| Objective: To investigate the predictive role of therapeutic drug monitoring(TDM)and pharmacometabolics in capecitabine adverse reactions in patients with colorectal cancer(CRC).Methods:Sample collection: Blood samples of patients after radical resection of colorectal cancer before and 1h,2.5h and 4h after taking Cap.At the end of each chemotherapy cycle,patients were followed up to record the type and grade of adverse reactions that occurred after Cap2.TDM: using the agilent 1290 series ultra high performance liquid chromatograph and agilent 6460 a type triple level 4 pole mass spectrometer(UHPLC-MS/MS)detection taking various time points after Cap Cap in the blood,cytidine 5 ’-DNA-5-fluorine(5’-DFCR),5 ’-DNA-5-floxuridine(5’-DFUR),2 ’-deoxy-5-floxuridine(2’-DFUR),5-fluorouracil(5 FU),concentration of fluorouracil(FUH2).The area under the drug time curve(AUC)of CAP and its metabolites was calculated by using Graph Pad software.According to the adverse reactions(nausea,vomiting,diarrhea,hand-foot syndrome(HFS),leucopenia,anemia,thrombocytopenia),the groups were grouped,and the independent sample T test was performed by SPSS to compare whether there were significant differences in CAP and its metabolite AUC among different groups.AUC with p value less than 0.05 was selected.Finally,binary logistic regression analysis and ROC(receiver operation characteristic curve)analysis were performed on the selected AUC to establish the model.3.Pharmacometabolics detection: UHPLC/Q-TOF-MS were used to detect characteristic ions at each time point before and after medication,and software such as MSDial and MSFinder were used for compound identification.Compounds with significant differences between groups were screened based on whether each adverse reaction occurred.The endogenous compounds were further screened by searching HMDB,Pub Med and other online databases.SPSS was used to conduct univariate binary logistic regression analysis,multivariate binary logistic regression analysis and ROC analysis for these endogenous compounds,and finally the model was established.4.Analyze and discuss the correlation and advantages and disadvantages of the two forecasting methods.We performed Pearson correlation analysis between these endogenous metabolites and the AUC and plasma concentrations of CAP and its metabolites.Results:1.The sample collection: 25 patients with conform to the requirements of the samples were collected,through the telephone,query clinical medical follow-up after chemotherapy adverse reactions,including 9 cases of patients occurred nausea,vomiting,diarrhea in 4cases,HFS in 8 cases,white blood cells reduce disease in 6 cases,anemia in 14 cases,12 cases of thrombocytopenia.2.The predictive role of TDM in adverse reactions of CAP chemotherapy:Based on UHPLC-MS/MS to explore the AUC of Cap and its metabolites and the relationship between the adverse reaction after chemotherapy,found through the analysis of the AUC of Cap and its metabolites and the relationship between the adverse reaction after chemotherapy,5-FU AUC and Cap after chemotherapy diarrhea statistically significant(p= 0.009,the area under the ROC curve is 0.724),5 ’-DFCR and FUH2 AUC and thrombocytopenia after chemotherapy has statistical significance(p values were 0.0031,0.0412,prediction area under the ROC curve is 0.814).3.Detection of predictive markers of adverse reactions to CAP chemotherapy based on pharmacometabolomics: Based on UHPLC/Q-TOF-MS/MS analysis of endogenous metabolite components in the blood of CRC patients before administration,it was found that OHCU and GCDC could predict nausea after CAP chemotherapy;3-amino-2-piperidone and OHCU can be used to predict vomiting after CAP chemotherapy.Erythronic acid and Demethylphylloquinone can predict diarrhea after CAP chemotherapy.Isovalerylalanine,L-2-amino-3-(4-aminophenyl)propanoic acid can predict HFS after CAP chemotherapy.Merodesmosine can predict leukopenia after CAP chemotherapy;POEA,Linoleoyl ethanolamide,N-oleoyl GABA and N-jasmonoyltyrosine can be used to predict anemia after CAP chemotherapy.N1-methyl-4-pyridone-3-carboxamide,S-methyl-L-cysteinesulfoxide,Glycylproline,Ethyl 2Z,4E-decadienoic acid,(R)-1-O-1,3-octanediol,prolyl-lysine,Saccharolipid,Myristoylglycine,n-oleoyl GABA,N-Jasmonoyltyrosine,PE(P-16:0E /0:0),LYSOPC(15:0),Lysope(0:0/22:6),and G1 Cn AC3 can be used to predict thrombocytopenia after Cap chemotherapy.Based on the above compounds associated with adverse reactions,the area under the ROC curve ranged from 0.929 to 1.4.The correlation and comparison between the two forecasting methods are analyzed and discussed:Correlation analysis showed that the AUC of 5-FU was correlated with Erythronic acid,while the AUC of Fu H2 was correlated with(R)-1-O-1,3-octanediol,Saccharolipid,Myristoylglycine and PE(p-16:0E /0:0).AUC of 5 ’-DFCR is correlated with S-methyl-L-cysteinesulfoxide,Glycylproline,Ethyl 2Z,4E-decadienoic acid,(R)-1-O-1,3-octanediol,Saccharolipid,Myristoylglycine,PE(p-16:0E /0:0),lyso PC(15:0),and G1 Cn AC3.Meanwhile,5 ’-DFCR at 1h,2.5h and 4h time points was found to be correlated with Glycylproline,Saccharolipid,PE(p-16:0E /0:0)and lyso PC(15:0)at 1h,2.5h and 4h time points,as well as S-Methyl-L-cysteinesulfoxide at 2.5h and 4h time points.By comparing the area under the ROC curve of the model,the number of markers,detection methods and other aspects,it was found that the role of pharmacometabolomics in predicting capecitabine adverse reactions in this study was better than that of TDM method.Conclusions:In Chinese colorectal cancer patients,both TDM and pharmacometabolics can be used to predict capecitabine adverse events,and pharmacometabolics(predicting seven adverse events,area under the ROC curve ranged from 0.929 to 1)and TDM(predicting two adverse events,The areas under the ROC curve were 0.724 and 0.814,respectively.The results showed that pharmacometabolomics had a better predictive effect.Meanwhile,correlation analysis of the two methods showed that endogenous metabolites were correlated with the metabolites of capecitabine. |
PDF Full Text Request