应用群体药动学-药效学结合模型评估大剂量甲氨蝶呤化疗后的骨髓抑制

摘要

目的:建立群体药动学(PPK)-药效学(PD)模型评估大剂量甲氨蝶呤(HDMTX)化疗后的骨髓抑制效应,促进个体化用药.方法:收集144例急性淋巴细胞白血病(ALL)患儿行HDMTX化疗的临床资料.以MTX血清药物浓度作为药动学指标,化疗5d后的白细胞计数减少率作为药效学指标,采用非线性混合效应建模法进行数据分析.药动学模型选用二房室开放式模型,药效学模型选用带效应室的Sigmoid-Emax模型.用拟合优度(goodness-of-fit)、自举法(Bootstrap)和正态化预测分布误差(NPDE)对最终模型的预测性能进行验证.结果:PPK-PD最终模型参数的群体典型值分别为:V1(中央室分布容积)=15.46 L,V2(周边室分布容积)=1.95 L,CL(表观清除率)=4.76 L·h-1,CL2(周边室清除率)=0.11 L·h-1,ke0(效应室消除速率常数)=0.003 6 h-1,EC50(达最大效应一半时效应室浓度)=0.87 mg·L-1,γ(陡度因子)=1.91,Emax(最大效应参数)=79.87％.化疗期间的总碱化量(碳酸氢钠)对MTX的表观清除率有显著影响.拟合优度、自举验证和NPDE结果表明,最终模型稳定,预测结果可靠.结论:本研究成功建立了用于评估HDMTX化疗后骨髓抑制程度的PPK-PD模型,可为临床优化给药方案提供帮助.%Objective:To establish a population pharmacokinetics-pharmacodynamics (PPK-PD) model to evaluate the bone marrow depression following high-dose methotrexate (HDMTX) chemotherapy,and to facilitate individualized therapeutic regimens.Methods:Data from 114 children with acute lymphoblastic leukemia (ALL) during HDMTX treatment were collected.The serum concentrations of MTX were selected as the pharmacokinetics index,and decreased percentage in white blood count (WBC) after 5 days from HDMTX infusion as pharmacodynamics index,respectively.A nonlinear mixed-effects population modelling approach was carried out for data analysis.A two-compartment,open kinetic model and a sigmoid Emax model with an effect compartment were selected as the structure models for pharmacokinetic and pharmacodynamic analysis,respectively.The predictive performance of final model was evaluated by goodness-of-fit,Bootstrapping and normalized predictive distribution error (NPDE).Results:The typical population values of pharmacokinetic and pharmacodynamic parameters estimated in final model were as follows:V1 (volume of the central compartment) =15.46 L,V2 (volume of the peripheral compartment) =1.95 L,CL (total body clearance) =4.76 L· h-1,CL2 (intercompartmental clearance) =0.11 L· h-1,ke0 (elimination rate constant of the effect compartment) =0.003 6 h-1,EC50 (the effect compartment concentration resulting in 50％ of maximum effect) =0.87 mg· L-1,γ(sigmoidicity factor) =1.91,and Emax (maximum drug effect) =79.87％.The final model structure indicated that MTX clearance was impacted by total sodium bicarbonate dose during chemotherapy.The stability and the predictive performance of final model were accepted by goodness-of-fit,Bootstrapping and NPDE.Conclusion:The PPK-PD model has been successfully established in this study to evaluate the bone marrow depression following HDMTX chemotherapy,which will be useful for optimizing the administration protocol of HDMTX.