The Value Of FMRI At 3.0 T MR On Differential Diagnosis Of Musculoskeletal Tumors: Associated With Pathology

ObjectivesTo explore the value of fMRI at 3.0 T MR in preoperative differential diagnosis of musculoskeletal tumors, the role of angiogenesis (involving MVD and VEGF as biomarkers) in biological behavior classification of musculoskeletal tumors, and to analyze the correlation of fMRI with angiogenesis of the musculoskeletal tumors.Materials and Methods1. Routine MRI, fMRI, surgeon and pathological diagnoses were carried out on forth-three patients with musculoskeletal tumors.2. Routine MRI scanning:The coronal T2WI, the axial T2WI, the axial T1WI and the axial FS T2WI were carried out on the patient sequentially at a 3.0 T MR scanner (GE Signa ExciteⅡ, Milwaukee, Wisconsin, U.S.A.) by using an 8-channel phased array surface coil or a QUADKENN coil.3. DWI:An SS-SE-EPI sequence was performed on the patient before and after dynamic contrast enhanced MRI (DCE MRI), the b value was 0 and 600s/mm2. The apparent diffusion coefficient (ADC) values of the tumors were measured.4. DCE MRI:A 3D-FAME sequence repeating 16 times was performed on the patient, the flip angle was 20°, the time span of a single phase was 18s-20s, and the total time span of the DCE MRI was 288s-320s. The contrast agent was injected at a rate of 3ml/s with a dose of 0.1mmol/kg at the end of the first phase. The time-intensity curve (TIC) type, the maximum slope of increase (MSI), the signal enhancement ratio (SER), the positive enhancement integral (PEI), the signal intensity of peak (SIpeak), the time to peak (Tpeak) were measured, and the relative MSI, the relative SER, the relative PEI, the delay of Tpeak (ΔTpeak), the signal enhanced extent (SEE), the early signal enhancement extent (ESEE) and the relative ESEE were calculated.5. Immunohistochemistry:The wax block and a 4μm slice were made out of the formalin fixed tumor specimens in the next day of the surgeon, consulting to ROI of the DCE MRI, assisting by the pathologist. Microvessel density (MVD) and vascular endothelial growth factor (VEGF) were selected as biomarker. The reagents were purchased from Beijing Zhongshan Golden Bridge Biotechnology Company. The immunohistochemistry was carried out using two-step immunohistochemical staining. PBS buffer was used insteading of primary antibody as a negative control and normal striated muscle specimen was used as a normal control. CD34 expressed in the cytoplasm and membrane of vascular endothelial cells, while VEGF expressed in the cytoplasm. To identify the MVD, the slice was first observed at a low power lens (×40) to find a'hot spots', and then at a high power lens (×400) to count the number of vessels. The VEGF stained cell proportion and the staining intensity were both concerned to determine a classification of VEGF.6. Statistic method:All data was processed by SPSS17.0 software. The categorical data was analyzed by Fisher exact test. The ADC value before and after DCE MRI were compared by the paired sample t test. The DCE MRI semi-quantitative parameters and MVD were normally tested by one sample Kolmogorov-Smirnov test. If the parameters satisfied the normal distribution, they were compared by independent sample t test or one-way ANOVA (pairwise comparisons among groups were undergone by LSD or Dunnett's T3 method). If not, a non-parametric test (Mann-Whitney U test in two independent samples or Kruskal-Wallis H test in multiple independent samples) was used. ROC curve was used to determine the threshold and the diagnostic reliability. Kendall's tau-b and Spearman correlation analysis were used in correlation test. P<0.05 was considered to be statistically significant. Results1. The pathological diagnoses were offered by QiLu Hospital Pathology Department. There were 9 benign tumors (3 fibrous dysplasias,2 neurilemmomas,1 aneurysmal bone cyst,1 chondroblastoma,1 angioleomyoma and 1 myofibroma),5 intermediate tumors (2 desmoid type fibromatosis,2 giant cell tumors,1 solitary fibrous tumors) and 29 malignant tumors(10 osteosarcomas,3 chordomas,3 malignant lymphomas,3 metastases tumors,2 malignant giant cell tumors,2 adult fibrosarcomas,1 malignant fibrous histiocytoma,1 leiomyosarcoma,1 myxofibrosarcoma,1 pleomorphic undifferentiated sarcoma,1 pleomorphic liposarcoma,1 alveolar soft part sarcomr).2. The ADC values before and after DCE MRI was satisfied normal distribution, they were 152.43±37.64 and 154.65±53.12 correspondingly and there was no significant difference between them. The ADC values of the benign, intermediate and malignant musculoskeletal group were 163.67±35.46,147.00±23.57 and 149.88±40.41 before DCE MRI, and 163.44±28.81,154.60±16.07 and 151.93±38.96 after DCE MRI. There were no significant differences among the ADC value of the benign, intermediate and malignant group in both cases of pre-enhancement and post-enhancement DWI.3. In 79.1%(34/43) cases of the patients the most significantly enhanced portion of the tumors were located in the largest cross-section of the tumor. There was significant difference among the TIC tpye of the benign, intermediate and malignant group. The Tpeak, the relative MSI, the relative PEI, theΔTpeak and the relative ESEE were significant different among the benign, intermediate and malignant group. The Tpeak, the relative MSI, theΔTpeak, the relative ESEE were useful in diagnosing malignant musculoskeletal tumors according to the ROC analysis; the sensitivity and specificity of them in diagnosing malignant tumors were 82.8% and 85.7%,86.2% and 78.6%,86.2% and 78.6%, and 86.2% and 85.7% correspondingly.4. The average MVD was 23.3±12.6 in all of the 43 musculoskeletal tumors, 10.0±7.7 in the benign group,30.6±10.9 in the intermediate group, and 26.1±11.5 in the malignant group. There were significant differences of MVD and VEGF expression among benign, intermediate and malignant group. There was significant difference between MVD of the VEGF weak positive group and the VEGF strong positive group.5. There was significant difference of MVD among typeⅡ, typeⅢand typeⅣTIC group. There were strong negative correlations between Tpeak andΔTpeak with MVD and strong positive correlationgs between the relative MSI and the relative ESEE with MVD. Tpeak, the relative MSI, the relative PEI,ΔTpeak and the relative ESEE showed significant differences between of VEGF weak positive and VEGF strong positive group.Conclusions1. Musculoskeletal tumors could be shown clearly using 3.0 T MRI with a body phased array surface coil or a QUADKNEE. A combination of the coronal FS T2WI, the axial T2WI, the axial T1WI, the axial FS T2WI and the DCE MRI was suggested as a standard MR examination.2. DWI was helpless for differential diagnosis of musculoskeletal tumors with a b value of 600s/mm2, while DCE MRI is helpful.3. MVD and VEGF represented angiogenesis of tumors; a large number of parameters of DCE MRI had strong correlation with the MVD and VEGF. DCE-MRI could reveal the angiogenesis in vivo.