Does MRI predict pathologic tumor response in women with breast cancer undergoing preoperative chemotherapy?

J Surg Oncol. 2007 Nov 1;96(6):474-80. doi: 10.1002/jso.20856.


Background: Neoadjuvant chemotherapy precludes the accurate pre-surgical pathologic measurement of tumor size. The purpose of this study is to review imaging studies performed in patients who received preoperative chemotherapy prior to surgery and determine whether MRI, ultrasound (US) or physical exam best predicted final pathologic tumor size.

Methods: Stage I, II, and III breast cancer patients were treated with neoadjuvant therapy on trial. As part of the trials, women underwent MRI, US, and physical exam prior to the start of therapy and 1 week after completion of neoadjuvant chemotherapy.

Results: Of the 68 patients with MRI data, the correlation coefficient (r) of MRI to pathologic size of tumor was r = 0.749. Among the 52 patients who had an US assessment the correlation of US to pathology was r = 0.612. Sixty-two patients had physical exam data, and the correlation of examination to pathology size was r = 0.439. MRI correctly predicted 8 of 11 complete responders and accurately evaluated the size of non-responders to neoadjuvant therapy (r = 0.869)

Conclusions: In a select group of women undergoing neoadjuvant therapy for invasive breast cancer, MRI best predicted pathology response. The use of MRI in neoadjuvant therapy may allow for accurate prediction of patients eligible for breast conservation.

Publication types

  • Clinical Trial

MeSH terms

  • Adult
  • Aged
  • Breast Neoplasms / diagnosis*
  • Breast Neoplasms / diagnostic imaging
  • Breast Neoplasms / therapy
  • Chemotherapy, Adjuvant*
  • Combined Modality Therapy
  • Female
  • Humans
  • Magnetic Resonance Imaging*
  • Mammography
  • Middle Aged
  • Neoadjuvant Therapy*
  • Ultrasonography