Fat necrosis of the breast: clinical, mammographic and sonographic features

OBJECTIVE: the purpose of this study was to describe and quantitate the clinical, mammographic and sonographic (US) features and to evaluate the evolution of fat necrosis in the breast. MATERIALS AND METHODS: a retrospective review of the clinical, mammographic and US findings of 126 fat necrosis lesions in 94 patients, diagnosed between 1989 and 1999, was done. All the cases included in the study had at least 3 years follow-up mammograms. In addition, 48 patients with a total of 62 fat necrosis lesions, also had an US follow-up. Fat necrosis was diagnosed on the basis of histologic (n=25) and initial or follow-up imaging (n=69) findings. RESULTS: the predominant mammographic features of the 114 lesions apparent on mammograms were radiolucent oil cyst (n=34, 26.9%), round opacity (n=16, 12.6%), asymmetrical opacity or heterogenicity of the subcutaneous tissues (n=20, 15.8%), dystrophic calcifications (n=34, 26.9%), clustered pleomorphic microcalcifications (n=5, 3.9%), and suspicious speculated mass (n=5, 3.9%). In five patients with 12 (9.5%) palpable masses, mammograms were normal. The predominant US features of the 112 lesions apparent on sonograms were solid (n=18, 14.2%), anechoic with posterior acoustic enhancement (n=21, 16.6%), anechoic with posterior acoustic shadowing (n=20, 15.8%), cystic with internal echoes (n=14, 11.1%), cystic with mural nodule (n=5, 3.9%) and increased echogenicity of the subcutaneous tissues (n=34, 26.9%). In five patients with 14 (11.1%) lesions, sonographic examination was normal. Mammographic follow-up showed that five of the radiolucent oil cysts developed curvilinear calcifications, six of the round opacities decreased in size and density, and another two disappeared. Eleven of the dystrophic calcifications became even more coarse. Six of the asymmetrical opacities became vague and one developed an oil cyst and coarse calcifications. The only nonoperated speculated mass developed a typical small radiolucent oil cyst in the centre. US follow-up showed that 18 of the 29 increased subcutaneous tissue echogenicity turned back to normal, while in the remaining 11 small cysts formed. In 19 solid appearing masses, 15 showed decrease in size, while four remained stable (biopsy disclosed fat necrosis). The four complex masses in two patients showed increase in size and appeared more cystic (FNAB was consistent with fat necrosis). CONCLUSION: a spectrum of imaging findings is associated with fat necrosis. Knowledge of the mammographic and US appearance and evolution of these patterns may enable imaging follow-up of these lesions, reducing the number of unnecessary biopsies.