography parameter.The general linear models were used to quantify the relationship between ultrasonic elastography parameters and the RCB scores.In this sense,R² value of greater than0.3 served as a significant cutoff level.The combination of ultrasonic elastography parameters and immunohistochemical marker by using the predicted RCB scores?predRC B?and parallel test were designed using a multivariable linear regression model,which would have an increase predictive performance.The predictive diagnostic performances of SWE parameters,immunohistochemical marker and predRCB deduced by linear regression model were compared.MedCalc Statistical Software version 15.8?MedCalc Software bvba,Ostend,Belgium?were used for the statistical data analysis.A P value of less than 0.05 was regarded as significantly significant.According to the relationship between matrix stiffness of breast cancer evaluated by ultrasound elastography and pathological responses,the applicant propose a novel working hypothesis“high matrix stiffness-high interstitial fluid pressure-tissue hypoperfusion hypoxia-HIF-1?/TWIST1 activation”.Results:Almost all elastography parameters showed independent relationships with pathological responses.Elastography parametersafter second cycle[?E_mean?t2?,R2]showed best diagnostic performance.For predicting favorable response,although?E_mean?t2?showed the best?AUC=0.93?,both SWE parameters after one cycle and all SE parameters performed well with no significant difference from?E_mean?t2?.For predicting resistance,?E_mean?t2?was significantly superior to that of all SE parameters?AUC=0.92?.Significant differences were noted for the breast thickness and lesion depth between the two groups with discrepant results between SE and SWE respectively.It was moderate and statistically significant between some SWE stiffness variables,K i-67 and RCB score.The?E_mean?t2?and K i-67 showed significantly better diagnostic performance than other parameters in the predicting favorable responses and resistance for NAC,respectively.However,the correlation between?E_mean?t2?and K i-67 index was significantly weak?r=0.29?.A new predictive index pred RCB for a combination of?E_mean?t2?and K i-67 index by using the multivariable linear regression model was generated,pred RCB=4.449+3.617×?E_mean?t2?-0.026×Ki-67.Furthermore,the pred RCB showed better diagnostic performances than SWE parameters and Ki-67 alone.A novel working hypothesis"high matrix stiffness-high interstitial fluid pressure-tissue hypoperfusion hypoxia-HIF-1?/TWIST1 activation"was proposed.The correlation between the expression of HIF-1?and TWIST1 of breast cancer tissue was also identified in our study by immunofluorescence.There was also a significant relationship between matrix stiffness of breast cancer evaluated by SWE and HIF-1?/TWIST1 in breast cancer tissue of patients.Conclusion:Ultrasonic elastography can predict the pathological response to N AC in patients with breast cancer.SE and SWE showed similar performance in predicting favorable response after NAC;SWE was better than SE in predicting resistance.Discrepant results may between SE and SWE be due to the breast thickness and lesion depth.Our findings also highlight the potential utility for adding K i-67 to SWE parameters,which maybe improve the prediction power of SWE,facilitate personalizing the treatment of patients with breast cancer.The hypoperfusion induced by extracellular matrix with high stiffness may be mediated by activation HIF-1?/TWIST1 pathways through the induced NAC resistance to breast cancer.