Construction Of A Risk Prediction Model For Axillary Lymph Node Metastasis In Invasive Breast Cancer

Objective: Breast cancer is the most common malignant tumor in women,and lymph node metastasis is closely related to the prognosis of breast cancer patients.In order to predict the risk of lymph node metastasis in patients with invasive breast cancer,the purpose of this study was to construct a risk prediction model for axillary lymph node metastasis of invasive breast cancer.Method: The relevant information of 203 patients with invasive breast cancer admitted to Shengjing Hospital Affiliated to China Medical University was analyzed retrospectively,and the relevant indicators were detected by immunohistochemistry to screen out the risk factors for axillary lymph node metastasis of invasive breast cancer.The risk prediction model was constructed by binary Logistic multivariate analysis,and the risk prediction model was validated and evaluated.And draw a column diagram.Results: Analysis of risk factors for invasive breast cancer with axillary lymph node metastasis.Univariate analysis showed that 10 factors were correlated with axillary lymph node metastasis.They were age(2=5.430,P=0.024),menstrual status(2=14.630,P < 0.001),diabetes(2=4.428,P=0.040),lymphatic/vascular invasion(2=15.397,P < 0.001),and maximum tumor diameter(2=20.107,P < 0).01)and pathological types(2 = 7.996,P = 0.018),E cadherin(2 = 6.848,P = 0.011),the TGF-beta(2 = 27.709,P < 0.001),p53(2 = 3.715,P = 0.062),molecular classification(2 =7.508,P < 0.057).Binary Logistic multivariate analysis showed that whether there was cancer thrombus(P=0.026),maximum tumor diameter(cm)(P=0.004),p53(P=0.044),TGF-β(P=0.004),menstrual status(P=0.006),diabetes(P=0.002),These 6 factors are independent risk factors for axillary lymph node metastasis of breast cancer.Finally,the risk prediction model of axillary lymph node metastasis in invasive breast cancer was constructed as follows: Logit(P)=-4.404+0.862× vascular tumor thrombi +1.073× maximum tumor diameter(> 2cm)+1.017×TGF-β(positive)+3.193×(diabetes)-0.294×p53(normal expression)-1.615× menstrual status(menopause).ROC curve was drawn based on the probability value of the model to predict axillary lymph node metastasis,and the area under ROC curve(AUC)of the training group was 0.855.Sensitivity was 82.26%,specificity was 73.77%,positive predictive value was 80.36%,and negative predictive value was 80.36%.The area under ROC curve(AUC)of the verification group was 0.738.Sensitivity was 69.23%,specificity was 60.98%,positive predictive value was 62.79%,and negative predictive value was 67.57%.Hosmer Lemeshow test results P=0.824 > 0.05,indicating that the model has a good calibration degree.Conclusions: 1.The presence of cancer embolus,maximum tumor diameter(cm),p53,TGF-β,menstrual status and diabetes are independent risk factors for axillary lymph node metastasis of breast cancer.2.The risk prediction model was as follows:Logit(P)=-4.404+0.862× vascular tumor thrombi +1.073× maximum tumor diameter(> 2cm)+1.017×TGF-β(positive)+ 3.193×(diabetes)-0.294×p53(normal expression)-1.615× menstrual status(menopause).3.The AUC of the training group and all patients was > 0.75,and the AUC of the verification group was close to 0.75,indicating that the risk prediction model was well differentiated and could be used to predict the risk of axillary lymph node metastasis of breast cancer.4.Hosmer-Lemeshow test showed that P=0.426 > 0.05,indicating that the risk prediction model had good predictive ability and high accuracy in predicting axillary lymph node metastasis of breast cancer.