Axillary nodal burden and preoperative predictors in biopsy-proven node-positive breast cancer undergoing upfront surgery.

INTRODUCTION: Recent trials support omitting axillary lymph node dissection (ALND) in selected sentinel node-positive patients, but those with preoperative biopsy-proven nodal metastasis remain underrepresented. We characterized axillary nodal burden and its preoperative predictors. MATERIALS AND METHODS: We retrospectively studied patients with cT1-3 breast cancer and biopsy-proven axillary metastasis who underwent upfront ALND between 2008 and 2023 with at least one positive node. Nodal burden was classified as limited (1-2 positive nodes) or extensive (≥3). Multivariable logistic regression identified predictors. RESULTS: Among 1671 patients (median 21 nodes retrieved), 662 (39.6%) had limited and 1009 (60.4%) had extensive nodal burden. Suspicious node count on axillary ultrasound (AUS) was independently associated with extensive burden (two nodes: OR 2.26, 95% CI 1.64-3.12; ≥3 nodes: OR 2.08, 1.67-2.59; both p < 0.001), as was higher clinical T stage (cT2: OR 1.35, p = 0.009; cT3: OR 1.57, p = 0.008). Extensive burden ranged from 41.0% (cT1, one suspicious node) to 75.0% (cT3, two suspicious nodes). The AUS ≥3 and AUS 2 groups did not differ for ≥3 positive nodes but differed for ≥10 positive nodes (24.8% vs. 14.0%). Among 574 patients with a non-palpable axilla, AUS remained associated with extensive burden whereas clinical T stage did not. CONCLUSION: Extensive nodal burden was common yet heterogeneous. AUS suspicious node count and clinical T stage were associated with extensive disease, but discrimination was modest. These findings do not support uniform omission of ALND and require prospective validation.