摘要
目的建立中国癫痫儿童丙戊酸钠(VPA)的群体药动学(PPK)模型,促进个体化用药。方法回顾性地收集246例癫痫患儿应用VPA的临床数据,血药浓度都是常规监测的稳态浓度。将患儿分为两组,PPK模型组(n=146):用NONMEM软件求算PPK参数值,建立基本模型和最终模型;PPK验证组(n=100):分别用基本模型和最终模型预测PPK验证组患儿的血药浓度,比较两模型的平均预测误差(MPE)、预测误差均方(MSPE)、平均预测误差平方根(RMSPE)和权重残差(WRES),进行模型验证。结果用NONMEM法拟合为一室一级吸收模型,PPK基本模型的参数为:Ka2.38 h-1,Vd 6.54 L,CL0.232 L.h-1;最终模型参数为:Ka2.38 h-1,Vd(2.03+0.189).体重(L),CL(0.097 8+0.010 4).年龄(L.h-1)。基本模型经PPK验证组检验的结果,MPE,MSPE,RMSPE,WRES及其95%C I分别为:-25.03(-32.74,-17.32),4 105.71(2 803.93,5 407.49),44.34(38.29,50.39),-0.03(-0.11,0.05);最终模型经PPK验证组检验的结果,MSE,MSPE,RMSPE,WRES及其95%C I分别为:-1.29(-6,3.4),1 268.38(1 050.64,1 486.12),28.46(25.66,31.26),0.12(-0.04,0.28)。最终模型比基本模型更优化、预测更准确。结论用NONMEM软件成功地建立了中国癫痫儿童VPA的PPK模型。
OBJECTIVE To set up a population pharmacokinetic(PPK) model of Valproate (VPA) in children with epilepsy in China, and promote reasonable use of antiepileptic drugs (AEDs) in clinical practice. METHODS Sparse data of VPA serum concentrations from 246 pediatric children were collected. These patients were divided into two groups: PPK-MOdel group, n = 146; and PPK- Valid group, n = 100. Using NONMEM software, PPK parameter values of VPA for 146 patients in PPK-Model group were calculated, and a basic model and a final model were built. To valid the basic model and the final model, the concentrations of 100 patients were predicted in PPK-Valid group by two models respectively. To assess the accuracy and precision of the concentration prediction, mean prediction error( MPE), mean squared prediction error(MSPE) , root mean squared prediction error( RMSPE), weight residue(WRES) and its 95% CI were all calculated. Then, these values of two models were compared. RESULTS PPK of VPA was described by one-compartment model with first order process. The parameters of basic model were as follows : Ka 2. 38 h^- 1, Vd 6. 54 L, CL 0. 232 L · h^ - 1 ; the parameters of final model were as follows : Ka 2. 38 h^ - 1, Vd (2.03 ± 0. 189 ) - WEIGHT ( L), CL (0. 097 8 ±0. 010 4) · AGE (L · h^-1 ). For the basic model, MPE, MSPE, RMSPE, WRES and its 95% CI were as follows: -25.03 ( -32.74, - 17. 32), 4 105.71 (2 803.93, 5 407.49), 44. 34 (38.29, 50. 39), -0. 03 ( -0. 11,0. 05). For the final model, MPE, MSPE,RMSPE, WRES and its 95% CI were as follows: - 1.29 ( -6,3.4), 1 268. 38 ( 1 050. 64,1 486. 12), 28.46 (25.66,31.26) ,0. 12 ( - 0. 04,0. 28). The final model was much better than the basic model. CONCLUSION A PPK model of VPA in children with epilepsy in China is successfully established using the NONMEM software.
出处
《中国药学杂志》
CAS
CSCD
北大核心
2007年第4期291-294,317,共5页
Chinese Pharmaceutical Journal
关键词
群体药动学
NONMEM法
癫痫
儿童
丙戊酸钠
population pharmacokinetics
NONMEM software
epilepsy
chilidren
valproate