Construction and Evaluation of A Risk Prediction Model for Axillary Lymph Node Metastasis in Breast Cancer Based on Ultrasound Characteristics and Conventional Tumor Markers

GAO Jianqing; GUO Hailan; XU Na; LI Jiajun; LU Min

doi:10.3639/j.issn.2095-3097.2025.05.017

Abstract：Objective To analyze the relationship between ultrasound features,conventional tumor markers and axil-lary lymph node metastasis(ALNM)in breast cancer patients,and build a risk prediction model.Methods A total of 120 breast cancer patients admitted to Meishan Third People's Hospital from April 2020 to December 2024 were selected as the research subjects and divided into the ALNM group(n=73)and the non-ALNM group(n=47)based on whether ALNM occurred.The in-fluencing factors of ALNM in breast cancer patients were explored by multivariate logistic stepwise regression analysis.A nomo-gram model was constructed and its predictive value for ALNM in breast cancer patients was evaluated.Results The proportions of patients with a maximum tumor diameter≥2 cm,calcification foci,rich blood flow signals in the tumor,lymph node aspect ratio<2,heterogeneous internal echoes of lymph nodes,unclear lymph node boundaries,and rich blood flow signals in lymph nodes in the ALNM group were all higher than those in the non-ALNM group,and the levels of serum carcinoembryonic antigen(CEA),carbohydrate antigen 125(CA125),and carbohydrate antigen 153(CA153)were higher in the ALNM group,and the differ-ence was statistically significant(P<0.05).Multivariate analysis showed that the presence of calcification foci(OR=2.487,95%CI:1.674-3.695),lymph node aspect ratio<2(OR=0.332,95%CI:0.214-0.516),rich blood flow signals in lymph nodes(OR=2.234,95%CI:1.552-3.217),high levels of CEA(OR=1.914,95%CI:1.388-2.639),high levels of CA125(OR=2.208,95%CI:1.524-3.198),and high levels of CA153(OR=2.323,95%CI:1.595-3.385)were risk factors for ALNM in breast cancer patients(P<0.05).The nomogram model showed that the concordance index(C-index)was 0.901(95%CI:0.848-0.954).In the goodness-of-fit test,χ2=0.759,P=0.438;the calibration curve was close to the ideal curve.The area under the curve(AUC)(95%CI)of the nomogram model for predicting ALNM in breast cancer patients was 0.893(0.840-0.946),with a specificity of 0.682 and a sensitivity of 0.921.Conclusion The nomogram model constructed based on ultrasound characteristics and conventional tumor markers has certain predictive value for ALNM in breast cancer patients and can be used as an important tool to guide clinical treatment.