Population Pharmacokinetics And Limited Sampling Strategy For Oral Tacrolimus In Chinese Liver Transplant Patients

OBJECTIVE:To establish a population pharmacokinetics model of tacrolimus orally administed from routing therapeutic drug monitoring in Chinese liver transplant recipients.To establisth limited sampling schedules of tacrolimus in Chinese liver transplant,and provide rational therapeutic optimization for individual dosage.METHODS:1.37 recipients after liver transplantation were given tacrolimus basedimmunosurpressive regime 24 hours after surgery.once dose per 12 h.Blood samples were drawn from vein into EDTA tube before mornig dose at 8 AM and the whole blood concentration was measured by MEIA.The CYP3A5 genotype was detected meanwhile.The sparse blood samples were analyzed by population pharmacokinetics software NONMEM.The genotype,demographic characteristics,biochemical indexs were screened with one by one added and eliminated method and general additive model method.2.26 recipients after liver transplantation were also given tacrolimus-based immunosurpressive regimen,11 blood samples were collected predose and 0.3,1,1.5,2,3,4,6,8,10,12 posedose,blood concentration was measured by MEIA.The pharmacokinetic parameters were calculated using WinNolin,the correlation between individual blood time between AUC were observed and limited sampling schedules were established by regression models and validated by plots.3.The rich blood samples from 26 recipients were estimated by NONMEM,and individual blood concentrations were feedback by Bayesian,the limited samples schedules were established base on these new concentrations.RESULTS:1.One compartment with fixed Ka fitted the sparse blood sample well.GAM has a good effect on screening covariates.The final population pharmacokinetics parameters were as follows respectively:CL/F = 22 L·h-1,V/F =348 L.CYP3A5 genotypes,BUN,ALP have influence on CL/F,ALP have influence on V/F after incorporated the covariates into the final model.CL/F with expressive CYP3A5 genotype was higher than non-expressive CYP3A5 when did not consider other factors.2.26 recipients blood samples were analyzed with Non-compartment method,and the pharmacokinetic parameters were as follows respectively:trough concentration was 5.75±.84μg·L-1,AUC0-12 was 115.6±31.5μg·h·L-1,CL/F was 0.31L·h-1·kg-1.The best correlation between one point with AUC was show on C3 and C4 with R2 value 0.9231 and 0.9397 respectively,the C0 with AUC was 0.599.The LSS equation was established with C0 and C4 and the bias and precision were<15%.3.Two compartment with a lag time fitted the 26 recipients blood concentration,population parameters were as follows:tlag was 0.206 h,CL/F was 27 L,V2/F was 390 L,V3/F was 484 L,Q was 77.8 L·h-1,Ka was 0.929 h-1.Non-compartment method was used to analyzed the Bayesian feedback individual concentrations and the parameters were as follows:Cmax was 10.38±5.64μg·L-1(3.15-24.59),AUC0-12 was 87.91±37.59μg·h·L-1(39.48-162.20),CL/F was 14.02±7.38L·h-1,V/F was 390.77±173.89 L(65.95-826.64).The one point C3,C4,C6 show a good correlation with AUC,and R2 values were 0.9229,0.9792,0.9383 respectively.C0 and AUC was 0.8258.Two-point equations C0-C2,C0-C3,C0-C4,three-point equations C0-C2-C3,C0-C2-C4,C2-C3-C4 have better predictive performance compare to other LSS models.CONCLUSION:A population model was established in 37 recipients after liver transplantation for oral administrated tacrolimus.The CYP3A5 genotype has a great influence on CL/F.LSS models may provide a measure to predict AUC and foundation for individual dosage.