Routine use of standard breast MRI compared to axillary ultrasound for differentiating between no, limited and advanced axillary nodal disease in newly diagnosed breast cancer patients

Thiemo van Nijnatten*, E. H. Ploumen, R. J. Schipper, B. Goorts, E. H. Andriessen, S. Vanwetswinkel, M. Schavemaker, P. Nelemans, B. de Vries, R. G. H. Beets-Tan, M. L. Smidt, M. B. I. Lobbes

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review


OBJECTIVES: To compare standard breast MRI to dedicated axillary ultrasound (with or without tissue sampling) for differentiating between no, limited and advanced axillary nodal disease in breast cancer patients. METHODS: All patients who underwent breast MRI and dedicated axillary ultrasound between 2009 and 2014 were eligible. Exclusion criteria were recurrent disease, neoadjuvant systemic therapy and not receiving completion axillary lymph node dissection after positive sentinel lymph node biopsy (SLNB). Two radiologists independently reassessed all MRI exams. Axillary ultrasound findings were retrospectively collected. Probability of advanced axillary nodal disease (pN2-3) given clinically node negative (cN0) or limited (cN1) findings was calculated, with corresponding negative predictive value (NPV) to exclude pN2-3 and positive predictive value (PPV) to identify axillary nodal disease. Histopathology served as gold standard. RESULTS: A total of 377 cases resulted in 81.4% no, 14.4% limited and 4.2% advanced axillary nodal disease at final histopathology. Probability of pN2-3 given cN0 for breast MRI and axillary ultrasound was 0.7-0.9% versus 1.5% and probability of pN2-3 given cN1 was 11.6-15.4% versus 29.0%. When cN1 on breast MRI was observed, PPV to identify positive axillary nodal disease was 50.7% and 59.0%. CONCLUSIONS: Evaluation of axillary nodal status on standard breast MRI is comparable to dedicated axillary ultrasound in breast cancer patients. In patients who underwent preoperative standard breast MRI, axillary ultrasound is only required in case of suspicious nodal findings on MRI.
Original languageEnglish
Pages (from-to)2288-2294
JournalEuropean Journal of Radiology
Issue number12
Publication statusPublished - Dec 2016


  • Breast cancer
  • Axilla
  • Lymph nodes
  • Magnetic resonance imaging
  • Ultrasound

Cite this