Construction and validation of a novel nomogram for prediction of lymph node metastasis in HER2-positive breast cancer: based on the optimal number of examined lymph nodes for accurate nodal staging

This study aimed to construct and validate a novel nomogram for prediction of lymph node metastasis in HER2-positive breast cancer based on the optimal number of examined lymph nodes (ELNs) for accurate nodal staging. We included 4,040 patients diagnosed with HER2-positive breast cancer from the SEER database, randomly allocating them into training and validation cohorts in a 7:3 ratio. The optimal number of ELNs was identified via piecewise linear regression. The association of ELNs count with nodal migration was evaluated through Logistic Regression (LR) analysis and Random Forest (RF). The nomogram was constructed, and its' performance was evaluated by the receiver operating characteristic curves, calibration curve and Decision curve analysis curves. The optimal number of ELNs was 13. LR and RF identified the optimal number of ELNs, radiotherapy status, chemotherapy status, T stage, and grade as independent predictive variables for node metastasis, which were used in the nomogram's construction. And the area under the curve values for the nomogram were 0.829 (95% confidence interval (CI): 0.813–0.845) and 0.833 (95% CI:0.808–0.858) in the training and test split respectively, surpassing those of the optimal number of ELNs (0.649, 95% CI: 0.631–0.667 and 0.676, 95% CI:0.648–0.704). Calibration plots exhibited low Brier scores (0.150 for training split, 0.145 for test split). This study developed a novel nomogram that integrates the optimal number of ELNs with other independent risk factors, facilitating individualized prediction of lymph node metastasis in patients with HER2-positive breast cancer.