Fibrocystic Changes of the Breast: Radiologic–Pathologic Correlation of MRI

Abstract

Breast MRI provides high sensitivity but modest positive predictive value for identifying breast cancers, with approximately 75% of MRI-guided biopsies returning benign pathologies. Fibrocystic change (FCC) is a descriptive term used colloquially by many radiologists (and falling out of favor with many pathologists) to refer to several benign entities encountered in the breast. Many of the benign entities believed to comprise FCC can show enhancement on MRI. Recognizing the pathologic correlates of these enhancing lesions should help guide management after such a result on MRI-guided biopsy. Premenopausal women may present with clinical symptoms attributed to FCC, including pain, nipple discharge, breast lumps, or discrete masses. Benign entities associated with FCC include proliferative lesions such as usual ductal hyperplasia and sclerosing adenosis, and nonproliferative lesions including cysts, apocrine metaplasia, and stromal fibrosis. Fibrocystic change can be diffuse or focal. Diffuse FCC usually presents as non-mass enhancement (NME), often with persistent kinetics. Focal FCC can present as an irregular mass or focus with variable enhancement patterns including washout kinetics. Following a benign concordant MRI-guided biopsy result of one or more of the above entities, follow-up with MRI in 12 months is reasonable. Accurate radiologic–pathologic correlation can be achieved when careful review of histologic findings is carried out in the context of MRI features.

Challenges of non-mass enhancement breast lesions mimicking malignancy encountered on breast magnetic resonance imaging

“…Broad histological variations characterize FCC of the breast, encompassing stromal fibrosis, cysts, adenosis, apocrine metaplasia, and epithelial proliferation to various degrees ( 3 ). This histological diversity may account for the wide spectrum of MRI features of FCC ( 4 ). Previous studies have reported a wide range (34–95.1%) of PPV for segmental distribution, and identified that the dynamic curves were less effective, a contrast to those of this study ( 5 - 7 ).…”