In situ and minimally invasive breast cancer: morphologic and kinetic features on contrast-enhanced MR imaging

PURPOSE: This retrospective study was undertaken to investigate the morphologic and dynamic features of in situ and minimally invasive breast cancer on contrast-enhanced (c.-e.) MR imaging and to examine possible associations to pathology features. MATERIAL AND METHODS: A total of 71 patients underwent MR imaging. T1-weighted FLASH-3D images were obtained before and after intravenous administration of Gd-DTPA. Histopathologic analysis of 78 lesions revealed ductal carcinoma in situ (DCIS) n = 50 and DCIS with microinvasion n = 28. MR features were correlated with histopathologic findings. RESULTS: Enhancement in DCIS was focal (73%), diffuse (10%) or ductal (17%). No enhancement occurred in two cases (4%). In 65% enhancement speed was classified as delayed. There was a tendency toward a more ill-defined (83 vs. 43%) enhancement pattern in high grade DCIS and a more ductal (29 vs. 12%) and faster (50 vs. 29%) enhancement in comedo type DCIS. However, significant differences in the enhancement behaviour could neither be demonstrated between high grade and non high grade DCIS nor between comedo and non comedo type DCIS. No significant differences were noted between pure and microinvasive DCIS. CONCLUSION: In this retrospective analysis the majority (96%) of DCIS lesions show contrast enhancement. However, in only about 50% of DCIS the criteria of a so-called 'typical' enhancement behaviour was fulfilled, that means strong, early, focal ill-circumscribed or ductal. Enhancement that follows a duct is often associated with malignancy, however this feature was only present in 17% of the cases. c.-e. MR imaging allowed the detection of 25 additional foci of DCIS. Therefore malignant in situ lesions can be present with atypical enhancement, and should be taken into consideration in high-risk patients in particular.     

Dynamic contrast-enhanced MR imaging of the entire breast with spectral-selective inversion fast three dimensional sequence

Dynamic contrast-enhanced images with high spatial and temporal resolutions were acquired with a fast 3D spoiled gradient echo (SPGR) sequence using spectral selective inversion recovery (IR) pulse. Five healthy volunteers and 12 patients with 14 pathologically proven breast lesions were studied. Fat suppressed volume image data covering the entire breast were obtained with a sufficient spatial resolution (0.9×1.5×3.0 mm³) and an imaging time of 57 s. By using the criteria including peripheral enhancement and presence of spiculation, sensitivity, specificity, and accuracy in detecting malignant lesions were 88.9, 80.0 and 85.%, respectively. Although the C/N and S/N ratios were approximately 30% less than those of the conventional fat suppressed 3D technique, fast 3D SPGR imaging with spectral IR method demonstrated sufficient image quality for both time intensity analysis and morphological evaluation of the breast lesions with a data acquisition time less than half of the conventional method. This technique can substantially improve spatial and temporal resolutions of dynamic MR images of the breast and will be useful in evaluating malignant and benign breast lesions.     

Retrospective analysis for evaluation of the value of contrast-enhanced MRI in patients treated with breast conservative therapy

Purpose The aim of this study was to assess the value of contrast-enhanced (c.-e.) MRI in the follow-up of patients with conservatively treated breast cancer since detection and exclusion of malignancy may interfere significantly with posttherapeutic changes within the treated breast. Material and methods A total of 207 patients with a history of limited surgery and radiation therapy underwent MR imaging, 40 patients were examined 0–12 months and 167 patients were examined later than 12 months after radiotherapy. Suspicious or indeterminate findings were suggested by clinical examination or conventional imaging in 80 studies. In 127 women, MRI was performed within breast tissue that was difficult to assess due to scarring or dense breast tissue. Results Recurrent carcinoma was confirmed in 27 patients by surgical biopsy. All 27 carcinomas, except for one with a slow signal increase, demonstrated early rise of signal intensity on dynamic T1-weighted contrast enhanced images. During the first year after therapy, the diagnostic accuracy could not be improved by additional use of c.-e. MRI. Differentiation between posttherapeutic changes and recurrent carcinoma was frequently not possible because of strong and sometimes early and ill-circumscribed enhancement. Later than 12 months after therapy enhancement decreased significantly, thus the false positive calls could be reduced from 49 (conventional imaging) to 12 (conventional imaging plus MRI). A total of 12 of 26 recurrences and multifocality in 4/5 cases were diagnosed by MR imaging alone at this time interval. Conclusion In the first year after therapy, c.-e. MRI is only indicated in selected cases. The results later than 12 months emphasize that c.-e. MRI may contribute significant additional information. It allows better distinction of posttherapeutic fibrosis from recurrent carcinoma and proved to be able to detect recurrent disease more sensitive and at an earlier stage.     

Rat cerebral ischemia induced with photochemical occlusion of proximal middle cerebral artery: a stroke model for MR imaging research

A stroke model in rats with photochemically induced thrombosis (PIT) of proximal cerebral middle artery (MCA) is introduced for magnetic resonance imaging (MRI) study. Thirty-seven rats subjected to surgical and optical procedures for inducing the PIT models were scanned using a 1.5-T scanner with T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), diffusion-weighted imaging (DWI), and contrast-enhanced perfusion-weighted imaging (PWI) at 1 h and 24 h after MCA occlusion. The penumbra evolution and PWI-derived parameters including relative cerebral blood volume (rCBV) and relative cerebral blood flow (rCBF) were monitored; and the relative lesion size (RLS) was compared with the final RLS on the gold standard triphenyl tetrazolium chloride (TTC) staining at 24 h. The results showed that the focal cerebral ischemic lesions were detectable in all rats with different MR approaches. The lesion on PWI at 1 h and on all MR images at 24 h was matched well with that seen on TTC staining; the peri-infarct area decreased from 6.2 ± 7.2% of the brain volume at 1 h to 0.3 ± 5.6% at 24 h. Compared to that in the contralateral hemisphere, rCBV in ischemic region was 52.6 ± 21.4 and 40.0 ± 15.8% (p > 0.05), and rCBF was 64.6 ± 11.2 and 47.3 ± 11.1% (p < 0.05) at 1 h and 24 h respectively. The present PIT model in rats has been successfully adopted for MRI research, which might be feasible for certain stroke studies and should be beneficial for the evaluation on effects of potential diagnostic and therapeutic approaches.