多模态分子磁共振成像在脑胶质瘤分级中的基础研究

Glioma is the most common primary brain tumor in adults. Accurate grading of gliomas before surgery is fundamental in determining clinical treatment strategies. However, routine magnetic resonance imaging (MRI) techniques sometimes cannot accurately grade gliomas, because overlaps exist in the imaging manifestations between different grade gliomas. Currently, new noninvasive MRI techniques for improving the grading of gliomas have been considered as emphases and hot points in the native and foreign studies. In recent years, the developments of noninvasive molecular MRI techniques have provided probabilities for improving the grading of gliomas. This study is planned to use diffusion-weighted imaging (DWI) with multi-models including monoexponential, biexponential and ultra-high b-values models, and chemical exchange saturation transfer (CEST) techniques, which are combined with pathological indices including aquaporin (AQP) 1, AQP4 and AQP9, Ki-67, VEGFR2 and MVD, to explore the capabilities of parameters obtained from molecular MRI with multi-models in representing the biological characteristics of gliomas and provide new noninvasive imaging biomarkers for improving the grading of gliomas and the assessment of prognoses.

脑胶质瘤是成人最常见的脑内原发肿瘤，术前准确分级对其临床治疗方案的选择至关重要。然而，由于不同级别的脑胶质瘤影像学表现有重叠，常规磁共振成像技术时常不能对脑胶质瘤进行准确分级。如何利用无创性磁共振新技术提高脑胶质瘤分级的准确性是目前国内外研究的重点和热点。近年来，无创性分子磁共振技术的发展为此提供了可能。本研究拟采用含有单指数、双指数和超高b值模型在内的多模态磁共振扩散加权成像(DWI)及化学交换饱和转移（CEST）相关成像技术，与包括水通道蛋白（AQP）1、AQP 4和AQP 9、Ki-67、VEGFR2和MVD在内的多项病理学检测指标相结合，探讨多模态分子磁共振成像参数反映脑胶质瘤组织不同方面生物学特征的能力，以期利用多模态分子磁共振成像技术提高脑胶质瘤分级的准确性，从而为脑胶质瘤的准确分级和改善预后提供新的无创性影像学生物标记物。

脑胶质瘤是成人最常见的脑内原发性肿瘤。脑胶质瘤的分级和病理指标与治疗方案的制定及预后密切相关。常规磁共振成像(MRI)技术在脑胶质瘤分级方面价值有限，而MRI技术的快速发展，使得无创性MRI新技术在准确分级和反应病理指标方面具有很大潜力。本项目采用无创性多模态MRI技术，包括单指数、双指数、超高b值模型的扩散加权成像(DWI)和基于化学交换饱和转移(CEST) 原理的酰胺质子转移(APT)等，对脑胶质瘤的术前分级开展研究，并进一步将多模态MRI参数与水通道蛋白(AQP)1、AQP4、Ki-67等病理检测指标相结合，无创地反映脑胶质瘤组织的生物学特征，从而为脑胶质瘤的准确分级提供影像学生物标记物。本项目的研究结果显示，与低级别脑胶质瘤相比，高级别脑胶质瘤的标准表观扩散系数(ADCstandard, p<0.001)和真扩散系数(ADCslow, p=0.005)显著减低，而伪扩散系数(ADCfast, p=0.003)和ADCultra-high(p=0.006)显著增高。脑胶质瘤组织的AQP1 (r=0.515, p<0.01)、AQP4 (r=0.532, p<0.01)与ADCultra-high呈显著正相关。脑胶质瘤组织的Ki-67与ADCstandard (r=-0.427, p<0.01)和ADCslow (r= 0.583, p<0.01)呈显著负相关。与ADCstandard相比，ADCslow与Ki-67的相关性显著增高(p<0.05)。APT加权(APT-weighted, APTw)信号强度与脑胶质瘤的Ki-67指数呈显著正相关(r=0.624, P<0.001), ADCstandard与脑胶质瘤的Ki-67指数呈显著负相关(r=-0.427, P<0.01)。与ADCstandard相比，APTw与脑胶质瘤的Ki-67指数具有更强的相关性(p<0.05)。无创性多模态MRI技术包括双指数、超高b值模型的DWI和APT等，不仅能够对脑胶质瘤进行术前准确分级，而且能够有效反映AQP1、AQP4和Ki-67等病理指标的表达，从而提供更加丰富的脑胶质瘤组织信息，对于指导治疗和预后预测具有重要价值，在临床诊疗中具有广阔的应用前景。