Premium
Physiologically based pharmacokinetic modeling of tramadol to inform dose adjustment and drug‐drug interactions according to CYP2D6 phenotypes
Author(s) -
Xu Miao,
Zheng Liang,
Zeng Jin,
Xu Wenwen,
Jiang Xuehua,
Wang Ling
Publication year - 2021
Publication title -
pharmacotherapy: the journal of human pharmacology and drug therapy
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.227
H-Index - 109
eISSN - 1875-9114
pISSN - 0277-0008
DOI - 10.1002/phar.2494
Subject(s) - cyp2d6 , physiologically based pharmacokinetic modelling , tramadol , pharmacokinetics , pharmacology , duloxetine , active metabolite , paroxetine , metabolite , population , drug , area under the curve , medicine , chemistry , cytochrome p450 , analgesic , antidepressant , environmental health , pathology , metabolism , hippocampus , alternative medicine
Objectives The objective of this study was to establish physiologically based pharmacokinetic (PBPK) models of tramadol and its active metabolite O‐desmethyltramadol (M1) and to explore the influence of CYP2D6 gene polymorphism on the pharmacokinetics of tramadol and M1. Furthermore, we used PBPK modeling to prospectively predict the extent of drug‐drug interactions (DDIs) in the presence of genetic polymorphisms when tramadol was co‐administered with the CYP2D6 inhibitors duloxetine and paroxetine. Methods Plasma concentrations of tramadol and M1 were used to adjust the turnover frequency (K cat ) of CYP2D6 for phenotype populations with different CYP2D6 genotypes. PBPK models were developed to capture the pharmacokinetics between CYP2D6 extensive metabolizers (EMs), intermediate metabolizers (IMs), poor metabolizers (PMs), and ultra‐rapid metabolizers (UMs). The validated models were then used to support dose adjustment in different CYP2D6 phenotypes and to predict the extent of CYP2D6‐mediated DDIs when tramadol was co‐administered with paroxetine or duloxetine. Results The PBPK models we built accurately describe tramadol and M1 exposure in the population with different CYP2D6 phenotypes. In our prediction, the area under the concentration‐time curve (AUC inf‐tDlast ) of M1 is 70% lower in PMs than in EMs, 27% lower in IMs, and 15% higher in UMs. Based on the models we built, we suggest that the oral dose of tramadol should be 50% higher for IMs and 25% lower for UMs to achieve an approximately equivalent plasma exposure of M1 as in EMs. When tramadol was co‐administered with paroxetine or duloxetine, the magnitude of the inhibitor‐substrate interaction was lowest in EMs (0.45), secondary in IMs (0.39), and highest in PMs (0.18) in terms of M1. Conclusion The current example uses the PBPK model to guide dose adjustment of tramadol and to predict the effect of CYP2D6 genetic polymorphisms on DDIs for rational clinical use of tramadol in the future.