OBJECTIVE: To evaluate the accuracy of magnetic resonance imaging (MRI) of the breasts in the identification of a pathological complete response in patients with breast cancer undergoing neoadjuvant chemotherapy (NAC). MATERIALS AND METHODS: This was a single-center, retrospective, observational study designed to validate a diagnostic test. The following variables were evaluated: age; results of the histological and immunohistochemical analysis of the biopsy; post-NAC MRI findings; and results of the histological analysis of the surgical specimen, using the residual cancer burden index. The radiological response, as assessed by MRI, was compared with the pathological response, as assessed by histological analysis of the surgical specimen (the gold standard method). RESULTS: We evaluated 310 tumors in 308 patients. The mean age of the patients was 47 years (range, 27–85 years). For identifying a pathological complete response, breast MRI had an overall accuracy of 79%, with a sensitivity of 75%, specificity of 83%, positive predictive value of 75%, and negative predictive value of 83%. When that accuracy was stratified by molecular subtype, it was best for the HER2 subtype, with a sensitivity and specificity of 82% and 89%, respectively, followed by the triple-negative subtype, with a sensitivity and specificity of 78% and 83%, respectively. CONCLUSION: Breast MRI showed good accuracy in the prediction of a pathological complete response after NAC. The sensitivity and positive predictive value were highest for the HER2 and triple-negative subtypes.

Keywords: Breast neoplasms; Drug therapy, combination; Magnetic resonance imaging; Triple negative breast neoplasms; Receptor, ErbB-2.