Background: New insights show that an axillary lymph node dissection (ALND) may not always be indicated for metastases detected by ultrasound (pathologically proven). This study investigated whether axillary ultrasound accurately predicts pN0, pN1 and pN2-pN3 status.
Methods: Data were retrospectively collected from all consecutive patients with invasive breast cancer who underwent (primary) surgery between 2008 and 2012. False negative percentages and negative predictive values (NPVs) for sonographic nodal staging were calculated for all patients and again for cT1-2 patients treated by breast conserving therapy (BCT).
Results: A total of 577 axillary ultrasounds were included. After negative ultrasound findings (cN0), pathology showed pN2-pN3 disease in 4.4% of these cases, with an NPV of 95.5% (93.4-97.1%). When cN1 (1-3 suspicious nodes) was predicted, pathology showed pN2-pN3 disease in 41.2%, with an NPV of 58.5% (44.2-71.5%). In the subgroup of patients with cT1-2 breast cancer that were treated by BCT, pathology showed pN2-pN3 disease in 2.3% after negative ultrasound findings (cN0), with an NPV of 97.7% (94.9-99.0%). When cN1 was predicted (n = 12), pathology showed pN2-pN3 disease in 50.0%, with an NPV of 50.0% (22.3-77.9%). A direct ALND was performed in these 12 cN1 cases; pathology showed six patients with pN1 (three patients with one and three with two macrometastases) and six with pN2-pN3 disease (4, 5, 11, 13, 16 or 22 macrometastases, respectively).
Conclusion: In conclusion, a negative axillary ultrasound generally excludes the presence of pN2-pN3 disease. An axillary ultrasound cannot accurately differentiate between pN1 and pN2-pN3. It could be argued that the standard performance of an axillary ultrasound in breast cancer patients is questionable; multidisciplinary discussion could guide decisions on the use of axillary ultrasound for the individual patient.
Keywords: Axillary lymph node dissection; Axillary ultrasound; Breast cancer; Nodal staging; Sentinel lymph node biopsy.
Copyright © 2013 Elsevier Ltd. All rights reserved.