346P - Concordance between radiological and pathological response in axillary lymph nodes in patients with breast cancer treated with neoadjuvant chemothe...

Date 27 September 2014
Event ESMO 2014
Session Poster Display session
Topics Anticancer Agents
Staging Procedures (clinical staging)
Pathology/Molecular Biology
Breast Cancer
Basic Scientific Principles
Basic Principles in the Management and Treatment (of cancer)
Biological Therapy
Presenter Sonia Servitja Tormo
Citation Annals of Oncology (2014) 25 (suppl_4): iv110-iv115. 10.1093/annonc/mdu328
Authors S. Servitja Tormo1, L. Garrigos Cubells1, A. Rodriguez Arana2, M.D. Sabadell Mercadal3, J.M. Corominas4, M. Martinez-Garcia5, I. Gonzalez Maeso6, T. Martos Cardenas7, J. Albanell8, I. Tusquets Trias Bes8
  • 1Medical Oncology, Hospital del Mar, 08003 - Barcelona/ES
  • 2Radiology, Hospital del Mar, 08003 - barcelona/ES
  • 3Gynecology, Hospital del Mar, 08003 - barcelona/ES
  • 4Pathology Department, Hospital del Mar, 08003 - barcelona/ES
  • 5Department Of Medical Oncology, Hospital del Mar, 08003 - Barcelona/ES
  • 6Medical Oncology Department., University Hospital del Mar-IMIM, 08003 - Barcelona/ES
  • 7Oncology, Hospital del Mar, 08003 - barcelona/ES
  • 8Medical Oncology, University Hospital del Mar, 08003 - Barcelona/ES



Magnetic Resonance Imaging (MRI) is accepted as the best method to predict pathological response to neoadjuvant chemotherapy (NAC) in breast cancer patients, but some studies reported differences in the efficacy of MRI to detect pathological response in axilla. The aim of the study is to evaluate the correlation between radiological and pathological response in axilla.


This study included patients with clinical stage I-III who were treated with NAC at our breast cancer unit. Patients underwent a baseline MRI and after NAC. The radiological response was assessed according to the RECIST version 1.1 and the pathological response according to the Miller & Payne grading system.


Between January 2006 and June 2013 247 patients were included. Patient characteristics are described in table 1. Anthracyclines and taxanes were the most frequent drugs used (88.7%). Regarding the type of surgery, 69.2% of patients underwent breast conserving surgery and 85.8% underwent lymphadenectomy. A pathological complete response (pCR) was achieved in 17.4% patients, 19% in breast and 24.2% in axilla We have analyzed the ability of MRI to detect pN0 in axilla after NAQ reporting sensitivity of 76% and specificity of 24%. Patients achieving pCR in breast have a probability of 63.8% to achieve pN0 also in axilla. MRI was not able to detect pN0 in 39% of patients with initial stage cN1(FNA+) and in 66% of patients with sentinel node (SN) biopsy positive before NAQ, whose underwent lymphadenectomy (39.6% classified as A and 56.6% classified as D). Table 1

Characteristics n(247) %
Median age 56 + /-12
Stage I 8 3.2
II 137 55.4
III 102 41.4
Immunophenotype Luminal A 45 18.2
Luminal B-HER2(-) 83 33.6
Luminal B-HER2(+) 32 13
Triple Negative 50 20.2
HER2(+) 30 12.1
Unknown 7 2.8
Axillary lymph node at initial stage cN1(FNA+) 109 44.1
cN1(SN+) 14 5.6
cN0 124 50.3


MRI achieves Se 76% and Sp 24% predicting pN0 in axilla. The probability of obtaining pN0 in axilla is higher in those patients with pCR in breast. 39% of patients with initial stage cN1 and 66% of patients with SN biopsy positive before NAQ have achieved an axillar pN0 in the lymphadenectomy. These results suggest the need to improve the detection of axillary dowstaging before surgery to avoid unnecessary lymphadenectomy.


All authors have declared no conflicts of interest.