Grime Kenneth H, Bird James, Ferguson Douglas, Riley Robert J
Department of Discovery DMPK, AstraZeneca R&D Charnwood, Bakewell Road, Loughborough LE115RH, UK.
Eur J Pharm Sci. 2009 Feb 15;36(2-3):175-91. doi: 10.1016/j.ejps.2008.10.002. Epub 2008 Nov 1.
The ability to use in vitro human cytochrome P450 (CYP) time-dependent inhibition (TDI) data for in vivo drug-drug interaction (DDI) predictions should be viewed as a prerequisite to generating the data. Important terms in making such predictions are k(inact) and K(I) but first-line screening assays typically involve characterisation of an IC(50) value or a time dependent shift in IC(50). In the work presented here, two key screening methods from the scientific literature were appraised both in terms of practicality and quality of k(inact)/K(I) estimation. The utility of TDI screening data in DDI predictions was investigated and particular reference given to a simple DDI simulation model based on a spreadsheet that calculates the systemic exposure of unbound inhibitor drug following the input of human pharmacokinetic parameters. Using several clinical mechanism-based CYP DDI examples, the effectiveness of the approach was assessed and compared to other widely available approaches (a simple algorithm that employs a single in vivo unbound inhibitor concentration, a seven-compartment physiologically based pharmacokinetic (PBPK) model that defines the extent of interaction as a result of hepatic inhibitor concentrations and the commercially available software SimCYP). All the methods gave predictions that compared favourably with the observed DDIs, but various advantages and disadvantages of each were also given full consideration. The new model facilitates rapid sensitivity analysis (parameters can be easily input and altered to give a visual representation of the impact on the active enzyme concentration) and it was therefore used to derive "rules of thumb" demonstrating the relationship between extent of DDI, time-dependent IC(50) and dose for typical acidic and basic drugs. Additionally, a TDI decision tree linking into reactive metabolite investigations is proposed for use in a Drug Discovery setting.
将体外人细胞色素P450（CYP）时间依赖性抑制（TDI）数据用于体内药物相互作用（DDI）预测的能力，应被视为生成此类数据的先决条件。进行此类预测的重要参数是k（inact）和K（I），但一线筛选试验通常涉及IC（50）值的表征或IC（50）的时间依赖性变化。在本文介绍的工作中，对科学文献中的两种关键筛选方法在k（inact）/K（I）估计的实用性和质量方面进行了评估。研究了TDI筛选数据在DDI预测中的效用，并特别参考了一个基于电子表格的简单DDI模拟模型，该模型在输入人体药代动力学参数后计算未结合抑制剂药物的全身暴露量。使用几个基于临床机制的CYP DDI实例，评估了该方法的有效性，并与其他广泛使用的方法进行了比较（一种采用单一体内未结合抑制剂浓度的简单算法、一个定义由于肝脏抑制剂浓度导致的相互作用程度的七室生理药代动力学（PBPK）模型以及商业可用软件SimCYP）。所有方法给出的预测与观察到的DDI结果相比都很好，但也充分考虑了每种方法的各种优缺点。新模型便于进行快速敏感性分析（可以轻松输入和更改参数，以直观显示对活性酶浓度的影响），因此用于推导“经验法则”，以证明典型酸性和碱性药物的DDI程度、时间依赖性IC（50）和剂量之间的关系。此外，还提出了一个与反应性代谢物研究相关的TDI决策树，用于药物发现环境。
