Perioperative Serum Metabolomic Changes And Prognosis Of Oral Squamous Cell Carcinoma

ObjectiveThe purpose of this study was to analyze the perioperative serum metabolomic changes of oral squamous cell carcinoma(OSCC)and the relationship between the changed metabolites and the prognosis of OSCC patients.MethodsIn this study,ultra high performance liquid chromatography quadrupole/electrostatic field orbital trap tandem high resolution mass spectrometer(UHPLC-Q-Orbitrap HRMS)was used to obtain the preoperative and postoperative serum fingerprints of 103 patients with OSCC collected in the First Affiliated Hospital of Zhengzhou University from December 2017 to March 2018.73 cases were the training set and the other 30 cases were the verification set.① After the serum fingerprint was obtained in the training set,the changes of perioperative metabolites of OSCC were further screened and identified by multivariate statistical analysis.②Import the resulting differential metabolites into metaboanalyst Metabolic pathway analysis and heat map construction showed the differences between the two groups.③ In the human metabolomics database,the genes corresponding to different metabolites were obtained,and the above genes were introduced into KEGG database for enrichment analysis.The visual path P<0.05 and the false discovery rate(FDR)<0.05 were used as alternative pathways to verify the results in metabolomics.④SPSS(IBM version 26.0)was used to draw receiver operating characteristic curves(ROC)of different metabolites and calculate the area under the receiver operating characteristic curves(AUC).The metabolites with AUC>0.8 were screened out.⑤ Then,divide the peak area of each patient’s corresponding preoperative group by the peak area of each patient’s corresponding postoperative group,and calculate the fold change(FC)value of each patient’s metabolites.Using X-tile software Calculate the best cut-off value,replace the average FC value,and divide the samples into high metabolite change group and low metabolite change group.⑥ Kaplan Meier method and log rank test were used to evaluate the univariate prognostic value of clinical parameters and metabolite changes.(clinical parameters include gender,age,smoking,drinking,primary site,tumor differentiation,tumor stage,clinical stage and lymph node invasion).In the validation set,the independent prognostic factors in the training set were verified by Kaplan Meier method and log rank method.The metabolic factors also were verified by Kaplan-Meier and log-rank method.⑦ The variables with statistically significant correlation with the 3-year RFS rate in univariate analysis were included in the multivariate Cox proportional hazards model to evaluate the independent prognostic factors of 3-year RFS.Results① In this study,serum samples were analyzed based on uhplc-q-orbitrap HRMS to determine the differential metabolites between the preoperative and postoperative groups of OSCC.A total of 14 differential metabolites were identified,including succinic acid,arginine,9-decanoylcarnitine,asparaginyl-Valine,glutamine,hypoxanthine,sphingosine,palmitoyl ethanolamide,hexanoylcarnitine,orotic acid,uric acid,vanillyl mandelic acid,ethyl acetate,thromboxane B2.② The metabolic pathways of 14 differential metabolites screened in the early stage were analyzed.It was found that the metabolic pathways such as tricarboxylic acid cycle metabolic pathway,purine metabolic pathway,alanine,aspartic acid and glutamate metabolic pathway,pyrimidine metabolic pathway and sphingolipid metabolic pathway changed before and after OSCC.③ KEGG enrichment analysis showed that compared with the preoperative group,13 pathways(P<0.05,FDR<0.05)were significantly disturbed in the postoperative group.The most important genes were involved in amino acid metabolism pathway and purine metabolism pathway.④ Combined with the area under the ROC curve,eight foreign bodies with poor metabolism with AUC>0.8 were screened:succinic acid(AUC=0.926,95%CI,0.875,0.965,P<0.05),hypoxanthine(AUC=0.905,95%CI,0.855,0.947,P<0.05),thromboxane B2(AUC=0.944,95%CI 0.887,0.98,P<0.05),glutamine(AUC=0.922,95%CI,0.870,0.922,P<0.05)Arginine(AUC=0.836,95%CI,0.836,0.946,P<0.05),9-decanoylcarnitine(AUC=0.873,95%CI,0.813,0.923,P<0.05),orotic acid(AUC=0.813,95%CI 0.735,0.889,P<0.05),asparaginyl valine(AUC=0.809,95%CI 0.737,0.882,P<0.05).⑤According to the maximum cut-off value of metabolite FC obtained by X-tile,the changes of patient metabolites were divided into two groups.Kaplan Meier method and log rank test were used to evaluate the univariate prognostic value of clinical parameters(gender,age,smoking,drinking,primary location,tumor differentiation,tumor stage,clinical stage and lymph node invasion)and the changes of the above eight metabolites.Univariate log rank test showed that tumor differentiation(grade I),tumor T stage(T1/T2),low change level of succinic acid and high change level of hypoxanthine were associated with better 3-year RFS(P=0.025,P=0.003,P=0.015,P<0.001).In the validation set of 30 OSCC patients,low changes in succinic acid and high changes in hypoxanthine were also associated with better 3-year RFS.⑥Tumor differentiation(grade I),tumor T stage(T1/T2),low change level of succinic acid and high change level of hypoxanthine were included in the multivariate Cox proportional hazards model,The results showed that low succinic acid change and high hypoxanthine change were independent prognostic factors for the 3-year RFS rate(hazard ratio[HR]=5.730,95%CI,1.667-9.694;[HR]=3.221,95%CI,1.233-8.414).⑦Single variable ROC curve analysis showed that T stage,tumor differentiation,low change level of succinic acid and high change level of hypoxanthine had similar prediction accuracy.ROC curve analysis of different factor combinations showed that the prognosis evaluation group with succinic acid change low,hypoxanthine change high and tumor grade(degree of differentiation)had the highest prediction accuracy(AUC=0.90;95%confidence interval,0.822-0.967)ConclusionsThe changes of perioperative metabolites may be related to the prognosis of OSCC patients.Serum metabolomic analysis based on uhplc-q-orbitrap HRMS can further stratify the prognosis of OSCC patients.These results can better understand the relevant mechanisms of early recurrence and help to develop more effective therapeutic targets.