基于药动-药效模型预测总体参数不确定性对临床试验效能的影响——药物临床研究的计算机模拟

Using CTS and PK-PD Models to Predict the Effect of Uncertainty about Population Parameters on Clinical Trial Power

传统的药物临床研究方案设计大多依赖于经验,缺乏统计学等方面的客观评价,我们提出一种方法,分析参数不确定性对药物临床研究的计算机模拟(Clinical trail simulation,CTS)中效能预测质量的影响及其重要性的评估,为方案的制定提供更为科学和客观的依据。论文基于药动-药效(pharmacokinetics-pharmacodynamics,PK-PD)模型,运用CTS技术,分别给出了仿真模型、处理效应模型和给定PK-PD的临床试验效能模型的建立方法及原理,并讨论了模型的不确定性。对仿真结果,采用方差分析的敏感度分析法并结合η2统计,探讨模型参数的不确定性对CTS模拟结果的影响。经实例仿真,得到相应效能预测值的标准差和其第五百分位数。通过分析计算得到的敏感指数,反映了输入参数的重要性与不重要性,简单直观地表示了参数不确定性对预测结果影响的大小。经仿真试验证明:该方法在一定程度上是有用并且有效的,可为临床试验方案的制定提供科学和客观的依据。

The traditional clinical traiI designs always depend on expert opinions and lack statistical evaluations. In this article, we present a method and illustrate how population parameter uncertainty may he incorporated in the overall simulation model. Using the techniques of clinical trail simulation （CTS） and setting up predictions on the ba- sis of pharmacokinetics-pharmacodynamics （PK-PD） models, we advance the modeling methods for simulation, for treatment effects, and for the clinical trail power under the given PK-PD conditions. Then we discuss the model of uncertainty, suggest an ANOVA-based method, add η2 statistics for sensitivity analysis, and canvass the effect of uncertainty about population parameters on clinical trail power. The results from simulations and the indices derived from this type of sensitivity analysis may be used for grading the influence on the prediction quality of uncertainty about different population parameters. The experiment results are satisfactory and the approach presented has practical value in clinical trails.