Determination Of Metabolites In Pleural Effusion By High Performance Liquid Chromatography

In the past half a century,lung cancer has been one of the most common malignant tumors with the highest rate of morbidity and mortality in many countries.Because of lacking sensitive and precise methods for early detection,the five-year survival rate of lung cancer is only arround 14%.Pleural effusion is the most specific human body fluid of lung disease and usually induced directly by lung disease.In this research,we have established the HPLC method for pleural effusion and combined with pattern recognition,and applied to the lung cancer metabonomics study.We have conducted the exploration research through the contrast between the malignant(caused by lung cancer)and non-malignant pleural effusion,and we hope to discover more sensitive and more specific tumor biomarkers.The main work and achievement of this thesis are as follows:1.A method by HPLC-DAD and HPLC-ELSD was developed for the simultaneous determination of tyrosine,2'-deoxyguanosine,inosine,guanosine,phenylanine and tryptophan in pleural effusion.A mixed solution with these six standard substances and Waters sunfire C18 was employed to optimize the mobile phase,detection wavelength,drift tube temperature and nitrogen flow.The result showed that the linear correlation coefficient of these six substances was above 0.9996.The relative standard deviation of seven parallel injection was less than 2.73%.The limit of detections(LOD)of these six substance were between 0.008-0.50?mol/L.The recoveries were ranged from 81.5%to 110.5%.The method was applied to the analysis of pleural effusion samples,and the results indicated that the method was simple,rapid,accurate and suitable for pleural effusion analysis.2.A method by HPLC was developed for the simultaneous determination of tyrosine,phenylanine and tryptophan in pleural effusion and applied to the differential diagnosis of pleural effusion causes.Pleural effusion and VertexTM AQ C18 was employed for the experimental conditions optimizing.The result showed that good linear relationship was achieved between the chromatographic peak area and the concentration of each component.The linear correlation coefficient of these three substance was above 0.999.The relative standard deviation of seven parallel injection was less than 3.2%.The limit of detection(LOD)of these three substance was 0.5,0.17 and 0.20 pmol/L,respectively.Pleural effusion samples were precipitated with methanol,then centrifuged to remove protein residue and finally analyzed by HPLC.The recoveries were ranged from 80.7%to 101.1%.The method was applied to the analysis of 72 patients with malignant(20 patients)and non-malignant(52 patients)and the results indicated that the difference of these three amino acids was statistically significant and had certain clinical value in differential diagnosis of pleural effusion cause.3.Using the HPLC-UV method optimized by VertexTM AQ C18 column,we analyzed 20 malignant and 50 non-malignant pleural effusion samples.The obtained data with 28 peaks for 70 patients was pretreated and analyzed by pattern recognition including PCA,PLS-DA and OPLS-DA in SIMCA.Meanwhile we used non-parametric t-test and the area under the ROC curve to evaluate the effect to distinguish the malignant and non-malignant groups of the 28 kind of substances.The study found that PLS-DA can distinguish the comparison group of malignant and non-malignant groups,malignant and tuberculous pleurisy groups.OPLS-DA can differentiate the comparison group of malignant and non-malignant groups,malignant and tuberculous pleurisy groups,malignant and pneumonia groups.The results of non-parametric t-test showed that eight substances were statistically significant between malignant and non-malignant groups,seven substances were statistically significant between malignant and tuberculous pleurisy groups,two substances were statistically significant between malignant and tuberculous pleurisy groups.The content of these substances in the lung cancer group was significantly higher than that of the control group,and found that the retention time of 3.54 min substances was all significant differences in the three comparison groups.According to the area under the ROC curve of 28 substances detected,two substances(RT=25.16 min,31.11 min)of which the area under the curve is greater than 0.6 and p<0.05 as lung cancer biomarkers.The initial establishment of the judgment of non-malignant and malignant models provide the basis for secondary clinical diagnosis of lung cancer.