Johannes Tillil 1,2, Charlotte Kloft 3, Wilhelm Huisinga 2
1 PharMetrX Graduate Research Training Program: Pharmacometrics & Computational Disease Modelling (Freie Universität Berlin and University of Potsdam, Germany), 2 Institute of Mathematics (University of Potsdam, Germany), 3 Institute of Pharmacy (Freie Universität Berlin, Germany)
Objectives
For patients receiving long-term anticoagulant therapy, various clinical situations may arise in which it is necessary to transition from one type of anticoagulant to another, for example, from a direct oral anticoagulant (DOAC) like rivaroxaban to a vitamin-K antagonist (VKA) like warfarin. When transitioning to and from warfarin, it is crucial that anticoagulation is adequately maintained, because warfarin has a narrow therapeutic window with a risk of bleeding or thrombosis if administered at super- or subtherapeutic concentrations [1].
At therapeutic concentrations, rivaroxaban increases the prothrombin time (PT) and international normalized ratio (INR) used for titrating warfarin and has a synergistic effect on PT prolongation if coadministered with warfarin [2]. Such interference may be a potential cause of adverse events in DOAC to VKA transitioning as inaccurate PT readings can lead to inadequate anticoagulation [3]. Combining the interindividual variability from rivaroxaban and warfarin PKPD models with a QSP model of the coagulation cascade, we assess the risk of subtherapeutic INRs during the transition from rivaroxaban to warfarin in an atrial fibrillation virtual patient population based on the recommendations from an up-to-date practical guide.
Methods
The QSP model [4] has previously been validated to predict the PT for warfarin and the activated partial thromboplastin time (aPTT) for rivaroxaban [5]. To accurately simulate the rivaroxaban effect on the PT we added additional inhibition of the factor Xa-mediated factor II activation, as the prothrombinase complex does not influence the PT test at high tissue factor inputs [6]. Characteristics of virtual patients with atrial fibrillation were sampled based on the parameter estimates and covariate distributions in a population PK model for rivaroxaban [7] and a population PKPD model for warfarin [8], with parameters adapted to racemic warfarin following [9]. Interindividual variability of the clotting factors was included by adding 20% CV on the factor synthesis rates.
1000 virtual patients were given a standard dose of 20 mg QD rivaroxaban in dosing steady state and transitioning to warfarin with a standard dose of 5 mg QD. As recommended in the “2021 European Heart Rythm Association practical guide on the use of non-vitamin K antagonist oral anticoagulants in patients with atrial fibrillation” [10], rivaroxaban therapy is continued until 4 days after initiation of warfarin therapy when the INR is measured at the rivaroxaban trough concentration. The therapeutic window of warfarin in atrial fibrillation is an INR between 2 and 3. In case of INR > 2, rivaroxaban is discontinued and the INR evaluated 1 day later. In case of INR < 2, rivaroxaban therapy is continued and the INR evaluated 1-3 days later. Results In 668 (67%) patients, an INR between 2 and 3 was measured, prompting discontinuation of rivaroxaban and no immediate change of the warfarin dose. For patients with low rivaroxaban clearance, the INR was still elevated at the rivaroxaban trough on the fourth day of coadministration, indicating a spurious therapeutic warfarin effect that decayed with the elimination of rivaroxaban. Thus, for 153 (23%) of these patients the INR dropped to subtherapeutic levels below 2 in the next day followup, persisting to the second day of followup for 130 (19%) patients. The INR at the rivaroxaban trough concentration on the 4th day of coadministration was increased by 28% on average compared to warfarin monotherapy. Due to the synergistic effect of warfarin and rivaroxaban, the INR was increased by 76% on average at peak rivaroxaban concentrations. In our simulated population, the mean INR after 10 days of warfarin therapy was 2.63 with 634 (63%) of patients achieving a therapeutic INR between 2 and 3 at the 5 mg QD dosing regimen. Conclusions The QSP model-based simulations recover the additive to synergistic effect of rivaroxaban and warfarin coadministration on PT test readouts. Assessing the transition from rivaroxaban to warfarin, our simulations indicated that 15% of a hypothetical atrial fibrillation patient population has the potential for subtherapeutic anticoagulation after discontinuation of rivaroxaban. However, the clinical significance of the subtherapeutic INR measurements remains to be investigated. In future work, we aim to link these measurements to in-vivo bleeding and thrombosis risk scenarios. References: [1] K. T. Moore et al. “Switching from Rivaroxaban to Warfarin: An Open Label Pharmacodynamic Study in Healthy Subjects”. In: British Journal of Clinical Pharmacology 79.6 (2015), pp. 907–917. [2] H.-U. Siegmund et al. “Contribution of Rivaroxaban to the International Normalized Ratio When Switching to Warfarin for Anticoagulation as Determined by Simulation Studies”. In: British Journal of Clinical Pharmacology 79.6 (June 2015), pp. 959–966. [3] M. R. Patel et al. “Outcomes of Discontinuing Rivaroxaban Compared With Warfarin in Patients With Nonvalvular Atrial Fibrillation”. In: Journal of the American College of Cardiology 61.6 (Feb. 2013), pp. 651–658. [4] T. Wajima, G. K. Isbister, and S. B. Duffull. “A Comprehensive Model for the Humoral Coagulation Network in Humans”. In: Clinical Pharmacology & Therapeutics 86.3 (2009), pp. 290–298. [5] S. Hartmann et al. “Quantitative Systems Pharmacology Model-Based Predictions of Clinical Endpoints to Optimize Warfarin and Rivaroxaban Anti-Thrombosis Therapy”. In: Frontiers in Pharmacology 11 (July 2020). [6] J. Knöchel, C. Kloft, and W. Huisinga. “Understanding and Reducing Complex Systems Pharmacology Models Based on a Novel Input–Response Index”. In: Journal of Pharmacokinetics and Pharmacodynamics 45.1 (Feb. 2018), pp. 139–157. [7] W. Mueck et al. “Rivaroxaban”. In: Clinical Pharmacokinetics 50.10 (Oct. 2011), pp. 675–686. [8] A.-K. Hamberg et al. “A Pharmacometric Model Describing the Relationship Between Warfarin Dose and INR Response With Respect to Variations in CYP2C9, VKORC1, and Age”. In: Clinical Pharmacology & Therapeutics 87.6 (2010), pp. 727–734. [9] U. Falkenhagen et al. “Deriving Mechanism-Based Pharmacodynamic Models by Reducing Quantitative Systems Pharmacology Models: An Application to Warfarin”. In: CPT: Pharmacometrics & Systems Pharmacology 12.4 (2023), pp. 432–443. [10] J. Steffel et al. “2021 European Heart Rhythm Association Practical Guide on the Use of Non-Vitamin K Antagonist Oral Anticoagulants in Patients with Atrial Fibrillation”. In: EP Europace 23.10 (Oct. 2021), pp. 1612–1676.
Reference: PAGE 34 (2026) Abstr 12141 [www.page-meeting.org/?abstract=12141]
Poster: Drug/Disease Modelling - Safety