Correlation Between BOLD Imaging And Angiogenesis, Invasion Of Rectal Cancer

Human colorectal cancer is one of the commonest malignant tumors, ranking second only to lung cancer as the leading cause of mortality in malignant tumors. There has been a marked increase in the incidence in recent years. Conventional MR scanning has a good parenchymal resolution at mulplanar and mulderection level which can better show the tumor form of rectal cancer and determine violations of the surrounding circumstances and metastasis, but can not evaluate the changes in organ function and the oxygen metaboilic level of the lesions. MR BOLD(Blood Oxygenation Level Dependent) imaging is different from the conventional MR scan. It can evaluate the oxygen metabolic level of the lesions by T2* and R2* changes, and make a quantitative and semi-quantitative analysis of oxygenation to help determine the sensitivity of radiotherapy and chemotherapy. It has brought about the change from morphological observation to the micro-analysis of metabolic and functional status and played an important role in diagnosis and the prediction and evaluation of curative effects. BOLD imaging was frequently used in brain functional imaging,it was mainly used in kidney in abdominal organs.With the development of MR technique,BOLD imaging allows a more efficient, higher temporal resolution, and it can be widely used in abdominal organs. The technique is applied to rectal cancer for the first time in this study in our country.The growth, metastasis and dissemination of malignant tumors can't do without its own blood supply, and the process of setting up blood supply is related to the increasing in the density of capiliaries. Invasion and metastasis are the most important and most essential biological characteristics of malignant tumors, and remain the main causes of death in patients with malignant tumors, and rectal cancer has no exception. Matrix metalloproteinase-2 (MMP-2) plays a crucial role during the course of tumor invasion and metastasis by promoting cancer cells to invade the surrounding tissues and the occurrence of metastasis by way of degrading the extracellular matrix. This experiment aims to provide a basis for the prediction of angiogenesis,invasion and metastatic ability in rectal cancer with BOLD imaging by studying the correlation between BOLD imaging parameters(R2* value) with MMP-2 expression, and MVD count of rectal cancer.Materials and Methods1. Clinical DataFrom June 2009 to December 2009, among treatments at the First Affiliated Hospital of China Medical University,63 cases clinically diagnosed with rectal cancer were examined by MR imaging. Fifrty-one cases which were confirmed by surgery within 2 weeks after diagnosis and had general access to the tumor specimens, as well as complete reports were involved in this study. There are 29 male cases and 22 female cases. Age ranges from 35 to 83 years old, with an average age of 60.7±10.4. Twelve cases of tumors were located in the upper rectum,23 cases at the middle rectum,12 cases at the lower rectum. All 51 cases were adenocarcinoma, including 12 well-differentiated,31 moderately differentiated,8 poorly differentiated cases. Clinicopathological stage was estimated by the Modified Duke Staging System. The standard is as follows:Modified Duke A, the tumor penetrates into the mucosa of the bowel wall but no further. Modified Duke B, B1:tumor penetrates into, but not through the muscularis propria (the muscular layer) of the bowel wall. B2:tumor penetrates into and through the muscularis propria of the bowel wall. Modified Duke C, C1:tumor penetrates into, but not through the muscularis propria of the bowel wall; there is pathologic evidence of rectal cancer in the lymph nodes. C2:tumor penetrates into and through the muscularis propria of the bowel wall; there is pathologic evidence of rectal cancer in the lymph nodes. Modified Duke D, the tumor, which has spread beyond the confines of the lymph nodes (to organs such as the liver, lung or bone). Among the 51 cases, there are 6 Duke A,15 Duke B,23 Duke C, and 7 Duke D cases.2. Image post-processing and analysis(1) MR BOLD scanningFast for six hours, administrate 20mg 654-2 (Raceanisodamine Hydrochloride Injection, 10mg/ml, shuanghe pharmaceutical,Ltd) orally 40 minutes before scanning. Free breath is permitted in the scan with partial bladder filling. The equipment are GE SignaHDx 3.OT MR scanner and 8 channel phased array superficial coil.Make patients lie on back to perform abdominal pelvic MR scan and inflate rectum with 200～600ml genial normal saline enema first of all(regarding the paitient's tolerability). After that, perform conventional T2WI MR image scanning at axial, coronal and axes planar and T1WI MR image scanning at axial planar.Then choose the maximum level of disease as the target two layer of BOLD imaging at axial and coronal or axes planar, and obtain 48 BOLD images altogether Specific scan parameters shown in Table 1.(2) MR BOLD image post-processing and analysisMR BOLD images will be sent to the GE image post-processing workstations(AW4.4 workstation). First, use pre-installed software--R2Star to generate R2* image, evaluate the BOLD image quality level,and calculate the SNR of BOLD image using the neighboring fat signal intensity as background noise; Then collect the R2* values of the lesions,calculate the inter-observer and inner-observer differences. ROI region needs to be selected from specific solid part with multi-point averaging acquisition3. Pathology specimens processing method and immunohistochemical analysis.Preparation of histopathological sections, staining were done under the guidance of teachers from Department of Pathology.(1) slice productionMake 4 serial slices of the selected cases of paraffin-embedded tissues with 4μm each slice. The first slice was proceed with the HE staining, and the remaining three were for immunohistochemical staining.(2) immunohistochemical analysisPeroxidase labeled streptomycin-avidin method (SP) was used. CD34, VEGF, MMP-2 monoclonal antibody and SP immunohistochemical kit were purchased from the Jinqiao Biotechnology Ltd, Beijing. Make known colorectal cancer-positive biopsy as positive control, PBS buffer as negative control.(3) Judgement of immunohistochemical resultsMVD (microvascular density, MVD) count:MVD unit area is defined as tissue stained with CD34-positive microvascular cells. Cancer isolated brown with vascular endothelial cells or cell clusters on behalf of an individual microvessel. First view the whole film at low magnification(100×), larger microvessel density to identify "hot spots" (hot spot) with high-power microscope (200×), each slice records five "hot spots". Choose the average number of cases as the MVD.criterion for MMP-2:Mainly positive for microscopic tumor cells, vascular endothelial, mesenchymal appear brown, yellow or particle. mass. Criteria:First observe a comprehensive cross-section with the low-fold mirror (40×) to select the strongest MMP-2 staining, then count 100 cells at 200 times magnification. Positive staining cells<10% is for the negative (--); 10～25% for weakly positive (+); 26～50% of (++);> 50% positive for the strong positive (+++)。4. Statistical treatment of experimental dataAll data were analysed with SPSS 13.0 and MedCalc statistical software. Measurement data is expressed with mean±standard deviation. For the normal distribution, homogeneity of variance of the measurement data, use t test or paired t test to compare between two groups, use one-way ANOVA analysis to compare among more groups. For the measurement data of skewed distribution or grade data, use non-parametric tests:comparison between two groups by Wilcoxon test, and multi-group comparison by Kruskal Wallis test. For the correlation analysis of parameters use Pearson or Spearman rank correlation analysis. Difference has stati-stical significance with p value<0.05 as the standard.Results1. Tumor morphology can be observed on BOLD images in this study and the scope of lesions were determined combining with unenhanced images.The image signal to noise ratio (SNR) was 25.37±3.41; All the images reached measurement requirements via quality grading. Tumor R2* values can be measured on BOLD image or composite image of R2* map and BOLD image.2. In the 51 cases of rectal cancer, R2* and T2* values at MR BOLD images in rectal cancer fluctuated within a certain range:R2* in the range of 19.16 43.07Hz,while T2* in the range of 24.98～52.81ms. The average value of R2* was 26.23±5.28Hz, T2* 40.52±6.72ms。3. the correlation analysis between the R2* values on BOLD images and the MVD counts,MMP-2 expressionThe difference in the MVD counts between MMP-2 positive groups and negative group had statistical significance (P<0.05).Tumor parenchyma R2* values in DukesA were lower than that of any other clinical stage (P<0.05); they had no relation with tumor pathological grade and lymph node metastasis. R2* values and the MVD count of the tumor parenchyma were positively correlated (P＜0.05); tumor R2* values had no significant difference among all intensity groups of MMP-2 expression.Conclusion1. BOLD imaging of rectal cancer has high image quality, the R2* value of lesions can be accurately measured.2. BOLD imaging can reflect the tumor oxygen metabolism, indirectly reflects the characteristics of blood flow, which lays the foundation for further tumor microvascularity study in order to identify the correlation between tumor R2* values and microvessel density and immunohistochemical indicator of MMP-2 as well as evaluating tumor invasion and metastasis.3. Tumor parenchymal R2* values were positively correlated with MVD counts. BOLD MRI may be used as a imaging method of evaluating tumor angiogenesis of rectal cancer in vivo...