Population Phamacokinetics Of Sodium Valproate In Epileptic Children

Subject:To establish the PPK model ofvalproate(VPA) in epileptic children using NONlinear Mixed Effect Model(NONMEM) which can give a basis for individualized therapy using Bayesian method.Method:We collected the CLinical information of 320 pediatric inpatients retrospectively.All patients were treated with VPA and we got 391 blood samples of VPA concentrations.The influence of the covariates on relative CLearance of VPA was investigated using NONMEM.Bootstrap was applied to the raw data as internal validation.Another 10 VPA concentrations of the inpatients were estimated using the model as extemal validation.Result:The final regression of CL(L·h-1) was: CL=0.0205+WT×0.0023+Tamt×0.000449+LXXP×0.0209,V=WT×0.593,where WT was total body weight(Kg),Tamt was total daily dose ofVPA(mg),LXXP was total daily dose of CLonazepam(mg),V was apparent volume of distribution(L).The results of Bootstrap resmpling proved the internal validity.In external validation,the regression between predictions(PRED) and dependant variable(DV) is:DV=1.0061×PRED-2.6326(r=0.9404).ConCLusions:(1)The CLearance of VPA will increase while highly dosed or co-administrated with CLonazepam.The concentration of VPA should be monitored while co-administrated with CLonazepam.(2)The model showed good stability and performance,and it can be used to estimate the relative CLearance of VPA.