Correlation of three-dimensional magnetic resonance imaging with precise histopathological map concerning carcinoma...

This study was initiated to clarify the ability of magnetic resonance imaging (MRI) in defining breast carcinoma extension by comparing MRI to detailed histopathological analysis. Mastectomy (n = 14) or quadrantectomy (n = 44) specimens were sub-serially sectioned and mapped in detail in 58 breast cancer patients. Morphologically, we classified the lesions utilizing MRI into three patterns in relation to their histology. Numerically, we assessed the maximum distance of carcinoma extension using MRI, mammography, and ultrasonography (US). Linear regression was calculated for each of the three imaging measurements versus histopathological measurements. Three imaging patterns were observed by MRI, (1) localized (n = 30), (2) segmentally extended (n = 19), and (3) irregularly extended (n = 5). The localized pattern showed a distinct focal mass, but in 10 cases, microscopic ductal carcinoma in situ (DCIS), or invasive lobular carcinoma, which were not depicted by MRI, existed. The segmentally extended pattern showed diffuse enhancement along duct-lobular segments, forming a 'cone' shape. Histologically, pure (n = 4) or predominant (n = 10) DCIS was distributed segmentally. The irregularly extended pattern showed thick branches extending out from the index tumor which were histologically revealed to be stromal invasion of ductal carcinoma. From the results of linear regressions, MRI was the most accurate modality in histologically measuring the extent of the cancer. When cases were limited to patients who were classified into segmentally or irregularly extended pattern by MRI (n = 24), MRI was more accurate than mammography and US, even if they were combined (P < 0.05). MRI may provide additional information concerning carcinoma extension prior to surgery, especially in patients classified into 'extended patterns' by MRI.