基于MR灌注成像的脑胶质瘤微血管管径评价及其应用价值研究

发布时间：2018-05-11 14:05

  本文选题：神经胶质瘤 + 磁共振成像　； 参考：《第三军医大学》2014年硕士论文

【摘要】：胶质瘤是最常见的原发脑肿瘤，根据2007年WHO中枢神经系统分类，分为低级别胶质瘤(WHOⅠ、Ⅱ级)及高级别胶质瘤(WHOⅢ、Ⅳ级)。不同级别及类型的胶质瘤具有不同的生物学行为及预后，为临床诊治研究的重点。研究证明，肿瘤微血管的生成和病理分级密切相关，已作为判断预后的重要因素。磁共振灌注成像(perfusionweighted magnetic resonance imaging，PWI)是一种功能性成像手段，可以术前评估活体组织的微血管分布及血流灌注情况。活体检测胶质瘤的微血管生成情况不仅有助于术前分级、指导临床治疗方案，更有利于监测抗肿瘤药物的疗效。其中，脑血容量(cerebral blood volume,CBV)已广泛应用于临床，并对胶质瘤术前分级显示出一定的优势。近年来，动物实验结果表明基于MR灌注成像的血管管径指数(vessel size index, VSI)可以非侵入性地定量描述病变的血管重构，VSI也称微血管结构成像(vesselarchitectural imaging，VAI)，优于血容量(CBV)、血流量(CBF)等提供间接信息的指标，但该技术应用于临床仍处于摸索阶段，缺乏病理验证。本研究利用SE-GE-EPI灌注成像技术测量人脑胶质瘤的微血管管径，与CBV结果作比较，并与病理学免疫组化结果相对照，分析不同级别脑胶质瘤微血管生成的特点及PWI-VSI测量的准确性，明确微血管管径与胶质瘤分级的相关性。 目的： 研究采用磁共振灌注成像(MR-PWI)测量胶质瘤微血管管径的有效性、准确性，分析不同级别胶质瘤微血管的组织学特点，并探讨胶质瘤血管大小指数(VSI)、血容量(CBV)与胶质瘤组织学分级的相关性，以期提高MRI对胶质瘤术前分级评估的准确性及为临床监测胶质瘤药物治疗效果提供更多的信息。 材料与方法： 1.研究对象 胶质瘤组：经病理证实的35例胶质瘤患者。男15例，女20例，，年龄18~69岁,平均47.26±13.71岁。所有病例为首诊病例，未接受过任何侵入性或非侵入性治疗。根据WHO2007年中枢神经系统肿瘤分类及分级方法，低级别肿瘤10例，其中星形细胞瘤7例(包括6例弥漫性星形细胞瘤及1例肥胖型星形细胞瘤)，少突胶质细胞瘤3例；高级别肿瘤25例，其中间变性少突胶质细胞瘤5例，间变性少突星形细胞瘤1例，间变性星形细胞瘤3例，胶质母细胞瘤16例。 正常对照组：10例，取自切除胶质瘤标本的周围正常脑组织。 2.扫描设备及数据测量 MR成像仪器：SIEMENS lagnetom Verio3.0T，完成常规磁共振扫描及SE灌注扫描，嘱患者次日24小时后行SE-GE-EPI灌注，采用GE SIGNA HDX1.5T超导型MR扫描仪，八通道线圈。分别测定肿瘤区VSIMRI最大值、VSIMRI平均值、最大血容灌注值(CBVmax)。测定对侧正常脑白质病变区VSIMRI值、CBV值。 3.病理切片染色及测定指标 将病理确诊为胶质瘤的35例患者的所有石蜡标本(每名患者2-8块)各制备一张切片，分别进行HE染色及CD34免疫组织化学染色，利用计算机图像分析软件(Image-ProPlus5.0)定量分析胶质瘤微血管密度、形态(包括微血管长径、短径)。 4.数据分析 采用皮尔逊相关分析评价VSIMRI、CBV值与微血管长径、短径两两之间的相关性。对不同级别胶质瘤的VSIMRI最大值、VSIMRI平均值、rCBVmax值行Mann-Whitney U检验，P0.05为差异有统计学意义。采用受试者工作特征(ROC)曲线分析VSIMRI最大值、VSIMRI平均值、rCBVmax值在胶质瘤分级诊断中的价值并确定其最佳诊断界值和取该界值时的灵敏度、特异度。 结果： 1.不同级别胶质瘤的免疫组化表现：CD34标记微血管呈棕黄色，低级别胶质瘤(II级)的微血管管径较均一，血管管径较小(长径18.93±5.11μm，短径9.62±2.68μm)，稍大于正常脑组织微血管(长径8.79±5.10μm，短径5.51±2.63μm)。在低级别胶质瘤中，少突胶质瘤血管管径(短径8.06±0.51μm)普遍小于弥漫性星形细胞瘤(短径10.41±3.03μm)，而MVD则大于后者。高级别胶质瘤(III、IV级)的微血管密集，管径相对较大(长径43.59±13.26μm。短径33.36±11.43μm)，尤其胶质母细胞瘤有更多畸形血管、分支状血管及发现肾小球样血管。微血管管径最大区域的微血管数量不多，即MVD最大值区域与微血管管径最大值区域不一致。随着胶质瘤级别的增加，肿瘤微血管分布的异质性增大。 2. VSIMRI表现：弥漫性星形细胞瘤及II级少突胶质细胞瘤的VSI彩图中色彩较均一，呈VSI值较低的蓝绿色，同一患者不同兴趣区测得的数据相差不大，其中少突胶质瘤VSIMRI值小于弥漫性星形细胞瘤(VSIMRI平均值分别为20.80±3.83μm及56.70±15.69μm)。高级别胶质瘤的VSI彩图色彩不均一，可见到呈红色(VSIMRI值较大，III级胶质瘤118.03±22.53μm，胶质母细胞瘤144.05±23.09μm)的区域，特别是胶质母细胞瘤，病灶中央有明显坏死区域，VSIMRI值为0，VSIMRI最大的区域多位于坏死病灶的边缘，这与病理所见相符合。 3. VSIMRI值和组织学切片微血管管径的相关性：通过皮尔逊相关分析得出，病理切片中测量的微血管长径、短径与MRI测量的VSIMRI最大值、平均值两两之间都有良好的相关性(p值均小于0.01，r值均大于0.7)，在MRI与病理测值两种方法之间，相关系数最高的为微血管短径与VSIMRI平均值(r=0.8190)。 4.不同级别胶质瘤VSIMRI值的比较及病理切片结果的比较：高级别胶质瘤较低级别胶质瘤管径大，经统计学两两比较发现，WHO Ⅱ级～WHO Ⅲ级，WHO Ⅱ级～WHO Ⅳ级，WHO Ⅲ级～WHO Ⅳ级胶质瘤间的VSIMRI平均值差异均有统计学意义(p＜0.05)。低级别胶质瘤(WHO II级)与高级别胶质瘤(WHO Ⅲ级和WHO Ⅳ级)的VSIMRI最大值也有显著差异(p＜0.05)，而WHO Ⅲ级和WHO Ⅳ级胶质瘤间VSIMRI最大值差异无统计学意义。低级别胶质瘤(WHO II级)与高级别胶质瘤(WHO III级和WHO IV级)的管径有显著差异，WHO Ⅲ级～WHO Ⅳ级胶质瘤的短径差异有统计学意义，长径差异无统计学意义。WHO Ⅱ级～WHO Ⅳ级胶质瘤的MVD差异有统计学意义。 5. ROC分析：ROC分析结果显示VSIMRI平均值和最大值在鉴别Ⅱ级和Ⅲ、Ⅳ级胶质瘤的曲线下面积为1，VSIMRI平均值取临界值81.02μm，VSIMRI最大值取临界值109μm，其鉴别Ⅱ级和Ⅲ、Ⅳ级胶质瘤的敏感度及特异度均可达100％(P0.01)。VSIMRI平均值取142μm时，鉴别WHO Ⅲ级胶质瘤与WHO Ⅳ级胶质瘤的敏感性为68.75%，特异性为88.89%。 6. VSI与CBV比较：在同时行两种灌注扫描的13例患者中，rCBV值与VSIMRI值呈正相关，VSIMRI平均值与病理学测量微血管管径呈正相关，rCBV值与微血管管径相关性无明显统计学意义。根据ROC分析结果得出的cutoff值(VSIMRI平均值取81.12μm，VSIMRI最大值取109μm，rCBV值取3.4)对此组病例进行高、低级别判定。VSIMRI平均值及VSIMRI最大值能准确分辨出13例，准确率为100%；rCBV值能准确分辨出12例，准确率为92%。 结论： 1. VSIMRI值与病理切片测量微血管管径有良好相关性，这与动物实验研究结果吻合，其中，短径均值(200倍镜下测得的所有微血管径线均值)与VSIMRI平均值相关性最好，说明MRI-VSI技术能很好反应测定区域所有微血管直径的平均值，从而反应肿瘤微血管结构变化。 2.不同级别胶质瘤的VSIMRI值差异有统计学意义，VSIMRI值鉴别高、低级别胶质瘤的敏感度及特异度可高达100%，更值得一提的是VSIMRI平均值可鉴别WHO Ⅲ级及WHO Ⅳ级胶质瘤。因此，利用VSIMRI值能很好鉴别良、恶性胶质瘤，有助于胶质瘤术前分级。 3.微血管结构(管径)比微血管密度更能反应胶质瘤的恶性程度，因此我们推测MRI-VSI技术是优于血容量等常规指标描述微血管生成的方法，其对胶质瘤分级诊断提供更多信息。
[Abstract]:Glioma is the most common primary brain tumor. According to the classification of WHO central nervous system in 2007, it is divided into low grade glioma (WHO I, grade II) and high grade glioma (WHO III, grade IV). Different grades and types of gliomas have different biological behavior and prognosis, which are the key points in the diagnosis and treatment of clinics. The PerfusionWeighted magnetic resonance imaging (PWI) is a functional imaging method, which can evaluate the microvascular distribution and blood flow of the living tissue before operation. The microvasculature of the glioma can not only be used to detect the microvasculature in the glioma. Preoperative classification, guiding clinical treatment plans, is more conducive to monitoring the efficacy of antitumor drugs. Cerebral blood volume (CBV) has been widely used in clinical, and shows a certain advantage for preoperative grading of glioma. In recent years, animal experimental results have shown that the vascular diameter index (vessel size index) based on MR perfusion imaging (vessel size index, VSI) can describe the vascular remodeling of the lesion noninvasively, and VSI also known as vesselarchitectural imaging (VAI), which is superior to blood volume (CBV) and blood flow (CBF) to provide indirect information. However, this technique is still in the stage of clinical exploration and lacks pathological examination. This study uses SE-GE-EPI perfusion imaging technology. The microvascular diameter of human glioma was measured, compared with the results of CBV, and compared with the results of pathological immunohistochemical staining, the characteristics of microvascular formation in different levels of glioma and the accuracy of PWI-VSI measurement were analyzed, and the correlation between the microvascular diameter and the grade of glioma was determined.
Objective:
To investigate the effectiveness and accuracy of magnetic resonance perfusion imaging (MR-PWI) to measure the microvascular diameter of glioma, the histological features of the microvessels of different levels of glioma were analyzed, and the correlation between the blood volume index (VSI) of glioma, the blood volume (CBV) and the histological grade of glioma was discussed in order to improve the accuracy of the evaluation of the preoperative grading of glioma by MRI. And provide more information for clinical monitoring of drug treatment effect of glioma.
Materials and methods:
1. research objects
Glioma group: 35 cases of glioma confirmed by pathology, 15 male and 20 female, age 18~69 years, average 47.26 + 13.71 years old. All cases were first diagnosed cases, without any invasive or noninvasive treatment. According to WHO2007 classification and classification of central nervous system tumor, 10 cases of low grade tumor, including 7 cases of astrocytoma. There were 6 cases of diffuse astrocytoma and 1 cases of obesity type astrocytoma, 3 cases of oligodendroglioma, 25 cases of high grade tumor, 5 cases of oligodendroglioma with intermediate degeneration, 1 cases of anaplastic oligodendroid astrocytoma, 3 cases of anaplastic astrocytoma, and 16 cases of glioblastoma.
Normal control group: 10 cases were taken from normal brain tissue around excised glioma specimens.
2. scanning equipment and data measurement
MR imaging instrument: SIEMENS lagnetom Verio3.0T, complete the routine magnetic resonance scan and SE perfusion scan, instruct the patient to perform SE-GE-EPI perfusion 24 hours after the next day, use the GE SIGNA HDX1.5T superconducting MR scanner and the eight channel coil. VSIMRI value and CBV value in white matter lesion area.
3. pathological section staining and measurement index
All paraffin specimens (2-8 blocks per patient) of 35 patients with glioma were prepared by HE staining and CD34 immunohistochemical staining respectively. The microvascular density, morphology (including microvascular length and short diameter) was quantitatively analyzed by computer image analysis software (Image-ProPlus5.0).
4. data analysis
Pearson correlation analysis was used to evaluate the correlation between VSIMRI, CBV value and microvascular diameter and short diameter 22. The maximum value of VSIMRI, the mean value of VSIMRI, the rCBVmax value of the glioma at different levels, Mann-Whitney U test, P0.05 were statistically significant. The maximum value of VSIMRI, VSIMRI average, R was analyzed by the working characteristics of the subjects (ROC). The value of CBVmax in grading diagnosis of glioma and determine its best diagnostic value and sensitivity and specificity.
Result锛

本文编号：1874265