Tumor-Proximal Biomarkers From Core Needle Biopsy Washout Fluid Improve Preoperative Prediction of Axillary Nodal Metastasis in Breast Cancer: A Combined Ultrasound-Based Nomogram

Abstract Background Preoperative evaluation of axillary lymph node (ALN) status guides decisions on axillary surgery and systemic treatment in breast cancer. Axillary ultrasound (US) is widely used but may miss occult nodal involvement. We examined whether vascular endothelial growth factor-C (VEGF-C) and cytokeratin 19 (CK19) measured in core needle biopsy (CNB) washout fluid could complement US for estimating ALN metastasis risk and developed a combined nomogram. Methods We retrospectively analyzed 216 patients with primary breast cancer who underwent preoperative US-guided CNB and subsequent standard ALN surgery. Washout fluid retained in the biopsy device was collected, and VEGF-C and CK19 concentrations were quantified by ELISA. Pathology served as the reference standard for ALN status (N0 vs N+). Predictors were assessed using univariate and multivariable logistic regression. Model performance was evaluated by ROC/AUC, bootstrap calibration, and decision curve analysis (DCA). Results ALN metastasis was present in 91 of 216 patients (42.1%). VEGF-C and CK19 levels in washout fluid were higher in the N+ group than in the N0 group (VEGF-C median 2080.4 vs 904.9 pg/mL; CK19 median 6.59 vs 2.06 ng/mL; both P < 0.001). In multivariable analysis, US nodal status (P < 0.001), tumor size (OR 1.560; P = 0.007), log2-VEGF-C (OR 2.969; P < 0.001), and log2-CK19 (OR 2.865; P < 0.001) independently predicted ALN metastasis. The nomogram achieved an AUC of 0.916 (95% CI 0.878–0.954) and outperformed US alone (AUC 0.767; P < 0.001, DeLong test). Calibration indicated good agreement between predicted and observed risk, and DCA showed higher net benefit across threshold probabilities of 5%–80%. Conclusions VEGF-C and CK19 measured from CNB washout fluid provide tumor-proximal information that complements ultrasound for preoperative nodal risk assessment. A nomogram integrating these biomarkers with tumor size and US findings improved discrimination and clinical utility for predicting ALN metastasis, and may support more individualized axillary management.