Abstract
We aimed to predict axillary pathologic complete response (pCR) in patients with clinical N2-3 breast cancer who underwent neoadjuvant chemotherapy (NAC) followed by breast and axillary surgeries. These patients were retrospectively analyzed between 2016 and 2020. Imaging metrics (lesion size and reduction ratios measured using magnetic resonance imaging (MRI) and ultrasonography (US)) and clinical and pathological factors were analyzed. Axillary pCR was achieved in 49.3% (n = 136/276) of the patients. Post-NAC significant predictors included breast lesion and lymph node (LN) sizes on MRI and LN cortical thickness on US. Pre-NAC tumor-specific factors were also associated with axillary pCR. In multivariable logistic regression analyses using three models, Model 3 (imaging variables, tumor characteristics, and molecular subtypes) demonstrated the best predictive performance (AUC: 0.9). Integrating molecular subtype-specific imaging and tumor characteristics enabled precise prediction of axillary pCR in patients with cN2-3 breast cancer, supporting tailored surgical approaches based on individual responses.