In vitro-in vivo extrapolation of CYP2D6 inactivation by paroxetine: Prediction of nonstationary pharmacokinetics and drug interaction magnitude
In vitro-in vivo extrapolation of CYP2D6 inactivation by paroxetine: Prediction of nonstationary pharmacokinetics and drug interaction magnitude
复制标题
DOI：
10.1124/dmd.105.004077
复制
发表时间：
2005-06-01
影响因子：
3.9
通讯作者：
Obach, RS
中科院分区：
文献类型：
--
作者：
Venkatakrishnan, K;Obach, RS
从体外数据预测帕罗西汀引起的药物-药物相互作用的尝试利用了可逆酶抑制模型，迄今为止尚未成功，严重低估了相互作用的程度。最近的数据为帕罗西汀对CYP 2D 6的失活机制提供了证据。我们使用基于抑制机制的数学模型，从体外失活动力学（k（inact）0.17 min（-1），未结合K-I 0.315 μ M）、体内抑制剂浓度和估计的51 h CYP 2D 6降解半衰期预测了帕罗西汀对CYP 2D 6失活的药代动力学后果。模型预测的帕罗西汀蓄积比是单次给药动力学预期的5倍，与观察到的5- 6倍蓄积高度一致。考虑到CYP 2D 6对其口服清除率的贡献，预测了帕罗西汀（20 - 30 mg/天）与地昔帕明、利培酮、奋乃静、托莫西汀、（S）-美托洛尔和（R）美托洛尔产生的相互作用的强度。预测的受害者药物AUC的倍数增加分别为5、6、5、6、4和6倍，并且分别与5、6、7、7、5和8倍的观察值合理一致。未考虑体外微粒体结合对预测准确性产生不利影响。这些预测对模型输入的敏感性模拟表明，当使用14 h的CYP 2D 6降解半衰期（报告为大鼠CYP 3A）时，相互作用幅度预测不足2倍。总之，基于机制的失活比例模型成功预测了帕罗西汀体外数据中CYP 2D 6失活的药代动力学结果。
Attempts at predicting drug-drug interactions perpetrated by paroxetine from in vitro data have utilized reversible enzyme inhibition models and have been unsuccessful to date, grossly underpredicting interaction magnitude. Recent data have provided evidence for mechanism-based inactivation of CYP2D6 by paroxetine. We have predicted the pharmacokinetic consequences of CYP2D6 inactivation by paroxetine from in vitro inactivation kinetics ( k(inact) 0.17 min(-1), unbound K-I 0.315 mu M), in vivo inhibitor concentrations, and an estimated CYP2D6 degradation half-life of 51 h, using a mathematical model of mechanism-based inhibition. The model-predicted accumulation ratio of paroxetine was 5 times that expected from single-dose kinetics and in excellent agreement with the observed 5- to 6-fold greater accumulation. Magnitudes of interactions produced by paroxetine ( 20 - 30 mg/day) with desipramine, risperidone, perphenazine, atomoxetine, (S)-metoprolol, and ( R)metoprolol were predicted, considering the contribution of CYP2D6 to their oral clearance. Predicted fold-increases in victim drug AUC were 5-, 6-, 5-, 6-, 4-, and 6- fold, respectively, and are in reasonable agreement with observed values of 5-, 6-, > 7-, 7-, 5-, and 8-fold, respectively. Failure to consider microsomal binding in vitro adversely affected predictive accuracy. Simulation of the sensitivities of these predictions to model inputs suggests a 2-fold underprediction of interaction magnitude when a CYP2D6 degradation half-life of 14 h ( reported for rat CYP3A) is used. In summary, the scaling model for mechanism-based inactivation successfully predicted the pharmacokinetic consequences of CYP2D6 inactivation by paroxetine from in vitro data.