Introduction
Rivaroxaban, an outstanding representative of a non−vitamin K oral anticoagulant, directly inhibits Factor Xa to block the production and reduce the activation of thrombin [1]. Compared with vitamin K anticoagulants, rivaroxaban exerts a more specific and powerful anticoagulant effect, and has been approved mainly for treatment and prevention of deep venous thrombosis, pulmonary embolism and systemic embolism from nonvalvular atrial fibrillation [2]. In the evaluation of safety and pharmacokinetic stability, rivaroxaban surpasses established anti-coagulant agents; however, bleeding risk still exits [3]. A systematic review and meta-analysis of the efficiency and safety of direct oral anticoagulants approved for treating or preventing cardiovascular thromboembolism complications showed that rivaroxaban did not outperform warfarin in terms of gastrointestinal bleeding risk [4]. Indeed, several bleeding events have been reported when rivaroxaban was applied to prevent stroke and systemic embolism for atrial fibrillation patients, especially when used in combination with other heart rate control drugs [5-7].
Previous studies have investigated the metabolism and elimination of rivaroxaban, with cytochrome P450 (CYP) enzymes, mainly CYP2J2 and CYP3A4, and a few liver hydrolytic enzymes playing an important role in the deactivation of rivaroxaban [8 9]. The major metabolites and metabolic pathways were identified by in vitro liver microsome incubation studies, and morpholinone 2- hydroxylation (M1) was identified as the structure of the major rivaroxaban metabolite by H1 NMR analysis (Figure 1) [8]. As previously reported, the proportion of rivaroxaban metabolized by CYP enzymes represents approximately two-thirds of a given dose, and the remaining one-third is eliminated by secretion mediated by P-glycoprotein (P-gp) and breast cancer resistance protein (BCRP) [8 10 11]. Pharmacokinetic interactions between rivaroxaban and drugs for regulating CYP3A4 and P-gp have been extensively evaluated, with outcomes indicating that caution is warranted when it is used concomitantly with strong CYP3A4 and P-gp inhibitors [12-14]. Notably, Mueck et al. found that rivaroxaban co-administrated with strong or moderate CYP3A4 inhibitors―such as clarithromycin and fluconazole―did not cause clinically relevant interactions for rivaroxaban [12 14 15]. In addition, bleeding events do exist for combining with other agents in clinical practice, which are not limited to CYP3A4 and P-gp inhibitors [5-7 16 17]. Taken together, we hypothesize that CYP3A4 is not the predominant isoform involved in the metabolism of rivaroxaban, and that other CYP isoforms likely participate to a larger extent in rivaroxaban morpholine 2-hydroxylation [12 18].
In the present study, we systematically evaluated the participation and contribution of a series of CYP isoforms in the metabolism of rivaroxaban by product formation analysis in human liver microsomes (HLMs) and recombinant human CYPs, as well as CYP-specific inhibition studies.