Biomarker Study Of Capecitabine-induced Diarrhea And Hand-foot Syndrome Based On Drug Concentration And Metabolomics

Objective:A capecitabine-induced hand-foot syndrome and diarrhea ICR mouse model was established,and the contents of capecitabine and metabolites were determined in the collected mouse skin and small intestine,and non-targeted metabolomics analysis of plasma samples was performed to elucidate the Metabolomics mechanism of capecitabine-induced HFS and diarrhea,screening of capecitabine-induced hand-foot syndrome and diarrhea biomarkers in mice.Methods:For 14 consecutive days,twice a day,mice were given capecitabine sodium carboxymethyl cellulose suspension by gavage at a dose of 275 mg/kg to establish a mouse model of HFS and diarrhea.To verify whether the animal model was successfully established,the characteristic changes of the feces,the color of the stool,and the morphological changes were used to verify whether the animal model was successfully established.The concentrations of capecitabine and its metabolites were determined by LC-MS/MS technology;in positive and negative ion mode,ultra-high performance liquid chromatography tandem mass spectrometry(UPLC-QTOF-MS)was used to determine the plasma mass spectrometry data of mice before and after administration.After the software Masslynxv4.2 was processed,Progenesis QI and EZinfo 3.0.3.0 were used for data correction and multivariate statistical analysis,including principal component analysis(PCA)and partial least squares-discriminant analysis(OPLS-DA)etc,and finally combined with the software SPSS 22.0 Statistical tests were performed;the chemical structures of different compounds were identified by comparing the HMDB and Metlin databases,and the molecular weight and secondary structure were accurate according to the score correlation;the metabolic pathways were identified using the IMPa LA and Metabo Analyst 5.0 databases.The samples were divided into three groups: diarrhea,HFS,and combined occurrence according to the baseline before administration,and the experimental group after administration.The classification is the same as above.Correlation analysis was carried out among the three adverse reactions caused by pettabine,and the international standardized ratio receiver operating characteristic curve(ROC)was used to evaluate the diagnostic value of differential metabolites.Logistic regression model was established to evaluate the early warning effect of adverse reactions and ROC diagnostic significance.Results:In animal experiments,the occurrence of adverse reactions was judged by visual observation and HE staining.15 rats had diarrhea,with an incidence rate of 45.46%;19rats had positive hand-foot reaction,with an incidence rate of 57.58%;and 22 rats had at least one adverse reaction,with an incidence rate of 66.67%.11 rats did not have any adverse reactions,accounting for 33.33%.Cap and metabolite concentration values(Mean ± SD)ng/mg: Cap 33.68 ±19.77ng/mg,5’-DFCR 7.80±14.70ng/mg,5’-DFUR 84.39±113.96ng/mg and 5’-DFUR in the small intestine of diarrhea group The content of-FU 74.00±51.27ng/mg has obvious statistical significance,and the concentration difference between occurrence and non-occurrence is large,especially the difference of 5’-DFCR is as much as 10 times,and the concentration of each substance in the plasma of diarrhea mice is similar Large,and mostly not statistically significant;2-DFCR 6.63±2.90ng/mg,5-FU 2.72±0.66ng/mg,5’-DFUR171.15±219.60ng/mg in HFS skin,Cap prototype The concentration content is high,but the difference is not significant,and there is no statistical significance.Only the Cap content in HFS plasma has statistical significance,but the concentration is lower.Non-targeted metabolism: 40 differential metabolites were identified as potential biomarkers.The baseline group included HFS compared with none-HFS: DL-citrulline,L-Ornithine,L-Glutamate,Cystinyl-bis-glycine,etc;Diarrhea compared with non diarrhea:Ornithine,DL-Glyceric acid,L-Glyceric acid,Taurocholic acid,Cholic acid,Cystinyl-bis-glycine,etc;At least one adverse reaction did not occur: DL-Glyceric acid,DL-citrulline,Calcitroic.acid,D-Lysinol,Tryptophyl.Leucine,Deoxycolol,etc.The above metabolites are involved in the following metabolic pathways: amino acid metabolism,bile acid metabolism,hormone metabolism,carbohydrate metabolism and uric acid purine metabolism.Among them,there are 15 amino acids(including different configurations,the same below),9 bile acids,3 fatty acids,5 sugar alcohols,1 hormone testosterone,1xanthine and others.ROC and logistic regression of diarrhea in the baseline group found that Ornithine,Taurocholic acid and Cystinyl-bis-glycine had high diagnostic value,among which Cystinyl-bis-glycine was the most valuable.After modeling alone,the cutoff value was-0.2705,the excellence index was 0.495,and the sensitivity of diarrhea diagnosis was61.5%.Combined with Taurocholic acid,Cystinyl-bis-glycine,Palmitoyglucitol Diagnostic equation of glucose alcohol diarrhea: Y =-0.002 * Palmitoyglucitol + 0.00000196 *Taurocholic acid + 0.0000239 * Cytinylbisglycine-1.8.At least one adverse reaction occurred.Three substances were screened out in the baseline group: D-Lysinol,Deoxycolol and DL-Glyceric acid.The combined diagnosis was better than that of a single substance.The AUC area of the combined diagnosis was 0.764,which had a high accuracy value.Due to the large original value,after log2 processing of the original data,the combined diagnosis equation was: Y = D-lysinol *(-0.6498)+ Deoxycolol *(-0.2603)+DL-Glyceric.Acid *(-0.1566).Finally,when the excellence index is 0.73,the sensitivity is0.733,the specificity is 1,and the cutoff value is 0.691.Conclusion:Taurocholic acid,Cholic acid,Ornithine,Arginine and other different metabolites are related to HFS and diarrhea.Identifying metabolites may help to understand the pathogenesis of diarrhea and HFS,and provide reference for the prevention and treatment of adverse reactions of Cap;The logistic regression model was established and the ROC curve evaluation model was used to evaluate the early warning effect and diagnostic value of adverse reactions.The results showed that the diagnostic model of diarrhea was established by combining Taurocholic acid,Cystinyl-bis-glycine and Palmitoyglucitol,which was found to have high diagnostic value for the occurrence of diarrhea;Combining D-lysinol,Deoxycolol and DL-Glyceric acid to establish a diagnostic model with at least one adverse reaction also has diagnostic value.Therefore,it is feasible to use logistic regression and ROC curve to predict diarrhea or at least one adverse reaction after capecitabine oral administration.