不同模式识别方法分析食管癌患者血浆代谢组的磁共振氢谱  

1 H.MRS metabonomic analysis of plasma samples of esophageal cancer patients based on different pattern recognition

作　　者：阿仙姑·哈斯木[1] 巴吐尔·买买提明[2] 伊力亚尔·夏合丁[1] 

机构地区：[1]新疆医科大学第一附属医院食管癌研究所,乌鲁木齐830054 [2]新疆医科大学药学院测试中心

出　　处：《中华肿瘤杂志》2010年第9期681-684,共4页Chinese Journal of Oncology

基　　金：新疆医科大学创新基金项目（YKD2007-10）,新疆高校科研计划项目（XJEDU2008S38）

摘　　要：目的 探讨不同模式识别方法应用于食管癌磁共振（MR）代谢组学研究的可行性.方法 应用MR技术对109例食管癌患者和50名健康人的血浆标本进行检测,记录血浆一维600MHz磁共振氢谱（1H-MRS）.分别采用主成分分析法（PCA）、偏最小二乘判别分析法（PLS-DA）和正交偏最小二乘判别分析法（OPLS-DA）对1H-MRS进行模式识别分析,确定食管癌患者和健康人血浆差异性代谢组分,并分析3种模式识别方法的判别能力.结果 OPLS-DA方法能够区分食管癌患者和健康人血浆代谢组的1H-MRS,其判别能力优于PCA法和PLS-DA法.OPLS-DA分析结果显示,与健康人比较,食管癌患者血浆中多种代谢物显著下降,包括极低密度脂蛋白、低密度脂蛋白、不饱和脂肪酸、异亮氨酸、亮氨酸、缬氨酸、乳酸、丙氨酸、乙酸、糖蛋白、丙酮、肌酸酐、胆碱、1-甲基组氨酸和异丁酸等,差异有统计学意义（均r＞0.27,P＜0.05）,而柠檬酸、β-葡萄糖、α-葡萄糖和二甲胺等明显增加,差异有统计学意义（均r＜-0.27,P＜0.05）.结论 1H-MRS的OPLS-DA模式识别方法可以剔除临床非实验因素的影响,为MR代谢组学技术应用于食管癌的临床诊断开辟了道路.Objective To analyze the metabonomic 1H-MRS of plasma samples from patients with esophageal cancer and healthy controls applying different pattern recognition methods, and to explore the potential of application of 1H-MR-based metabonomics in clinical research. Methods 1H-MR was performed on plasma samples from 109 EC patients and 50 health controls to analyze the metabonomic variation between EC patients and healthy subjects and the corresponding 1H-MRS were recorded on Varian Unity ANOVA 600 MHz to perform principal components analysis （PCA）, partial least squares discriminant analysis （PLS-DA）, orthogonal partial least squares discriminant analysis （OPLS-DA）, respectively.Results OPLS-DA analysis could correctly separate almost all the plasma samples from EC patients and health controls, better than both the PCA and PLS-DA. The plasma levels of leucine, alanine, isoleucine,valine, glycoprotein, lactate, acetone, acetate, choline, isobutyrate, unsaturated lipid, VLDL, LDL,1-methylhistidine were significantly decreased in EC patients （r total 〉 0.27, P 〈 0.05）, while that of dimethylamine, α-glucose, β-glucose, citric acid increased in the EC patients （r total 〈 -0.27, P 〈0.05）. Conclusions The analysis of metabonomic 1H-MRS of plasma samples by OPLS-DA method can eliminate the influence of non-experimental factors and decrease the heterogeneity of samples. It is useful and of great potential for application in clinical diagnosis and research of esophageal cancer.

关 键 词：食管肿瘤 磁共振氢谱 代谢组学 正交偏最小二乘判别分析法 

分 类 号：R686[医药卫生—骨科学]