Damoiseaux David, Li Wenlong, Martínez-Chávez Alejandra, Beijnen Jos H, Schinkel Alfred H, Huitema Alwin D R, Dorlo Thomas P C
Department of Pharmacy & Pharmacology, The Netherlands Cancer Institute, 1066 CX Amsterdam, The Netherlands.
Division of Pharmacology, The Netherlands Cancer Institute, 1066 CX Amsterdam, The Netherlands.
Pharmaceuticals (Basel). 2022 Jul 13;15(7):860. doi: 10.3390/ph15070860.
The extrapolation of drug exposure between species remains a challenging step in drug development, contributing to the low success rate of drug approval. As a consequence, extrapolation of toxicology from animal models to humans to evaluate safe, first-in-human (FIH) doses requires high safety margins. We hypothesized that a human-CYP3A4-expressing transgenic (Cyp3aXAV) mouse is a more predictive model for human drug exposure of CYP3A4-metabolized small-molecule drugs. Population pharmacokinetic models based on wild-type (WT) and Cyp3aXAV mouse pharmacokinetic data of oral lorlatinib, brigatinib, ribociclib and fisogatinib were allometrically scaled and compared to human exposure. Extrapolation of the Cyp3aXAV mouse model closely predicted the observed human exposure for lorlatinib and brigatinib with a 1.1-fold and 1.0-fold difference, respectively, compared to a 2.1-fold and 1.9-fold deviation for WT-based extrapolations of lorlatinib and brigatinib, respectively. For ribociclib, the extrapolated WT mouse model gave better predictions with a 1.0-fold deviation compared to a 0.3-fold deviation for the extrapolated Cyp3aXAV mouse model. Due to the lack of a human population pharmacokinetic model for fisogatinib, only median maximum concentration ratios were calculated, resulting in ratios of 1.0 and 0.6 for WT and Cyp3aXAV mice extrapolations, respectively. The more accurate predictions of human exposure in preclinical research based on the Cyp3aXAV mouse model can ultimately result in FIH doses associated with improved safety and efficacy and in higher success rates in drug development.
物种间药物暴露的外推仍然是药物开发中具有挑战性的一步，这导致药物获批成功率较低。因此，从动物模型外推毒理学以评估安全的首次人体（FIH）剂量需要很高的安全边际。我们假设，表达人CYP3A4的转基因（Cyp3aXAV）小鼠对于CYP3A4代谢的小分子药物的人体药物暴露是一个更具预测性的模型。基于野生型（WT）和Cyp3aXAV小鼠口服洛拉替尼、布加替尼、瑞博西尼和菲索加替尼的药代动力学数据的群体药代动力学模型进行了异速缩放，并与人体暴露情况进行了比较。Cyp3aXAV小鼠模型的外推分别以1.1倍和1.0倍的差异紧密预测了洛拉替尼和布加替尼观察到的人体暴露，相比之下，基于WT的洛拉替尼和布加替尼外推的偏差分别为2.1倍和1.9倍。对于瑞博西尼，外推的WT小鼠模型预测更好，偏差为1.0倍，而外推的Cyp3aXAV小鼠模型的偏差为0.3倍。由于缺乏菲索加替尼的人体群体药代动力学模型，仅计算了最大浓度中位数比值，WT和Cyp3aXAV小鼠外推的比值分别为1.0和0.6。基于Cyp3aXAV小鼠模型在临床前研究中对人体暴露更准确的预测最终可导致与安全性和有效性改善相关的FIH剂量，并提高药物开发的成功率。
