Author/Authors :
Cho، Yun Hee نويسنده Department of Radiology, Anam Hospital, Korea University
College of Medicine, Seoul, Korea , , Cho، Kyu Ran نويسنده Department of Radiology, Korea University College of
Medicine, Seoul, Korea , , Park، Eun Kyung نويسنده Department of Radiology, Korea University College of
Medicine, Seoul, Korea , , Seo، Bo Kyoung نويسنده Department of Radiology, Korea University College of
Medicine, Seoul, Korea , , Woo، Ok Hee نويسنده Department of Radiology, Korea University College of
Medicine, Seoul, Korea , , Cho، Sung Bum نويسنده Department of Radiology, College of Medicine, Korea
University, Seoul, Korea , , Bae، Jeoung Won نويسنده Department of Surgery, College of Medicine, Korea
University, Seoul, Korea ,
Abstract :
In preoperative assessment of breast cancer, MRI has been shown to
identify more additional breast lesions than are detectable using
conventional imaging techniques. The characterization of additional
lesions is more important than detection for optimal surgical treatment.
Additional breast lesions can be included in focus, mass, and non-mass
enhancement (NME) on MRI. According to the fifth edition of the breast
imaging reporting and data system (BI-RADS®), which includes several
changes in the NME descriptors, few studies to date have evaluated NME
in preoperative assessment of breast cancer. We investigated the
diagnostic accuracy of BI-RADS descriptors in predicting malignancy for
additional NME lesions detected on preoperative 3T dynamic contrast
enhanced MRI (DCE-MRI) in patients with newly diagnosed breast cancer.
Between January 2008 and December 2012, 88 patients were enrolled in our
study, all with NME lesions other than the index cancer on preoperative
3T DCE-MRI and all with accompanying histopathologic examination. The
MRI findings were analyzed according to the BI-RADS MRI lexicon. We
evaluated the size, distribution, internal enhancement pattern, and
location of NME lesions relative to the index cancer (i.e., same
quadrant, different quadrant, or contralateral breast). On
histopathologic analysis of the 88 NME lesions, 73 (83%) were malignant
and 15 (17%) were benign. Lesion size did not differ significantly
between malignant and benign lesions (P = 0.410). Malignancy was more
frequent in linear (P = 0.005) and segmental (P = 0.011) distributions,
and benignancy was more frequent in focal (P = 0.004) and regional (P
< 0.001) NME lesions. The highest positive predictive value (PPV)
for malignancy occurred in segmental (96.8%), linear (95.1%), clustered
ring (100%), and clumped (92.0%) enhancement. Asymmetry demonstrated a
high positive predictive value of 85.9%. The frequency of malignancy was
higher for NME lesions located in the same quadrant with the index
cancer (P = 0.006), and benignancy was higher in the contralateral
breast (P = 0.015). On multivariate analysis, linear (P = 0.001) and
segmental (P = 0.005) distributions were significant predictors of
malignancy. The possibility of malignancy is strongly indicated when
additional NME lesions show linear or segmental enhancement on
preoperative 3T DCE-MRI in patients with recently diagnosed breast
cancer.