Search

To assess which specific morphologic features, enhancement patterns, or pharmacokinetic parameters on breast Magnetic Resonance Imaging (MRI) could predict a false-negative outcome of Proton MR Spectroscopy ((1)H MRS) exam in patients with invasive breast cancer. Sixteen patients with invasive ductal carcinoma of the breast were prospectively included and underwent both, contrast-enhanced breast MRI and (1)H MRS examination of the breast. The MR images were reviewed and the lesions morphologic features, enhancement patterns and pharmacokinetic parameters (k21-value) were scored according to the ACR BI-RADS-MRI lexicon criteria. For the in vivo MRS studies, each spectrum was evaluated for the presence of choline based on consensus reading. Breast MRI and (1)H MRS data were compared to histopathologic findings. In vivo (1)H MRS detected a choline peak in 14/16 (88%) cancers. A false-negative (1)H MRS study occurred in 2/16 (14%) cancer patients. K21 values differed between both groups: the 14 choline positive cancers had k21 values ranging from 0.01 to 0.20/second (mean 0.083/second), whereas the two choline-negative cancers showed k21 values of 0.03 and 0.05/second, respectively (mean 0.040/second). Also enhancement kinetics did differ between both groups; typically both cancers that were choline-negative showed a late phase plateau (100%), whereas this was only shown in 5/14 (36%) of the choline positive cases. There was no difference between both groups with regard to morphologic features on MRI. This study showed that false-negative (1)H MRS examinations do occur in breast cancer patients, and that the presence of a choline peak on (1)H MRS as malignancy marker is related to the k21 value of the invasive tumor being imaged.

Background: We performed this study to determine rates of close or transected cancer margins after magnetic resonance imaging-guided bracket wire localization for nonpalpable breast lesions., Study Design: Of 243 women undergoing MRI-guided wire localizations, 26 had MRI bracket wire localization to excise either a known cancer (n = 19) or a suspicious MRI-detected lesion (n = 7). We reviewed patient age, preoperative diagnosis, operative intent, mammographic breast density, MRI lesion size, MRI enhancement curve and morphology, MRI Breast Imaging Reporting and Data System (BI-RADS) assessment code, number of bracket wires, and pathology size. We analyzed these findings for their relationship to obtaining clear margins at first operative excision., Results: Twenty-one of 26 (81%) patients had cancer. Of 21 patients with cancer, 12 (57%) had negative margins at first excision and 9 (43%) had close/transected margins. MRI size > or = 4 cm was associated with a higher reexcision rate (7 of 9, 78%) than those < 4 cm (2 of 12, 17%) (p = 0.009). MRI BI-RADS score, enhancement curve, morphology, and preoperative core biopsy demonstrating ductal carcinoma in situ (DCIS) were not predictive of reexcision. The average number of wires used for bracketing increased with lesion size, but was not associated with improved outcomes. On pathology, cancer size was smaller in patients with negative margins (12 patients, 1.2 cm) than in those with close/transected margins (9 patients, 4.6 cm) (p < 0.001). Reexcision was based on close/transected margins involving DCIS alone (6, 67%), infiltrating ductal carcinoma and DCIS (2, 22%), or infiltrating ductal carcinoma alone (1, 11%). Reexcision pathology demonstrated DCIS (3, 33%), no residual cancer (5, 55%), and 1 patient was lost to followup (1, 11%). Interestingly, cancer patients who required reexcision were younger (p = 0.022), but breast density was not associated with reexcision., Conclusions: To our knowledge, this is the first report of MRI-guided bracket wire localization. Patients with MRI-detected lesions less than 4 cm had clear margins at first excision; larger MRI-detected lesions were more likely to have close/transected margins. Reexcision was often because of DCIS and was the only pathology found at reexcision, perhaps because MRI is more sensitive for detecting invasive carcinoma than DCIS.

Objective: Our study aimed to correlate the dynamic contrast-enhanced MR appearance of infiltrating lobular carcinoma of the breast with histopathologic findings., Materials and Methods: We retrospectively reviewed the high-resolution, fat-suppressed and dynamic contrast-enhanced MR images of 13 of 20 women diagnosed with pathologically proven infiltrating lobular carcinoma of the breast. Twelve of the 13 women presented with breast symptoms and underwent mammography. Five of the women also had breast sonography. MR imaging was performed for evaluation of disease extent before the patients underwent modified radical mastectomy (n = 11) or lumpectomy (n = 2). Three experienced radiologists reviewed the MR scans. The tumor pattern types described on imaging were correlated with a detailed analysis of the pathology., Results: We found three patterns of infiltrating lobular carcinoma on MR imaging. The tumor pattern on imaging correlated with pathologic tumor morphology. We found the following patterns of infiltrating lobular carcinoma: a solitary mass with irregular margins (n = 4) that corresponded to the same appearance at pathology; multiple lesions, either connected by enhancing strands (n = 6) or separated by nonenhancing intervening tissue (n = 2), that correlated with the pathologic appearance of noncontiguous tumor foci, with malignant cells streaming in single-file fashion in the breast stroma or small tumor aggregates separated by normal tissue; and enhancing septa only, which were correlated with the histopathologic appearance of tumor cells streaming in the breast stroma (n = 1)., Conclusion: Infiltrating lobular carcinoma may be detected on MR imaging as solitary or multiple lesions that correspond to tumor morphology on pathologic examination. The appearance of multiple lesions or of enhancing fibroglandular breast elements on MR imaging is suggestive of infiltrating lobular carcinoma.

Purpose: To compare patients with visualized sentinel lymph nodes (SLNs) and patients with nonvisualized SLNs, with a focus on variables affecting SLN visualization at preoperative lymphoscintigraphy and on nodal drainage basins as related to tumor location., Materials and Methods: One hundred thirty-six patients who had breast cancer underwent preoperative lymphoscintigraphy before SLN biopsy. Patients with visualized and nonvisualized SLNs were compared for age; tumor site, size, and histologic findings; injection guidance method; diagnostic biopsy type; interval between biopsy and lymphoscintigraphy; intraoperative identification method; and surgical identification rate. Visualized SLN drainage basins were noted., Results: Ninety-nine patients had visualized and 37 had nonvisualized SLNs, without statistically significant differences in tumor site, size, and histologic findings; injection guidance method; diagnostic biopsy type; and interval between biopsy and lymphoscintigraphy. Ninety-nine (73%) of the 136 SLNs were visualized at lymphoscintigraphy; 30 (81%) of the 37 nonvisualized SLNS were identified at surgery. Of the seven SLNs not identified at surgery, five were mapped with radiocolloid only. Patients with nonvisualized SLNs were older than those with visualized SLNs. Eleven (46%) of 24 tumors with internal mammary drainage were in the outer part of the breast., Conclusion: Patients with and those without visualization differed in age, SLN identification at surgery, and surgical identification method. Nonvisualized status does not preclude axillary metastasis. In older patients with nonvisualized SLNs, blue dye may aid in SLN detection, as compared with isotope-only localization.

Ductal Carcinoma In Situ (DCIS) is the earliest form of ductal cancer, with a high rate of care if treated early. This article outlines the use of breast imaging in DCIS diagnosis, including mammography, MR imaging, and nuclear medicine studies. While MR imaging and nuclear medicine show great promise in DCIS diagnosis, mammography remains the mainstay of DCIS detection by the presence of microcalcifications in early tumors on the mammogram.

A three-dimensional (3)D spiral sequence was developed for dynamic breast magnetic resonance (MR) imaging with much improved image quality. Partial Z phase encoding was applied to obtain thinner slices for a coverage of the whole breast. Comparison between the 3D and a previously developed multi-slice 2D spiral sequences was performed on ten healthy volunteers without contrast and five breast patients with gadolinium-diethylene triamine pentaacetic acid (Gd-DTPA). The 3D spiral images had significantly less off-resonance blurring and spiral artifacts. With a small compromise on temporal resolution (7.7 seconds in 2D and 10.6 seconds in 3D), we obtained 32 interpolated 3-5 mm slices (with 20 Z phase encodes) for a full coverage of 10-16 cm breast with the same 1 x 1 mm2 in-plane resolution as the 2D sequence, which had 12 8-13 mm slices. Contrast between glandular and soft tissue in normal breasts was increased by about 25%. The reduced repetition time in the 3D spiral acquisition led to an increased Gd-enhanced signal. The difference between the enhancement of malignant and benign lesions increased by sevenfold. We expect that this new development could lead to improved specificity in characterizing breast lesions using MR imaging., (Copyright 2000 Wiley-Liss, Inc.)

The purpose of the study was to determine the sensitivity and specificity of various morphologic criteria in distinguishing malignant from benign breast lesions using a new sequence (3DSSMT) performed immediately after dynamic breast MRI. 3DSSMT combines a water-selective spectral-spatial excitation and an on-resonance magnetization transfer pulse with three-dimensional spoiled gradient-echo imaging. Morphologic features of 87 pathologically confirmed lesions were analyzed. The presence of either skin thickening, or a combination of a spiculated or microlobulated border, with a rim, ductal, linear, or clumped enhancement pattern was 94% specific and 54% sensitive for malignancy. Conversely, the presence of either a perfectly smooth border, a well-defined margin, non-enhancing internal septations, or a macrolobulated border was 97% specific and 35% sensitive for a benign diagnosis. In conclusion, delayed 3DSSMT discriminates a significant number of benign and malignant breast lesions; it has the potential to improve the diagnostic accuracy of dynamic breast MRI.