摘要
目的研究氢质子磁共振波谱(protonmagnetic resonance spectroscopy,lH-MRS)和弥散张量成像(diffusion tensor imaging,DTI)对颅内良、恶性肿瘤鉴别诊断的价值。方法经病理证实的颅内肿瘤33例,其中包括脑膜瘤10例,听神经瘤4例,血管母细胞瘤1例,混合性胶质神经元肿瘤1例,胶质瘤11例,转移瘤5例,原始神经外胚层肿瘤1例。根据病理结果,将颅内肿瘤分为良性组(19例)和恶性组(14例)进行观察分析。所有患者均进行常规MRI和DTI扫描,22例行lH-MRS分析。lH-MRS感兴趣区包括肿瘤组织和瘤周组织,分别观察各个区域的胆碱(choline,Cho)、N-乙酰天冬氨酸(N-acetyla spartate,NAA)、肌酸(creatine,Cr)、脂质(lipid,Lip)和乳酸(lactate,Lac)峰,并比较Cho/NAA、Cho/Cr、NAA/Cr值和Lip、Lac峰值。测量肿瘤实质区、坏死囊变区、周边水肿区、水肿周边正常白质区和对侧白质区感兴趣区的平均弥散系数(meandiffusivity,MD)、各向异性(fractionalanisotropy,FA)值。结果良、恶性肿瘤在1H-MRS代谢图上都表现为Cho峰升高、NAA峰降低,良性肿瘤实质区Cho/NAA和Cho/Cr值均显著低于恶性肿瘤,差异具有统计学意义(P<0.01)。虽然良性肿瘤NAA/Cr值高于恶性肿瘤组,但差异无统计学意义(P>0.05)。13例良性肿瘤中8例检测到Lac峰,2例检测到低Lip峰。9例恶性肿瘤中7例检测到Lac峰,5例检测到Lip峰;良性肿瘤实质部分和瘤周正常白质区FA值明显高于恶性肿瘤,且差异具有统计学意义(P<0.05),而2组肿瘤间坏死囊变部分和瘤周水肿区FA值比较无显著差异(P>0.05)。良性肿瘤水肿周边正常白质区FA值明显高于对侧相应部位脑白质,且有统计学差异(P<0.05);恶性肿瘤水肿周边白质FA值虽低于对侧脑白质,但差异不显著(P>0.05);MD值在良性肿瘤实质部分显著高于恶性肿瘤(P<0.05),但在坏死囊变区、水肿区和周边正常白质区差异均无统计学意义(P>0.05);肿瘤实质部分MD值和FA值均与Cho/Cr呈线性负相关。结论1H-MRS能无创地分析颅内肿瘤代谢信息,Cho/NAA和Cho/Cr有助于肿瘤良、恶性的鉴别;Cho/Cr与MD、FA值呈负性相关,Cho/NAA与Cho/Cr值和FA、MD值可以有效的鉴别颅内良、恶性肿瘤。
Objective To investigate the value of proton MR spectroscopy ( ^1H-MRS) and diffusion tensor imaging in differentiating benign from malignant intracranial tumors. Methods Thirty-three patients with pathological confirmed brain tumor,including 10 meningiomas,4 acoustic neuromas,1 hemangioblastoma,1 mixed glioneuronal tumor,11 gliomas,5 metastasis and 1 primitive neuroectodermal tumor underwent conventional MRI and diffusion tensor imaging,and 22 of them performed proton magnetic resonance spectroscopy in our hospital from September to December 2008. Choine ( Cho) ,Nacetylaspartate ( NAA) ,creatine ( Cr) ,lipid ( Lip) and lactate ( Lac) level were measured in the tumor masses and peritumor tissues. Mean diffusivity ( MD) and fractional anisotropy ( FA) was obtained in solid area,necrosis area,peritumor edema area,white matter surrounding the edema and contralateral white matter. Results All the patients were divided into benign ( n = 19) and malignant ( n = 14) groups according to the pathological results. Both benign and malignant tumors showed low NAA and high Cho. The Cho/NAA and Cho/Cr value in solid area of the benign tumor was markedly lower than that of the malignant tumor,with a significant difference ( P〈0. 01) ,while the NAA/Cr ratio in the former group was higher than that of the later,though having no significant difference between them ( P〉0. 05) . Eight of 13 benign tumors had Lac peak,and 2 had Lip peak. Seven of 9 malignant tumors had Lac peak,and 5 had Lip peak. The FA value in solid area and white matter surrounding the edema of benign tumors was significant higher than that of the malignant ( P〈0. 05) ,however the difference of FA in the necrosis and peritumor edema area between the 2 groups was not significant ( P〉0. 05) . FA value in the white matter surrounding edema of benign tumor was significant higher than contralateral white matter ( P〈0. 05) . Although the FA value in the white matter surrounding edema of malignant tumor was lower than the contralateral white matter,though the difference didn't reach statistical significance ( P〉0. 05) . The difference of MD in solid area of benign and malignant tumor was significant ( P〈0. 05) ,but difference of necrosis, peritumor edema and white matter surrounding edema between the 2 groups was not significant ( P〉0. 05) . For the FA and MD value in solid part of tumor there was linear negative correlation with Cho/Cr ratio. Conclusion^1 H-MRS,a noninvasive technique for investigating the metabolic changes in brain tumor,is helpful to differentiate benign from malignant intracranial tumors through Cho/NAA and Cho/Cr ratios. There is negative correlation between Cho/Cr and FA,MD value. Cho/NAA,Cho/Cr and FA,MD value may effectively differentiate benign from malignant intracranial tumor.
出处
《第三军医大学学报》
CAS
CSCD
北大核心
2009年第21期2061-2064,共4页
Journal of Third Military Medical University
基金
重庆市科技攻关项目(CSTC2007AC5014)~~
关键词
颅内肿瘤
磁共振波谱分析
弥散张量成像
代谢产物
鉴别诊断
intracranial tumor proton magnetic resonance spectroscopy diffusion tensor imaging metabolites differential diagnosis