Dynamic MRI of the breast

Abstract

Objective. To describe the rationale, the technical requirements and the examination technique of dynamic magnetic resonance studies of the breast and to assess the role of this method in the clinical diagnostic protocol. Material and methods. We reviewed the relative literature and compared the results with our personal experience. Results. The earliest reports on the possibility of differentiating carcinoma from benign tissue with magnetic resonance imaging came from Germany, where in 1986 Heywang and coworkers used T1-weighted spin-echo sequences before and after i.v. Gd-DTPA administration with an imaging time of approximately 5 mins and 5 mm slice thickness. With advances in magnetic resonance technology, Heywang–Koebrunner adopted a static three-dimensional fast low-angle shot technique permitting <3 mins' imaging time and providing high resolution with thin slices. This approach may detect even the carcinoma which occasionally does not enhance in the typical rapid, intense way. A review of 400 biopsy-proved lesions showed that all carcinomas enhanced strongly, all but 5% rapidly and 85% focally. Over 70% of benign masses exhibited no major contrast uptake, although some benign tumors and proliferative dysplasias enhanced strongly. At present, Heywang–Koebrunner use rapid three-dimensional gradient-echo sequences (TR/TE 14/7, FA 25°; at 1.0 T) which have the advantages of three-dimensional imaging and permit dynamic studies (<1 min/sequence). The same sequence is used at our institution. The dynamic technique is advocated by Kaiser who in 1989 reported preliminary sensistivity and specificity values over 95%. The time/signal intensity curves revealed the rapid and strong enhancement of malignancies, the gradual and strong enhancement of the only fibroadenoma studied, and the gradual and mild contrast uptake of benign dysplasia. In 1992, in nearly 1000 dynamic examinations, Kaiser and Reiser reported 98.3% sensitivity, 97.0% specificity, 82.1% predictive value and 97.2% accuracy. The combination of rapid acquisition with techniques that preserve high spatial resolution may improve specificity by allowing the study of lesion morphology as well as of enhancement patterns. A whole-breast imaging technique has been reported which permits acquisition times <15 s by partial sampling of the central k-space region superimposed on high-resolution three-dimensional images. Conclusions. Dynamic magnetic resonance imaging of the breast is currently an important step of the clinical protocol of breast diseases, but there is no established study protocol yet.