| Objective To analyse the relationship between the sonograms of conventional ultrasound(US)and real-time shear wave elastrography(SWE)in breast cancer lesion and metastasis of axillary lymph nodes,to evaluate the diagnostic efficiency of US and SWE to metastasis of axillary lymph nodes(ALN),and to analyse the relationship between the sonograms of US and SWE and p53 and TopoⅡ expression.To investigate the accuracy of prognosis of lesion’s US and SWE in breast cancer,in order to improve the evaluation value of ultrasound in breast cancer.Methods According to inclusion and exclusion criteria of the study,select 451 cases breast cancer patients(466 lesions)with US and SWE in our hospital from January 2015 to October 2017.In 451 cases,there are 327 cases(342 lesions)with US,SWE and results of postoperative pathology.According to the diagnostic results of ALN-US,they are divided into 2 groups: ALN-US(-)group: 162 cases(167 lesions)ALN-US(+)group: 165cases(175 lesions),Includes: 75 cases(82 lesions)uspected malignant likeing 90 cases(93 lesions)malignant likeing According to the results of pathological diagnosis of ALN,they are divided into 2 groups: ALN Benign group: 174 cases(176 lesions)ALN Malignant group: 153 cases(166 lesions)In 451 cases,There are 253 cases(256 lesions)with the complete information of expression of p53 and TopoⅡ.According to the expression of p53,they were divided into 2 groups: p53(-)group: 117 cases(119 lesions)p53(+)group: 136 cases(139 lesions)According to the expression of TopoⅡ,they are divided into 2 groups: TopoⅡ(-)group: 93 cases(94 lesions)TopoⅡ(+)group: 160 cases(162 lesions)Firstly,analyse the relationship between the sonograms feature of lesion’s US and the diagnosis performance of quantitative and qualitative features of lesion’s SWE to postoperative pathological diagnosis of axillary lymph nodes in breast cancer respectively,and then select out significative features into logistic regression analysis to establish logistic regression model which was used to investigate metastasis of ALN in breast cancer with US and SWE jointly,and then combine the model with ALN-US.Compare AUC of the ROC curve,sensitivity,specificity and accuracy of every method.Respectively compare the sonograms feature of US and the diagnosis performance of quantitative and qualitative features of SWE in breast cancer of p53(-)with p53(+)and TopoⅡ(-)with TopoⅡ(+).Analyse the interdependency between each sonograms feature and the state p53 and TopoⅡ.The value in prognosis of lesion’s US and SWE in breast cancer was assessed.Results 1.Mean maximum diameter(2.59±1.13)cm in TopoⅡ(+)is bigger than(2.43±1.25)cm in TopoⅡ(-)(P<0.05),the correlation coefficient(r)= 0.135,P=0.030.2.In US,comparing to the breast lesions without irregular shape,boundary fuzziness,spiculate margin,peripheral hyperechoic zone or hyperecho,lesions with irregular shape,boundary fuzziness,peripheral hyperechoic zone and hyperecho have higher incidence rate of ALN+,the difference is significant(all P<0.05).3.In SWE,comparing to the breast lesions without black hole sign,stiff rimsign,visual pattern classificationg type 4,Ecol red,Ehomo heterogeneous,lesions with black hole sign,stiff rim sign,visual pattern classificationg type 4,Ecol red and Ehomo heterogeneous have higher incidence rate of ALN(+),the difference is significant(all P<0.05).4.The mean value of Emax、Emean、SD and Eratio of ALN(+)is higher than that of ALN-,the difference is significant(all P<0.05).5.According to the result of logistic analysis of lesion’s US and SWE,there are 7 features including shape、boundary、hyperecho、visual pattern classificationg、Emax、Emean and Eratio going into the ALN model.The AUC,best cut off point,sensitivity and specificity of ALN model in prognosis metastasis of ALN is 0.706,0.4885,74.7% and 58.0%.6.ALN model indicates remarkably larger AUC to 7 independent variable(all P<0.05).However,all diadynamic criterias of model show no difference to those of ALN-US(all P>0.05),there is no reason regarding that they are different.7.The AUC of ALN model+ALN-US is bigger than that of ALN-US and Emax+ALN-US(P<0.05).The sensitivity of ALNmodel+ALN-US is bigger than that of Emax+ALN-US(P<0.05),but smaller than ALN-US、13 features+ALN-US and hyperecho+ALN-US(P<0.05).The specificity and accuracy of ALNmodel+ALN-US is bigger than that of other methers(all P<0.05).8.The Emax range of lesion’s SWE in p53(+)is smaller than that in p53(-),the difference is significant(P<0.05).The correlation coefficient(r)between Emax range and PR is-0.196,P=0.023.9.The TopoⅡ(+)show lower detection rate of peripheral hyperechoic zone than TopoⅡ(-),the difference is significant(P<0.05).Their correlation coefficient(r)is 0.135,P=0.031.Conclusions 1.The sonograms of irregular shape,boundary fuzziness,hyperecho and peripheral hyperechoic zone in US of breast lesions shows important value in predicting metastasis of ALN.2.The qualitative features of lesion’s SWE including black hole sign,stiff rim sign,visual pattern classificationg type,Ecol and Ehomo can provide important information to predict metastasis of ALN.3.The diagnosis performance of qualitative parameters of breast lesion’s SWE including Emax、Emean、SD、Eratio,show huge value in predicteding metastasis of ALN in breast cancer.4.ALN model has better clinical diagnostic value in determining ALN metastasis,it is not better than ALN-US.But the combined method between them has better diagnostic efficiency and is more helpful for for the preoperative staging and prognosis of breast cancer than other combined methods.5.The Emax range as one of qualitative parameters of SWE of breast lesions is in negative correlation to the expression of p53.Emax range may reflect the expression of p53 in breast cancer to a certain extent to a certain extent.6.The expression of TopoⅡ is in positive correlation to size and is in negative correlation to peripheral hyperechoic zone of breast lesions in US.Size and peripheral hyperechoic zone may reflect TopoⅡ expression to a certain extent. |
PDF Full Text Request