Correlation Ananysis Of Dynamic Contrast-enhanced Magetic Resonance Angiography And Prongnostic Factors In Breast Cancer

1. PurposeMaximum intensity projection image (MIP) was obtained first based on the Phase I or Phase II dynamic scan subtraction image after dynamic contrast enhanced MRI. Then, Blood 3D MR vessel image was acquired after reconstruction. Peripheral vascular symptoms and the increased vascular status of ipsilateral breast in breast lesions were obsereved from this Blood 3D MR vessel image. Statistical analysis were conducted among peripheral vascular symptoms, the increased vascular status of ipsilateral breast, ACR MRI BI-RADS prescript term (just including size, burrs, and edge enhancement), and prognostic factors related to breast cancer, such as histopathological result of lesions (including imrnunohistochemical ER, PR, Her-2), ipsilateral axillary lymph node status. The study aimed to discover some correlations between dynamic contrast-enhanced magnetic resonance angiography results and prognostic factors of breast cancer.2. Materials and methods(1) Study subjectsFrom May 2012 and October 2014,93 female patients suspicious for malignant breast lesions based on breast X-rays, ultrasound or clinical examination were analyzed rretrospectively. They were checked for dynamic enhanced MRI in our hospital, for those who were premenopausal patients, try to control the checking time in the second menstrual cycles.All patients were conducted for core needle biopsy or surgical resection into one week after MRI examination, in which 59 cases were excluded due to benign lesions (n=35), surgical history (n=21), or history of neoadjuvant chemotherapy (n=3). The remaining 34 patients with a mean age±standard deviation of 41.62 ± 9.51 years (from 29 to 63 years) were included in the analysis. For the surgical options, breast-conserving surgery was chosed for 13 cases and modified radical mastectomy was chosed for 21 cases. Pathological type was invasive breast cancer, including 30 cases of non-specific type and 4 cases of intraductal carcinoma.(2) The scanning device and methodThe 3.0T large aperture superconducting MRI scanner was Siemens product (Magnetom verio, SIEMENS, Germany). Bilateral breast 8-channel phased-array surface coil was matched when scanning breast. The checking time was in the second menstrual cycles. Body portion 8 channel phased array surface coil was matched when scanning armpit. Scanning sequence was chosed for FSE T2WI, T1WI of tirm, and dynamic contrast-enhanced scan. T1WI sequence of 3D FLASH of axial was applied. DCE-MRI included 9 images, and they were before injection of contrast agent (mask), and continuous 8 times acquisition after injection. All images were processed and analyzed in Siemens workstation. MIP image was obtained by reconstructing the image after the first phase of enhanced image subtracted with mask, and mammary artery was analyzed based on this MIP image. After normal enhanced scanning, replace the body coil, and scan axial Vibe T1WI sequence, to observe bilateral axillary lymph nodes.(3) Analysis for breast MRI imagesThe validation was performed separately in a blind manner or without knowledge of the eventual clinical parameters by two experienced radiologists. Morphological characteristics and enhancement features of lesions were described by MRI BI-RADS terminology, and analysis was conducted just for burr-like and edge enhancement signs. Additionally, peripheral vascular symptoms and whole milk vessel status were evaluated by application of silhouette image and MIP image. Tumor size was got and measured in the maximum level of MR image. The adjacent vessel sign were Visually evaluated in the silhouette image, and arteries in strengthen lesions or contacted with the edge of the lesion were all considered as positive. Arteries, only close or bypass but not contacted with the lesion, or not in the lesion, were all excluded. Increased vascular in ipsilateral breast was assessed under Sardanelli methods, based on vessels number, length, and inner diameter of MIP images. The number of blood vessels with a length greater than 3 cm and diameter greater than 2mm in each breast was counted. The whole increased vascular of ipsilateral breast was defined when vessels in the parent breast increased more than two compared with contralateral normal breast. When differences between inter-observers occurred, the respective slides were reinvestigated jointly by both investigators. Rounded nodules whose short diameter was greater than 1cm and significantly enhanced were defined as positive lymph nodes.(4) Pathological analysisThe validation was performed separately in a blind manner or without knowledge of the eventual clinical parameters by two experienced pathologists. For tumor sizes measuring, the following principles were followed:If more than 2 cm, the maximum tumor cross section was measured as specimens; If less than 2 cm, the maximum tumor cross section was measured under microscope. Multifocal or multicentric breast lesions measurement was conducted as described above, just measuring the maximum size of the lesion. Histological types of breast cancer were on the basis of Bloom-Richardson histologic grading, with histological 1, well-differentiated; histological 2, differentiation medium; and histological 3, poor differentiation. For estrogen receptor (ER), progesterone receptor (PR), and oncogene expressions, Allred scoring system was applied. Scoring at 0-2 was thinked as ER, PR-negative, and scoring at 3-8 was thinked as ER, PR-positive. Expression of Her-2 was scored by the proportion and intensity of cell membranes staining (0,≥1,≥2,≥3). The Her-2-negative group was scored as 0,≥1, and≥2, while Her-2-positive group was≥3. Additionally, whether or not Axillary lymph node metastasis existed and the metastatic lymph node number were both recorded.(5) Statistical analysisSpss17.0 was used for statistical analysis. The MRI findings were compared with the histological results. Interobserver reproducibility for MRI features was calculated using unweighted kappa statistics. The kappa value for agreement was interpreted as follows: poor, less than 0.20; fair,0.21-0.40; moderate,0.41-0.60; good,0.61-0.80; and very good,0.81-1.00. The correlation between the MRI features and histopathological predictors analysis using chi-square (chi-square test), there were correlation between them while P< 0.05, and P= 1.000, said completely unrelated.3. Results(1) MRI resultsThirty-four breast cancer patients who were confirmed by surgery or core needle biopsy received dynamic contrast-enhanced magnetic resonance scans, six cancers were multifocal or multicentric nature. If there are multiple lesions, the diameter of the maximum one is standard. Diameters of their tumors ranges from 0.5 cm to 2.9 cm at last, the average diameter is 1.31 ± 0.37 cm, in which 12 cases exceed 2cm and 22 cases are less than 2cm. The amount of lesions with speculated margin is 22, with the edge of other states is 12 cases. The edges of 26 cases are found to be enhanced status, and ones of 8 cases are other forms. Kinetics of 3 patients is persistent type,22 platform types and 9 outflow types. The number of positive and negative peripheral vascular signs is 27 and 7 respectively. There are 15 cases whose breast blood vessels increased and 19 cases do not.(2) Histopathology and immunohistochemical resultsOut of 34 breast cancer cases,2 cases are poorly differentiated ductal carcinoma,1 moderately-differentiated,1 well-differentiated,and 30 non-specific type invasive cancer, including 3 cases of Grade Ⅰ-Ⅱ,1 with medullary,22cases of Grade Ⅱ, and 5cases of grade Ⅲ.25 of these cases are ER-positive (1%-90%), and 9 negative. As for PR,23 positive cases (2%-95%) and 11 negative ones.The number of Her-2 positive cases (3+) is 6, and negative cases 28. A total of 8 patients with axillary lymph node metastasis were found, no lymph node metastasis was found in the rest cases.(3) Statistical resultsInterobserver agreements between two radiologists were moderate.Through the correlation analysis between MRI features and histopathological predictors, we concluded:1). There were correlation between blood vessels increased and tumor size, lymph node metastasis, and ER, PR, and HER2 expression, while there was significantly positive correlation with tumor size, the greater the tumor the more of the blood vessels; In the tumors of larger than 2 cm, higher positive rate of the adjacent vessel sign than tumor less than 2 cm2). There were correlation between the adjacent vessel sign and tumor size, lymph node metastasis, and ER, PR expression except the HER2 expression3). Tumors with speculated margin correlated with the expression of ER4).Tumor rim enhancement characteristic was usually related to the tumor size and the expression of ER and Her-25). While strengthening curve to a certain extent, indicates of benign and malignant tumors, but in this study, there was no obvious correlation between strengthen curve and the different histopathological predictors4. ConclusionThe adjacent vessel sign and increased ipsilateral whole-breast vascularity were important predictors for the prognosis of breast cancer, which usually suggest a relatively poor prognosis.