DCE-MRI、3D-ASL与DSC-PWI在脑胶质瘤术前分级诊断中的应用研究

发布时间：2018-04-23 17:30

  本文选题：脑胶质瘤 + 磁共振灌注成像　； 参考：《郑州大学》2017年硕士论文

【摘要】：背景和目的脑胶质瘤(brain glioma,BG)是最初起源于神经上皮组织胶质细胞的肿瘤,为最常见的脑部原发性恶性肿瘤,约占所有颅内肿瘤的50%以上。BG可以发生于任何年龄阶段,儿童主要发生于颅底和中线区,多为毛细胞星形细胞瘤和室管膜瘤等;成人则以大脑半球肿瘤多见,如星形细胞瘤、胶母细胞瘤、室管膜瘤等。BG的临床表现可分为以颅内压增高引发的头痛、恶心、呕吐及视乳头水肿等症状和以神经功能损害引起的肢体运动障碍、偏盲及癫痫等两大类症状。不同级别的BG的治疗方案和临床预后差别很大,因此,手术前准确判断BG的病理级别对治疗方案的设计、手术计划的确定,术后放化疗措施的制定及评估预后有显著的临床意义。患者一旦出现临床症状后,就诊时常规检查包括头颅计算机断层扫描(Computed Tomography,CT)和核磁共振成像(Magnetic Resonance Imaging,MRI)。头颅CT只是初步判定是否存在着颅内占位,而MRI在显示BG的特点和性质方面明显优于CT检查,故MRI已经成为评价BG的首选影像学检查方法并已获得临床上的认可,而BG的术前准确分级往往只依靠常规MRI检查往往是不够的。肿瘤的恶性程度与微血管结构、肿瘤细胞增殖性有很大的关系,所以准确地评估BG血管生成情况对鉴定其恶性程度有着重要的意义。由于BG具有富血管性,其血管结构和正常脑组织血管结构有明显的差异,表现在固有脑血管与新生血管两方面;BG的新生血管具有密度大、严重异型扭曲以及血管内径均匀性差的特点,因此,新生血管的活跃程度是评价BG恶性程度的重要指标,这就给磁共振灌注成像(perfusion weighted imaging,PWI)对BG的诊断、分级及鉴别诊断提供了理论基础。基于示踪对比剂原理的不同,目前临床上应用的灌注技术分为:经静脉快速团注外源性对比剂后,根据其影响纵向驰豫的是T1-加权动态对比增强(dynamic contrast-enhanced,DCE)成像技术(DCE-MRI)以及影响横向驰豫的是T_2*/T_2加权动态磁敏感对比(dynamic susceptibility contrast,DSC)灌注加权成像(DSC-PWI);不需要团注外源性示踪剂、以水质子作为自身内源性示踪剂的动态自旋标志灌注加权成像(ASL-PWI)技术。本研究通过对比DCE-MRI、DSC-PWI与3D-ASL三种灌注技术的各种参数在BG病理分级中的价值,探讨这三种技术的应用优势及不足之处,为BG的术前诊断和病理分级提供新的思路。资料与方法1研究对象:搜集纳入郑州大学第一附属医院从2015年9月至2016年5月期间内首发的临床资料完整的经常规影像学检查发现脑实质内占位病变的患者为研究对象。术后证实为BG患者82例,其中行DCE-MRI扫描的共45例,DSC-PWI扫描的共37例,所有患者均行3D-ASL扫描。术后病理根据2016 WHO CNS肿瘤新分类标准进行了分级。所有患者均在郑大一附院神经外科手术,获得的手术标本由本院病理科两位资深病理专家共同确定BG种类及分级,其中Ⅰ级和Ⅱ级为低级别胶质瘤(low grade glioma,LGG),Ⅲ级和Ⅳ级为高级别胶质瘤(high grade glioma,HGG)。使用德国Siemens公司3.0T Prasima超导磁共振扫描仪和64通道专用头颈联合专用线圈及配套专用高压注射器。MRI对比剂为钆喷酸葡胺注射液(Gd-DTPA),商品名:马根维显,注射速度均为4ml/s。2方法扫描序列:1 常规MRI平扫:T1WI、T_2WI、T_2flair轴位、T1WI矢状位、DWI;2 3D-ASL-PWI;3 DCE-MRI或DSC-PWI序列检查。(4)增强后T1FLAR轴矢冠扫描。3图像的处理将获得的所有原始的灌注数据导入Siemens Syngo.via后处理工作站,采用MRI Perfusion软件进行后处理。由三位高年资磁共振科医师分别选择ROI,进而得到兴趣区域的定量值。4数据处理本研究82例BG患者的灌注成像资料和数据应用了SPSS 22.0软件包进行统计学分析,实验数据采用均数±标准差((x|-)±s)表示,多组间参数的比较采用单因素方差分析,组间两两比较,采用LSD-t(Least—Significant Difference)检验。最后采用受试者工作特征(ROC)曲线进行统计学分析,获得HGG与LGG的最佳灌注鉴别诊断阈值及相应的敏感性、特异性、准确度,用以判断相应的诊断效能。结果1、扫描DCE-MRI的45例患者中,HGG的平均Ktrans值与Ve值均明显高于LGG(P0.01)。根据单因素方差分析显示平均Ktrans及Ve值在Ⅱ级和Ⅲ级、Ⅱ级和Ⅳ级之间均有着显著的统计学差异(P0.01),而Ⅲ级和Ⅳ级之间的没有统计学意义(P0.05)。2、扫描DSC-PWI的37例患者中,HGG的r CBV与r CBF值均明显高于LGG(P0.01)。根据单因素方差分析Ⅱ级和Ⅲ级、Ⅱ级和Ⅳ级,Ⅲ级和Ⅳ级之间的r CBV和r CBF值均具有明显的统计学差异(P0.01)。3、扫描3D-ASL的82例患者中,在ASL灌注成像的脑血流图中,肿瘤各项相对血流量灌注指数TBFmax/对侧半球、对侧白质、对侧灰质的CBF在高、低级别BG之间有着显著的统计学的差异(P0.01)。4、同时扫描DSC与ASL的37例患者中,在ASL与PWI灌注成像的脑血流图中,肿瘤各项相对血流量灌注指数TBFmax/对侧半球、对侧白质、对侧灰质的CBF之间没有的统计学意义(P0.05)。5、绘制HGG与LGG的Ktrans值、Ve值、r CBV和r CBF的ROC曲线,可分别获得的4个的AUG分别是:0.912、0.879、0.999、0.997。依据约登指数,可获得Ktrans、Ve、r CBV和r CBF鉴别HGG与LGG的最佳阈值分别为0.059/min-1、0.357、2.152、2.11;根据此阈值获得各个参数意见别HGG与LGG的敏感度和特异性,各个参数鉴别的敏感度分别为96.4%、93.7%、99.4%、97.4%;特异性分别为84.75%、73.2%、98.2%、95.2%。结论1、三种MRI灌注技术均能准确获得肿瘤微血管的生长情况和血流动力学改变信息,对BG术前病理分级诊断有重要指导意义。2、DCE-MRI技术中测量出的Ktrans值、Ve值可以作为区分BG级别的定量指标。但二者在鉴别Ⅲ级和Ⅳ级BG间不具有统计学差异,这说明了Ⅲ级和Ⅳ级BG的血管透通性相近。3、DSC-PWI技术中测量出的r CBV、r CBF值与BG的级别密切相关,这可以说明r CBV值与r CBF值可以作为区分BG级别的定量指标。4、3D-ASL与DSC-PWI技术具有一致的灌注效果,由于ASL安全无创、不用对比剂、可重复多次检查,所以在BG肿瘤分级诊断中3D-ASL值得推广。5、将BG的病理级别和不同的灌注参数进行关联比较时,会发现r CBV是区分HGG与LGG最敏感的参数。
[Abstract]:Background and objective brain glioma (BG), the most common primary malignant tumor of the brain, is the most common primary malignant tumor of the brain, which is the most common primary malignant tumor of the brain. More than 50% of.BG in all intracranial tumors can occur at any age stage. Children mainly occur in the cranial and midline areas, mostly hair cell astrocytomas and ventricular tubes. The clinical manifestations of.BG in adults, such as astrocytoma, gellonoma, and ependymoma, can be divided into two major symptoms, such as headache, nausea, vomiting and papillematous edema caused by increased intracranial pressure, Limb Dyskinesia caused by neurologic impairment, hemiblindness and epilepsy. The treatment plan and the clinical prognosis of BG are very different. Therefore, it is significant to determine the pathological grade of BG before the operation for the design of the treatment plan, the determination of the operation plan, the formulation of the postoperative radiotherapy and chemotherapy measures and the evaluation of the prognosis. Puted Tomography, CT) and nuclear magnetic resonance imaging (Magnetic Resonance Imaging, MRI). Skull CT only preliminarily determines whether there is an intracranial space occupying position, and MRI is obviously superior to CT examination in showing the characteristics and properties of BG, so MRI has become the first choice imaging examination method and has obtained clinical approval, and the preoperative accuracy is accurate. The grade often depends on the conventional MRI examination. The malignancy of the tumor is closely related to the microvascular structure and the proliferation of the tumor cells. Therefore, it is of great significance to evaluate the angiogenesis of BG accurately for the identification of its malignancy. As BG is rich in blood vessel, its vascular structure and normal brain tissue have a vascular structure. The obvious difference is manifested in two aspects of the intrinsic cerebral vessels and the neovascularization; the BG's neovascularization has the characteristics of large density, severe abnormal pattern distortion and poor homogeneity of the blood vessel diameter. Therefore, the activity of the neovascularization is an important index for evaluating the degree of BG malignancy, which gives the perfusion weighted imaging (PWI) to BG. Diagnosis, classification and differential diagnosis provide a theoretical basis. Based on the principle of tracer contrast agent, the clinical application of perfusion technique is divided into: after intravenous rapid injection of exogenous contrast agent, the effect of the longitudinal relaxation is T1- weighted dynamic contrast enhancement (dynamic contrast-enhanced, DCE) imaging (DCE-MRI) and influence The transverse relaxation is T_2*/T_2 weighted dynamic magnetic sensitivity contrast (dynamic susceptibility contrast, DSC) perfusion weighted imaging (DSC-PWI); the dynamic spin marker perfusion weighted imaging (ASL-PWI) technology is not needed for the injection of exogenous tracers and water protons as their own endogenous tracer. This study compares DCE-MRI, DSC-PWI and 3D-ASL three. The value of various parameters of the perfusion technique in the BG pathological classification, to explore the advantages and disadvantages of the three techniques, and to provide new ideas for the preoperative diagnosis and pathological classification of BG. Data and methods 1 subjects were collected and included in the First Affiliated Hospital of Zhengzhou University from September 2015 to May 2016. The whole routine imaging examination found that the patients with cerebral parenchyma occupying lesions were studied. 82 cases of BG patients were confirmed after operation, including 45 cases with DCE-MRI scan, 37 cases of DSC-PWI scan, all patients underwent 3D-ASL scan. The postoperative pathology was classified according to the new classification criteria of 2016 WHO CNS tumor. All patients were in Zheng Dayi. The surgical specimens in the Department of neurosurgery were determined by two senior pathologists in the Department of pathology in our hospital together with the BG type and classification, among which grade I and grade II were low grade glioma (low grade glioma, LGG), grade III and IV were high grade gliomas (high grade glioma, HGG). The German Siemens company 3.0T Prasima superconducting magnetic resonance scan was used. The drawing instrument and 64 channel special head and neck joint special coil and matching special high pressure syringe.MRI contrast agent are Dimeglumine Gadopentetic Acid Injection (Gd-DTPA), commodity name: Ma root dimension show, the injection speed is 4ml/s.2 method scanning sequence: 1 conventional MRI plain scan: T1WI, T_2WI, T_2flair axis, T1WI sagittal, DWI; 2 3D-ASL-PWI; 3 DCE-MRI or DSC-PWI sequence inspection (4) all the original perfusion data obtained by the enhanced.3 image of the enhanced T1FLAR axis sagittal scan were introduced into the Siemens Syngo.via post processing workstation and the MRI Perfusion software was used for post-processing. ROI was selected by three senior senior magnetic resonance physicians respectively, and then 82 cases of BG suffering from the quantitative value.4 data processing in the region of interest were obtained. The perfusion imaging data and data were applied to the SPSS 22 software package for statistical analysis. The experimental data were represented by mean number + standard deviation (x|-) + s. The comparison of multiple groups of parameters was compared with single factor analysis of variance. 22 comparison between groups was compared with LSD-t (Least Significant Difference) test. Finally, the working feature (ROC) curve of the subjects was used. Statistical analysis was carried out to obtain the optimal threshold of HGG and LGG perfusion differential diagnosis and corresponding sensitivity, specificity and accuracy to determine the corresponding diagnostic efficiency. Results 1, the average Ktrans value and Ve value of HGG were significantly higher than that of LGG (P0.01) in the 45 cases of scanning DCE-MRI. The average Ktrans and Ve values were shown to be in II. There were significant statistical differences between grade II and grade II and grade IV (P0.01), but there was no statistical significance between grade III and IV (P0.05).2. In 37 patients with DSC-PWI, HGG R CBV and R CBF values were significantly higher than LGG (P0.01). The values of V and R CBF have significant statistical differences (P0.01).3. In the 82 patients with 3D-ASL, the relative blood flow perfusion index TBFmax/ to the lateral hemisphere, the contralateral white matter, the contralateral gray matter CBF is high, and the low grade BG has a significant difference (P0.01).4, simultaneously scanning the ASL perfusion imaging in the cerebral blood flow chart of ASL perfusion imaging. In the 37 patients with L, the relative blood flow perfusion index TBFmax/ to the lateral hemisphere, the contralateral white matter and the CBF of the side gray matter (P0.05).5 was not found in the cerebral blood flow chart of the perfusion imaging of ASL and PWI, and the Ktrans value of HGG and LGG, Ve, and the 4 of them were respectively: 79,0.999,0.997. can obtain Ktrans, Ve, R CBV and R CBF to identify HGG and LGG as 0.059/min-1,0.357,2.152,2.11, respectively. According to this threshold, the sensitivity and specificity of individual HGG and LGG are obtained, and the sensitivity of each parameter identification is 96.4%, 93.7%, 99.4%, 97.4%; the specificity is 84.75%, 7, respectively. 3.2%, 98.2%, 95.2%. conclusion 1, three kinds of MRI perfusion techniques can accurately obtain the growth of tumor microvessels and the information of hemodynamic changes. It has important guiding significance for the diagnosis of pathological grading before BG, Ktrans value measured in DCE-MRI technology, Ve value can be used as a quantitative index to distinguish BG grade, but the two is to distinguish grade III and grade IV BG. There is no statistical difference. This shows that the vascular permeability of grade III and grade IV BG is similar to that of.3. The R CBV and R CBF value in DSC-PWI technology are closely related to the level of BG. This shows that the R CBV value and R CBF value can be used as a quantitative index to distinguish the grade. Creation, without contrast agent, can be repeated multiple times, so in the BG tumor grading diagnosis, 3D-ASL is worth promoting.5. When comparing the pathological grade of BG with the different perfusion parameters, it will be found that R CBV is the most sensitive parameter to distinguish between HGG and LGG.

【学位授予单位】：郑州大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R445.2;R739.41
，

本文编号：1792919