Feasibility And Application Value Of Ultrasonographic Lymph Node Classification In Predicting Axillary Lymph Node Metastasis Of Breast Cancer

Objectives:The aim of this study was to construct an ultrasonographic classification model for predicting axillary lymph node metastasis(ALNM)of breast cancer.And to preliminarily explore the feasibility and application value of the ultrasonographic classification model for axillary lymph nodes of breast cancer.Methods:From September 2016 to December 2017,528 patients with breast cancer admitted to the First Hospital of China Medical University were retrospectively analyzed.The ultrasonographic parameters belonging to 528 lesions of axillary lymph nodes were collected.Univariate chi-square test and multivariate logistic method were used to analyze the predictive value of acoustic parameters of axillary lymph nodes.The parameters were selected and scored according to the weight of diagnosis of nature of lymph nodes.The ROC curve of multivariate model was drawn according to the scores.Similarly to the BI-RADS classification of breast,we constructed a 1-6classification model,i.e.the six-classification method.Each classification represents the different significance of axillary lymph node metastasis of breast cancer.From January 2018 to March 2018,100 cases of breast cancer patients in our hospital were collected.Two doctors classificationd patients according to the ultrasonic six-classification method in the first part.The accuracy of the ultrasonic six-classification method in classing axillary lymph nodes of breast cancer was calculated.Kappa test was used to judge the consistency of ultrasonic six-classification method used by different physicians.Paired chi-square test was used to judge the difference of the positive rate of diagnosis of six-classification method used by different doctors.Results: In univariate analysis,cortical thickness,lymphatic hilum,lymph node blood flow,lymph node morphology,microcalcification,bilateral contrast,lesion quadrant,BI-RADS classification and size of ultrasound mass were correlated with axillary lymph node metastasis(P<0.001).After multivariate logistic method,BI-RADS classification,cortical thickness,portal structure,lymph node morphology and bilateral contrast were predictive of breast axillary lymph node metastasis(P<0.05).Area under the ROC curve(AUC)was 0.914(P<0.05),and the multi-factor data wasused to construct the expected six-classification model.Kappa test showed that the Kappa value of the six-classification method was 0.636(P<0.001),which showed moderate consistency among physicians with different seniority.Paired Chi-square test showed that there was no significant difference in the positive rate of six-classification method among different physicians(P>0.05).Conclusion: The multi-factor data can be used to construct the expected lymph node classification model.It is feasible to classify ALN of breast cancer by ultrasound.And it is conducive to repeated clinical application.