Construction and Validation of a Nomogram Model for Predicting Lymph Node Metastasis in Breast Cancer Based on
Xiangyang Huang1, Xin Li2, Huiping Lyu1
1Department of Ultrasound, Nan'an Hospital Affiliated to Shanghai University, Quanzhou, Fujian, China.
Purpose:
This study aimed to construct a multifactorial predictive model by integrating ultrasound imaging parameters of primary breast tumors and axillary lymph nodes, along with clinical pathological indicators. The goal was to enhance the accuracy of predicting axillary lymph node metastasis and provide a basis for clinical decision-making.
Materials And Methods:
A total of 268 breast cancer patients treated at Nan'an Hospital in Fujian Province from March 2022 to December 2024 were included in this study. They were randomly divided into training (187 cases) and validation (81 cases) cohorts in a 7:3 ratio. Ultrasound examinations were conducted to measure parameters such as maximum tumor diameter, lymph node cortex-to-medulla area ratio, the peak systolic velocity and lymph node resistive index of lymph node vascular, along with the collection of clinical pathological indicators, including pathological grade, tumor type, tumor location, menopausal status and age. Univariate and multivariate logistic regression analyses were performed to construct and validate a predictive model for breast cancer lymph node metastasis.
Results:
The results of the multivariate logistic regression analysis indicated that tumor maximum diameter (OR = 1.463, 95% CI: 1.250-1.712), lymph node cortex-to-medulla area ratio (OR = 11.878, 95% CI: 1.351-104.449), lymph node peak systolic velocity (OR = 1.165, 95% CI: 1.066-1.273) and lymph node resistive index (OR = 5.136, 95% CI: 2.721-9.693) were independent risk factors for breast cancer lymph node metastasis. The nomogram model constructed based on these factors demonstrated good predictive ability in both the training and validation cohorts, with area under the curves of 0.958 (95% CI: 0.930-0.986) and 0.958 (95% CI: 0.922-0.995), respectively. Internal validation using the Bootstrap method with 1000 repetitions showed consistent calibration curves and Hosmer-Lemeshow goodness-of-fit tests (p > 0.05), indicating good agreement between the predicted and observed probabilities of breast cancer lymph node metastasis. Decision curve analysis revealed that within a risk threshold range of 0.1-0.6, the model provided the greatest net benefit, indicating its strong clinical utility.
Conclusion:
Tumor maximum diameter, lymph node cortex-to-medulla area ratio, lymph node peak systolic velocity and lymph node resistive index are independent risk factors for axillary lymph node metastasis in breast cancer. The predictive model built on these factors demonstrates high accuracy and clinical utility in predicting breast cancer lymph node metastasis, offering important guidance for individualized treatment of breast cancer patients.
