Is the performance of MRI in preoperative staging of breast cancer independent of clinical and histological factors? A subgroup analysis.

OBJECTIVE To determine whether preoperative breast MRI is more useful in patients according to their breast density, age, menopausal status, and biopsy findings of carcinoma in situ. MATERIAL AND METHODS We retrospectively studied 264 patients treated for breast cancer who had undergone mammography, ultrasonography, and MRI. We compared the size of the tumor on the three techniques and the sensitivity of the techniques for detecting additional lesions both in the overall group and in subgroups of patients classified according to their breast density, age, menopausal status, and histological findings of intraductal carcinoma. The definitive histological diagnosis was used as the gold standard. RESULTS MRI was the technique that was most concordant with the histological findings for the size of the lesion, and it was also the technique that detected the most additional lesions. With MRI, we observed no differences in lesion size between the overall group and the subgroups in which MRI provided added value. Likewise, we observed no differences in the number of additional lesions detected in the overall group except for multicentric lesions, which was larger in older patients (P=.02). In the subgroup of patients in which MRI provided added value, the sensitivity for bilateral lesions was higher in patients with fatty breasts (P=.04). Multifocal lesions were detected significantly better in premenopausal patients (P=.03). CONCLUSIONS MRI is better than mammography and better than ultrasonography for establishing the size of the tumor and for detecting additional lesions. Our results did not identify any subgroups in which the technique was more useful.