液相色谱质谱法对66例胃癌患者血清代谢组学特征的分析

Serum metabonomics study in 66 patients with gastric cancer by liquid chromatography-mass spectrometry

目的采用液相色谱质谱(LC-MS)的代谢组学分析方法,研究胃癌患者是否有独特的血清代谢组学特征,寻找与胃癌早期相关的潜在标志物,初步探讨其涉及的相关代谢通路。方法收集2017年7月至2018年1月苏州大学附属第一医院66例胃癌患者和44例良性胃病患者,以50名健康人群作为健康对照组。另从66例胃癌患者中选取25例胃癌Ⅰ、Ⅱ期患者和25例胃癌Ⅲ、Ⅳ期患者,以及从50名健康对照组中选取25名对照者,采用LC-MS法检测胃癌、良性胃病和健康对照者血清中小分子代谢物。采用多因素logistic回归分析,构建主成分分析和偏最小二乘法-判别分析(PLS-DA)模型并验证筛选差异代谢物,通过ROC曲线评价差异代谢物的临床效能。结果主成分分析和PLS-DA模型均显示,胃癌具有明显独特的代谢谱,良性胃病的代谢谱与胃癌和健康对照组有部分交叉。多因素logistic回归分析证实,异亮氨酸、苯甲酮、1-磷酸鞘氨醇和吡喃型半乳糖4种物质建立最优化诊断模型,AUC值(95%CI)为0.963(0.930~0.997),最佳界值为0.871,灵敏度为93.1%,特异度为94.0%。PLS-DA模型显示胃癌Ⅰ+Ⅱ期、Ⅲ+Ⅳ期患者与健康对照者相比具有明显区分聚类趋势,在胃癌Ⅰ+Ⅱ期、Ⅲ+Ⅳ期患者血清中共鉴定出24种差异代谢物,其中赖氨酸、苯磺酰胺、肉碱、精氨酸和二十二碳六烯酸乙酯5种代谢物伴随胃癌的进展浓度逐渐升高,吡哌酸和犬尿氨酸可能作为早中期(Ⅰ+Ⅱ期)胃癌筛查的标志物。结论 LC-MS代谢组学可以有效证实胃癌患者具有独特的血清代谢物变化,筛选出的差异代谢物对预测胃癌发生风险具有潜在的临床应用价值。

Objective To detect whether patients with gastric cancer had unique serum metabolomic characteristics by liquid chromatography-mass spectrometry (LC-MS) metabolomic analysis, and to screen potential markers for early gastric cancer and to preliminarily explore the related metabolic pathways. Methods At the First Affiliated Hospital of Soochow University, 66 patients with gastric cancer and 44 patients with benign gastric disease from July, 2017 to January, 2018 were enrolled, and 50 healthy subjects served as controls. Twenty-five patients with gastric cancer at stage Ⅰ and Ⅱ and 25 patients with gastric cancer at stage Ⅲ and Ⅳ were selected from the 66 patients with gastric cancer, and 25 subjects were also selected from 50 healthy controls. The plasma small molecule metabolites of patients with gastric cancer and benign gastric disease and healthy controls were detected by LC-MS method. Multivariate logistic regression analysis was used to establish and validate the principal component analysis (PCA) model and partial least squares-discriminant analysis (PLS-DA) model and screen the differential metabolites. The receiver operating characteristic curve analysis was used to evaluated the clinical efficacy of differential metabolites. Results PCA and PLS-DA models showed that gastric cancer had a obviously specific metabolites profile, the profile of benign gastric disease overlapped with that of gastric cancer and healthy controls. The results of multivariate logistic regression analysis confirmed that four metabolites including isoleucine, benzophenone, sphingosine-1-phosphate and galactopyranose set could be used to establish an optimal diagnostic model. The area under the curve (AUC)(95% confidence interval (CI)) was 0.963 (0.930 to 0.997), and the best cut off value, sensitivity and specificity were 0.871, 93.1% and 94.0%, respectively. Meanwhile, patients with gastric cancer at stage Ⅰ+Ⅱ and stage Ⅲ+Ⅳ had a distinct clustering trend compared with the control group. In the serum of patients with gastric cancer at stage Ⅰ+Ⅱ and stage Ⅲ+Ⅳ, a total of 24 differential metabolites were identified, the concentration of five of which including lysine, carnitine, benzenesulfonamide, arginine and docosahexaenoic acid ethyl ester, increased along with the progression of gastric cancer. Pipecolic acid and kynurenine might served as biomarkers for early and mid gastric cancer (stage Ⅰ+Ⅱ) screening. Conclusions LC-MS metabolomic effectively confirm the unique changes of serum metabolites in patients with gastric cancer. The screened differential metabolites have potential clinical application value for predicting the risk of gastric cancer.