Objective: To study the changes of urinary metabolic profile, screen metabolic ions characterization with clinical diagnostic value, and a disease differentiation model establishment, in an attempt to help the clinical diagnosis of hepatocellular carcinoma patients. Methods: A case-control study was conducted. Ultra-performance liquid chromatography/mass spectrometry (UPLC-MS) were used to analyze urine samples of 32 patients with hepatocellular carcinoma, 28 patients with liver cirrhosis and 28 healthy persons, respectively. The orthogonal partial least squares discriminant analysis (OPLS-DA) and the principal component analysis (PCA) model were constructed using MZmine2.0 and SIMCA-P + 12.0.1.0 software for preliminary screening of metabolites. The metabolic ions selected in the final test were analyzed by SPSS, and the markers were analyzed and screened by one-way analysis of variance. Finally, the sensitivity and specificity of the selected markers were analyzed by calculating the area under the receiver operating characteristic (ROC) curve. One-way analysis of variance was used to compare quantitative indicators between groups. Results: OPLS-DA model parameters were R2X = 35.3%, R2Y = 86.9%, and Q2 = 72.2%, which had a good identification value. A total of 26 characteristic ions were screened, of which 17 were identified. 14, 19-Dihydroaspidospermatine had a high value in distinguishing healthy person with hepatocellular carcinoma patients, and the area under the receiver operating characteristic curve was higher than 0.9. The area under the ROC curve for distinguishing liver cancer with liver cirrhosis patients was 0.88, which was higher than the ROC curve of alpha-fetoprotein (0.75). Conclusion: Based on the UPLC-MS platform, the PCA and OPLS-DA models were successfully constructed, and the characteristic metabolic ions in the urine were extracted and identified, which has a certain value in assisting clinical screening of hepatocellular carcinoma patients.
目的： 研究肝细胞癌患者尿液代谢轮廓的变化，筛选具有临床诊断价值的特征代谢离子，建立疾病区分模型，从而帮助临床诊断。 方法： 采用病例对照研究，通过超高效液相色谱与质谱联用（UPLC-MS）法分析肝细胞癌、肝硬化患者和健康人尿液标本，分别有32、28、28份。用MZmine2.0和SIMCA-P+12.0.1.0软件构建正交偏最小二乘判别分析（OPLS-DA）模型和主成成分分析（PCA）模型进行代谢物初筛，对最终检验所选择的代谢离子通过SPSS进行数据分析，用单因素方差分析对标志物进行分析筛选。最后通过计算受试者操作特征（ROC）曲线下面积对所筛选出的标志物进行敏感性和特异性分析。组间定量指标的比较采用单因素方差分析。 结果： OPLS-DA的模型参数R2X = 35.3%、R2Y = 86.9%、Q2 = 72.2%，具有较好的鉴别价值。检测共筛选出26个特征离子并鉴定了其中的17个。14，19-Dihydroaspidospermatine对健康人和肝细胞癌患者具有较高的鉴别价值，曲线下面积> 0.9，对肝细胞癌患者和肝硬化患者鉴别的ROC曲线下面积为0.88，高于甲胎蛋白的ROC曲线下面积(0.75)。 结论： 基于UPLC-MS平台，成功构建PCA和OPLS-DA模型，提取并鉴定出肝细胞癌患者尿液的特征代谢离子，对辅助临床筛选出肝细胞癌患者具有一定价值。.
Keywords: Hepatocellular carcinoma; Metabolism profile; Urine.