Construction of a prediction model for axillary lymph node metastasis of breast cancer based on real-time shear wave elastography parameters
HOU Xiaoke
GUO Qiang
MA Junqiang
LI Panpan
HAN Guohui
Abstract：Objective To explore the efficacy of shear wave elastography(SWE)combined with clinical parameters in predicting axillary lymph node(ALN)metastasis of breast cancer.Methods A total of 305 breast cancer patients admitted to Yuncheng Central Hospital from January 2023 to December 2024 were selected as the research subjects.They were randomly divided into the training set(n=210)and the test set(n=95)in a ratio of 7∶3.According to the histopathological results,the patients were divided into the ALN positive group and the ALN negative group.Before treatment,all patients'conventional ultrasound and SWE images were collected,and laboratory indicators were tested.The receiver characteristic operating curve was used to evaluate the value of predicting ALN positivity,and the area under the curve(AUC),sensitivity and specificity were calculated.Results The training set included 113 cases of ALN-negative and 97 cases of ALN-positive.The test set included 47 ALN-negative cases and 48 ALN-positive cases.In the training set and test set,clinical cT2 stage,positive ALN reported by ultrasound,breast imaging reporting and Data System(BI-RADS)classification,and SWE hard ring sign were associated with histological metastasis of ALN(P<0.05).In both the training set and the test set,the average SWV of ALN in the ALN positive group was significantly higher than that in the ALN negative group(P<0.05).In the training set,the AUC of SWV for differentiating positive and negative ALN metastases was 0.846(95%CI:0.790-0.892),with a sensitivity of 80.41%and a specificity of 92.04%.Furthermore,in the training set and test set,the neutrophil count/lymphocyte count(NLR)in the ALN positive group[training set:3.05(2.47,3.95)vs.1.87(1.56,2.74);test set:2.95(2.20,3.93)vs.1.87(1.37,2.40)],platelet count/lymphocyte count(PLR)[Training set:190.10(138.57,254.82)vs.128.39(107.81,165.92);test set:188.00(143.31,269.78)vs.122.94(104.37,153.26)],systemic immune inflammation index(SII)[Training set:860.81(534.71,1 349.74)vs.639.78(423.18,789.57);test set:691.25(501.42,1 188.75)vs.586.14(417.64,834.54)]was significantly higher than that of the ALN negative group(P<0.05).In the training set,the results of multivariate Logistic regression showed that BI-RADS class 5,SWV,SWE hard ring sign,NLR and PLR were significantly correlated with positive ALN metastasis(P<0.05).These five characteristic variables were included to construct a Logistic regression model:Logit(P)=-14.938+1.675×BI-RADS Category 5+1.958×SWV+2.682×hard ring sign+1.062×NLR+0.017×PLR.Through ROC curve analysis,the AUC value of this model for predicting ALN metastasis was 0.971(95%CI:0.938-0.989),with a sensitivity of 89.69%and a specificity of 96.46%.The AUC value of this model for predicting ALN metastasis in the test set was 0.980(95%CI:0.928-0.998),with a sensitivity of 89.58%and a specificity of 95.74%.Conclusion The prediction model constructed based on BI-RADS5,SWE hard ring sign,SWV,NLR and PLR can effectively predict the risk of ALN metastasis in breast cancer patients,providing a new tool for non-invasive assessment of the risk of ALN metastasis in breast cancer patients.
Keywords：breast canceraxillary lymph-node metastasisshear wave elastographyneutrophil count/lymphocyte countplatelet count/lymphocyte count
Publication Date：2025-12-28
Online Publishing Date：2026-01-12(First online date of this platform, not the publication date of the document)
Pages：7( 1192-1198 )