A Serum Metabolomics Study Of Gastric Cancer Based On Pseudotargeted LC/MS Approach

Background:Gastric cancer is one of the most common malignancy and thesecond most frequent cause of cancer-related mortality in the world.Noninvasive screen methods including tumor markers are intensivelyneeded in the clinic. Therefore to adopt new technology to find newmarkers for gastric cancer diagnosis is of great significance to improvethe diagnostic rate of disease. Metabonomics is defined as “thequantitative measurement of the dynamic, multiparametric response ofliving systems to pathophysislogic stimuli or genntic modification” byJeremy Nicholson in1999. It is a hotspot and key area of currentscientific research, metabolic activity in biology and the pathogenesis anddiagnosis and treatment of disease has played an increasingly importantrole, especially in the tumor pathogenesis and tumor markers found inliving organisms is the research focus in recent years.Objective:Our study used a metabolomics research platform to analyze thedifferent metabolic substances in the sera of Gastric cancer and normalcontrols, aimed to find potential biomarkers for the diagnosis of Gastriccancer. Methods:To discover metabolic markers for gastric cancer, the sera metabolicprofiles of20gastric cancer (GC) patients and40healthy controls wereanalyzed using a liquid chromatography coupled to mass spectrometry(LC/MS) pseudotargeted approach. In this method, the metabolic featureswere firstly scanned and extracted on a ultra-performance LC coupled toquadrupole time of flight MS. Then, extracted ion-pairs were fetched to aultra-performance LC-triple quadrupole MS for multiple reactionmonitoring detection. Peak detection and alignment from the raw datawere performed using Analyst Quantitation (AB Sciex company, USA)software, and exported peak table. The exported data after internalstandard calibration, using SIMCA-p11.0(Umetrics Sweden ABcompany) software for multivariate data analysis, principle componentanalysis (PCA) and partial least squares-discriminant analysis (PLS-DA)model was constructed, nonparametric test was performed to find out thestatistical significance between the normal group and gastric cancer groupusing SPSS18.0software (IBM, United States). The metabolites withvariable importance factor value (VIP)>1and p <0.05were selected asdifferential ions.Results:Using this so-called pseudotargeted methods, there are totally373positive ions and247negative ions was measured stably. And thechemical structure of212metabolites were accordingly identified. Theresults of PCA and PLS-DA indicated that obvious classification could be observed between GC and the healthy controls. According to the resultsof non-parametric statistical test,57ions were identified among all thedifferential metabolic features. Of all the identified differentialmetabolites, dihydrocholesterol provide good diagnostic performance ofGC, which was also validated by another cohort including20gastriccancer patients and39healthy controls.Conclusion:Patients with gastric cancer have state of metabolic disorders inphospholipids, cholesterol, amino acids, free fatty acids, and carnitinemetabolites. Of all the identified differential metabolites,dihydrocholesterol provide good diagnostic performance of GC, may bethe potential biomarker for the diagnosis of Gastric cancer.