Gas chromatography-mass spectrometry technology for finding serum markers of neonatal sepsis

Objective To find the potential biomarkers of neonatal sepsis through comparing neonates with sepsis and without sepsis in serum metabolin differences. Methods Thirty neonates, who were diagnosed as sepsis from the Neonatal Center of Bayi Children's Hospital Affiliated to Beijing Military Region General Hospital during Aug.2012 to Feb.2013, were enrolled as the experimental group; while another 30 non-sepsis neonates were selected as the control group, the data of them were not significantly different from those of the experimental group.1.5 mL supernatant of radial artery blood was saved in -80 ℃ refrigerator after centrifugation in 4 ℃.Gas chromatography-mass spectrometry(GC-MS) technology was used for data acquisition after the serum sample thawinged at the room temperature and pretreated.The software of the GC-MS instrument was available for denoising, peak detection and peak area integral.Principal components analysis(PCA) model and partial least squares-discriminant analysis(PLS-DA) model were established by Matlab2011 software after the data standardized.The potential biomarkers were screened by t-test. Results The differences of the serum metabolites between sepsis neonates and non-sepsis neonates in PCA and PLS-DA model were visible.Compared with the control group by t-test, the experimental group had increased in the contents of glucose, lactic acid, urea, pyruvic acid, phenylalanine, and decreased in the contents of leucine, isoleucine, alanine, glutamine, palmitic acid, palmitoleic acid, linoleic acid, stearic acid.The sensitivity, specificity, Youden's index, compliance rate of the discriminant composited by the above metabolites were 93.33%, 100.00%, 0.9333 and 96.66%. Conclusions GC-MS technology can be used to detect the neonatal sepsis.The discriminant formed by the above metabolites can be used for the diagnosis of neonatal sepsis. Key words: Neonatal sepsis; Gas chromatography-mass spectrometry technology; Biomarker