Magnetic resonance imaging of the human female breast. Current status and pathologic correlations.

Over the past decade, research in MR of the female breast has focused on in vitro studies of excised breast tissue samples or mastectomy specimens and on in vivo clinical imaging of patients, usually with biopsy correlation. The results of in vitro measurement of T1 and T2 times of excised benign and malignant breast tissue do not conclusively demonstrate a clear separation based on these values alone. In particular, benign tumors (fibroadenomas) cannot be distinguished from carcinomas and there is considerable overlap in values for cancer and fibrocystic disease. MR images produced from mastectomy specimens allow detailed correlation with mammographic and histologic findings but are incompletely applicable to in vivo images because of numerous factors including tissue changes associated with disruption of blood supply. In vivo studies have shown that benign tumors (fibroadenomas), fibrocystic changes, and cancers can be imaged successfully by MR. To date the best images show a resolution of about 2 mm. Small cancers and cysts have been imaged successfully but proliferative patterns of fibrocystic disease cannot be differentiated at this time. Carcinomas occurring in dense fibroglandular breasts which are a problem to image mammographically have not been successfully imaged by MR in vivo and in only one instance in vitro in a mastectomy specimen. Since MR cannot image microcalcifications, it does not appear likely to replace mammography as the imaging modality of choice at the present time. Newer magnets of increased strength, specialized surface coils specifically for breast imaging, and the use of contrast enhancing agents all may increase the usefulness of MR as an imaging modality for the detection of breast disease. In particular, increased resolution and contrast enhancing agents which may aid in tissue differentiation should be studied for their ability to detect carcinomas in dense fibroglandular breast tissue and to identify lesions of carcinoma in situ and proliferative patterns of fibrocystic disease. The need for careful collaborative studies between radiologists and pathologists will be essential for the success of these research efforts.