Amide Proton Transfer (APT) Imaging In Brain Glioma

BackgroundGlioma is the most common primary brain tumor, mainly characterized by significant heterogeneity of nerve pathology, genetics and prognosis. Though neurosurgery, radiotherapy and chemotherapy, etc technologies have made certain progress in recent years, the average survival period of glioma is still relatively low. It is very important for treatment and prognosis to make correct classification, grading diagnosis, comprehensive assessment of tumor extent and its relationship with the surrounding anatomical structure in the early stage. The significant heterogeneity of glioma leads to difference degree of tumor differentiation in different positions, moreover parts of the high grade glioma show no obvious enhancement with conventional MRI enhancement scanning. These factors will make significant affect when undergoing biopsy for pathology examination and radiographic grading. Amide proton transfer (APT) imaging is a new molecular imaging method based on chemical exchange-dependent saturation transfer (CEST), which can detect the amide protons of endogenous proteins and peptides. The imaging manifestions between different grades of glioma and the pathology basis of APT imaging have not been deeply investigation. APT imaging was used in this study to analyze the image feature of different grades of glioma, combined with the features of pathology and immunohistochemical staining with CD34and Ki-67expression, to discuss the diagnostic value of APT imaging in glioma and the correlation between APT signal intensity and tumor cellular density, microvessel density and proliferation index to define the pathological basis of APT imaging.ObjectiveTo study the feature of APT imaging in glioma, evaluate the ability of showing the heterogeneity of glioma and distinguishing different grades of glioma. To analyze the feature of pathology and immunohistochemical staining with CD34and Ki-67expression; and to investigate the correlation between APT signal intensity and pathologic features of different grades of glioma and their effects on APT signal.Materials and Methods1. Subjects34patients who investigated and confirmed to be glioma by postoperative pathology were divided into two groups,11cases in low grade (WHO Ⅰ-Ⅱ) and23cases in high grade (WHO Ⅲ-Ⅳ) group. Low grade group contains7cases of oligodendroglioma,2cases of astrocytoma and2cases of mixed neuronal-glial tumor. High grade group contains8cases of anaplastic astrocytoma,11cases of glioblastoma,2cases of anaplastic oligodendrocytes,1case of anaplastic oligoastrocytomas and1case of mixed neuronal-glial tumor, while8cases of them were recurrence.28cases of all were processed immunohistochemical staining with CD34and Ki-67, including11cases of low grade and17cases of high grade glioma. Low grade group contains7cases of oligodendroglioma,2cases of astrocytoma and2cases of mixed neuronal-glial tumor. High grade group contains5cases of anaplastic astrocytoma,11cases of glioblastoma,2cases of anaplastic oligodendrocytes,1case of anaplastic oligoastrocytomas.2. MRI examining methodsPhilips Aiehiva3.0T MR scanner was used as scanning equipment. APT scanning was performed after routine scanning sequence including T2-weighted, FLAIR, and pre-contrast T1-weighted imaging and before T1-weighted enhancement imaging to avoid interference effect. The maximal slice of the tumor was selected to undergo APT single slice scanning. Second-order shimming was used before APT scanning to reduce the effect of the inhomogeneity of the radio frequency (RF) field. The off-resonance RF pulse was used to presaturation.3. APT rate measurementUse Sun UltraSparc Station (Sun Microsystems, Mountain View, CA) software to analyze the original data of APT and IDL (Research Systems, Inc., Boulder, CO, USA) program to measure the APT rate and build pcolor.The APT rates of three ROIs (regions of interest) were measured, including the tumor core, edema and the CNAWM (the contralateral normal-appearing white matter), if there was not obverse edema area in low grade glioma, only the tumor core and the CNAWM were selected to measure. Paired-Sample T Test was used to compare the different of APT rates between the tumor core and edema, the tumor core and the CNAWM, and the different between edema and the CNAWM. Independent-Samples T Test was used to compare the different of APT rates in the tumor core between high and low grade glioma. The ROC curve was used to analyze the best critical value of APT rate in the tumor core to differential the grade of glioma, and the sensitivity and specificity of grading diagnoses were confirmed with this value.4. OperationAll the patients were underwent the tumor resection, and the samples of the ROIs which were selected before operation were taken for the pathology test. Then the tumors were resected using the microscope.5. Pathologic stainingThe specimens were processed with routine hematoxylin-eosin (HE) staining and immunohistochemistry for Ki-67and CD34. The Cellularity, CD34microvessel density and Ki-67labeling indices of tumor cores were measured, which then analyzed using Independent-Samples T Test to compare the different between low and high grade group. Spearman correlation analysis and linear regression analysis were used to find their relation with APT rate of the tumor core.ResultsStatistical discrepancies of the APT rate for the tumor core was present comparing with edema region and the CNAWM (P<0.01) in high grade glioma. And the APT rate in edema region was higher than in the CNAWM (P=0.013). In low grade group, statistical discrepancies of the APT rate for the tumor core was present comparing with the CNAWM (P<0.05). Statistical difference of the APT rate for the tumor core was present between high and low grade glioma, while the APT rate of high grade glioma was higher than low grade (P<0.01).The sensitivity of differential grades using T1-weighted enhancement was91.3%, while the specificity was80%. ROC curve showed the Cut-off value of APT rate for diagnosis of malignancy was2.172. Tumor with APT rate in the core greater than2.172was considered as the high grade glioma. And the sensitivity of differential grades was95.7%, while the specificity was81.8%.Statistical discrepancies of The APT rate, Cellularity, microvessel density and proliferation index in the tumor core were present comparing high grade and low grade glioma (P<0.01). The APT rate and Cellularity, microvessel density and proliferation index were positive correlation in the tumor core. The coefficient of correlation of r wsa respectively0.904(P<0.01),0.598(P<0.01), and0.750(P<0.01). Multiple linear correlative analysis showed that the increase in Cellularity (X1), microvessel density (X2) and proliferation index (X3) were the main factors for the increasing APT rate. The equation was as follows: Y=0.801+0.12X1-0.003X2+0.26X3, F=46.437, P=0.000, R2=0.853, and t value: Cellularity (6.610)> proliferation index (2.383).ConclusionThe APT rate in tumor core is higher than edema and the CNAWM. What’s more, the APT rate in edema is higher than the CNAWM in high grade glioma while the T1-weighted enhancement shows no enhancement, which has important significance in dividing tumor infiltration area. It has the role of guiding the resection range of tumor during the surgery, and the resection range can be appropriately expanded according to the increased APT value of the area.The APT rate is obviously higher in high grade glioma than in low grade, even in the cases that show no enhancement in conventional scanning. So we consider that the APT imaging is helpful for differential the grade of glioma. APT imaging combining with conventional scanning can upgrade the diagnostic accuracy and is contributing to improve doctors’ diagnostic level.The APT rate of the tumor core is positive correlated with cellularity, microvessel density and proliferation index, which means that the APT rate can exactly reflect the level of malignancy of glioma and growth characteristics in different region and grade. A high APT rate prompts the increasing of tumor cell density, microvessel density and highly proliferative. Multiple linear regression analysis shows cellularity contributes more to proliferation index, while microvessel density is of no significant effect on APT rate, which is further confirmed the basis of APT imaging, that APT signal is higher in the hypercellular areas, in other words, where the proteins and peptides are higher in the tissue.