SEARCH

SEARCH BY CITATION

Keywords:

  • anticoagulants/administration and dosage;
  • anticoagulants/pharmacokinetics;
  • atrial fibrillation/drug therapy;
  • humans;
  • rivaroxaban PK/PD

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Declaration of Conflicting Interests
  8. Funding
  9. Acknowledgments
  10. References
  11. Supporting Information

Two once-daily rivaroxaban dosing regimens were compared with warfarin for stroke prevention in patients with non-valvular atrial fibrillation in ROCKET AF: 20 mg for patients with normal/mildly impaired renal function and 15 mg for patients with moderate renal impairment. Rivaroxaban population pharmacokinetic (PK)/pharmacodynamic (PD) modeling data from ROCKET AF patients (n = 161) are reported and are used to confirm established rivaroxaban PK and PK/PD models and to re-estimate values of the models' parameters for the current AF population. An oral one-compartment model with first-order absorption adequately described rivaroxaban PK. Age, renal function, and lean body mass influenced the PK model. Prothrombin time and prothrombinase-induced clotting time exhibited a near-linear relationship with rivaroxaban plasma concentration; inhibitory effects were observed through to 24 hours post-dose. Rivaroxaban plasma concentration and factor Xa activity had an inhibitory maximum-effect (Emax) relationship. Renal function (on prothrombin time; prothrombinase-induced clotting time) and age (on factor Xa activity) had moderate effects on PK/PD models. PK and PK/PD models were shown to be adequate for describing the current dataset. These findings confirm the modeling and empirical results that led to the selection of doses tested against warfarin in ROCKET AF.

The oral, direct factor Xa active site inhibitor rivaroxaban has been approved in the United States and European Union for stroke prevention in patients with non-valvular atrial fibrillation (AF),[1, 2] based on results from the global phase III ROCKET AF (Rivaroxaban once daily Oral direct factor Xa inhibition Compared with vitamin K antagonism for prevention of stroke and Embolism Trial in Atrial Fibrillation) trial.[3] Phase I clinical pharmacology studies determined that rivaroxaban has predictable pharmacokinetics (PK) and pharmacodynamics (PD), displaying no relevant accumulation on reaching steady state,[4, 5] with an accumulation ratio approximating unity at all doses tested up to 30 mg twice daily.[5] Predictable PK and PD, coupled with few drug–drug and food–drug interactions, distinguishes rivaroxaban from vitamin K antagonists, allowing fixed dosing regimens without routine coagulation monitoring.[1] Rivaroxaban has a dual mode of elimination: approximately one-third of the drug is eliminated unchanged via the kidneys and two-thirds of the drug undergoes metabolic degradation in the liver, with half being excreted via the kidneys and half via the hepatobiliary route.[6] There are no active metabolites.[6] Rivaroxaban elimination occurs with terminal half-lives of 5–9 hours in young patients and 11–13 hours in elderly patients.[5, 7]

Before ROCKET AF, rivaroxaban had only been evaluated for stroke risk reduction in phase II trials involving Japanese patients with AF (NCT00779064, NCT00973323). The 20 mg once-daily (q.d.) dosing regimen chosen for ROCKET AF was based on results from two phase II dose-ranging studies[8, 9] conducted to assess rivaroxaban for treatment and secondary prevention of deep vein thrombosis (DVT). The objective of these studies was to provide a dosing regimen for the phase III EINSTEIN clinical trial program (oral direct factor Xa inhibitor rivaroxaban in patients with acute symptomatic DVT),[10, 11] which showed that, after an initial treatment phase, 20 mg q.d. was best suited for secondary prevention. Sparse PK and PD data obtained from these phase II studies were used to develop a population PK/PD model for patients with DVT.[12] This model was also employed to simulate exposure for patients with AF by adjusting for demographics reflective of a typical population of patients with AF (by creating a virtual population with increased mean age and decreased renal function).[12] The estimated plasma concentration–time profile for the simulated patients with AF was similar to that for patients with DVT receiving the same dose. Therefore, the efficacy and safety of rivaroxaban 20 mg q.d. for the secondary prevention of DVT in the phase II dose-finding studies[8, 9] and rivaroxaban pharmacology in these patients[12] supported evaluation of 20 mg q.d. in patients with AF participating in ROCKET AF. This dose selection was also supported by similarities in thrombogenesis between AF and DVT (see the Discussion Section).

In ROCKET AF, rivaroxaban exhibited a safety profile comparable with warfarin and exhibited non-inferior efficacy to warfarin for the prevention of stroke and systemic embolism.[3] Herein, we report findings from pre-specified PK/PD modeling analyses, based on matched PK and PD data obtained from a subset of patients enrolled in the trial. The models were employed to confirm previously established population PK/PD structural models in patients with AF and to determine rivaroxaban exposure and its effects on several established measures of factor Xa inhibition—factor Xa activity, prothrombin time (PT), and prothrombinase-induced clotting time (PiCT).[12]

The main objectives of this work were to verify that the previously developed structural population PK and PK/PD models[12] could also describe overall exposure to rivaroxaban and the relationship between rivaroxaban pharmacokinetics and pharmacodynamics (PT, PiCT, and factor Xa activity), estimate the models' parameters using current data from the time-matched PK/PD substudy in patients with AF, and confirm the effect of dosing modification on exposure for moderately renally impaired patients; specifically, compare 20 mg q.d. in patients with creatinine clearance (CrCl) >50 mL/min to 15 mg q.d. in patients with CrCl 30–49 mL/min.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Declaration of Conflicting Interests
  8. Funding
  9. Acknowledgments
  10. References
  11. Supporting Information

Study Design

ROCKET AF was a randomized, multicenter, event-driven, double-blind, placebo-controlled phase III trial that evaluated the efficacy and safety of rivaroxaban in patients (n = 14,264) with non-valvular AF who were at moderate to high risk of stroke.[3] The study population consisted of adult subjects with non-valvular AF who were at risk of stroke or systemic embolism. Eligible subjects were those with a prior stroke, transient ischemic attack, or systemic embolism, or who had two or more of the following risk factors: age ≥75 years, hypertension, heart failure and/or left ventricular ejection fraction ≤35%, or diabetes mellitus. The number of subjects without a prior stroke, transient ischemic attack, or systemic embolism and with only two risk factors was limited to approximately 10% of the total number of subjects enrolled. A total of 14,264 participants were randomized to treatment in ROCKET AF. Patients were randomized to receive either rivaroxaban 20 mg q.d. (15 mg q.d. in patients with moderate renal impairment—that is, calculated creatinine clearance of 30–49 mL/min at baseline according to Cockcroft–Gault equation) or dose-adjusted warfarin titrated to achieve a target international normalized ratio of 2.5 (range 2.0–3.0). Study drug was taken in the evening with food. The median duration of treatment exposure was 590 days and the median follow-up period was 707 days.[3] Pertinent national regulatory authorities and review boards/ethics committees at participating centers involved in ROCKET AF approved the study protocol.

Population PK/PD analyses were performed on a subset of patients receiving rivaroxaban in ROCKET AF who had time-matched PK and PD samples (n = 161); the resulting dataset contained 801 PK samples in total and a similar number of samples for PD evaluations (786, 742, and 799 for PT, PiCT, and factor Xa activity, respectively).

Blood Sampling

Three time-matched PK and PD blood samples were taken at steady state on a convenient day between week 2 and the end-of-study visit: a pre-dose sample was collected up to 1 hour before evening drug administration; post-dose samples were collected over intervals of 1–3 and >3–16 hours after drug administration—exact times were recorded for each patient. A subsequent PK/PD sampling was performed ≥1 month afterwards, but as close to the end-of-study visit as possible (preferably 6–12 months), while patients were still receiving daily study drug; this consisted of samples obtained at both pre-dose and 1- to 3-hour post-dose (if agreed by subjects). Study drug administration was supervised at the individual study clinics on PK/PD sampling days.

Bioanalysis

PK assays were performed by the Bioanalytical Assays Department of Bayer HealthCare AG. Rivaroxaban plasma concentrations were determined by a validated liquid chromatography–tandem mass spectrometry method using an Applied Biosystems (Foster City, CA) Sciex API 3000 or API 4000 mass spectrometer, as described previously.[13] Concentrations above the lower limit of quantification (0.50 µg/L) were determined with a precision of 3.6–8.0% and an accuracy of ≥95.6%. Concentrations below the lower limit of quantification were not included in the analysis.

PD marker bioanalysis was performed by the Duke University Medical Center Hemostasis & Thrombosis Center Core Laboratory. Factor Xa activity, PT, and PiCT were determined using methods described previously.[14, 15] In brief, factor Xa activity was determined in an ACL TOP coagulation autoanalyzer (Beckman Coulter, Fullerton, CA) using a two-step photometric assay. Total factor X (in platelet-poor plasma) was activated to factor Xa using Russell's viper venom in the presence of calcium ions. Subsequently, a chromogenic substrate (DiaPharma Factor X kit, West Chester, OH) was hydrolyzed by factor Xa, releasing p­-nitroaniline, which was quantified by spectrophotometry at 405 nm. Data were expressed as a percentage of factor Xa activity in control plasma.

PT was measured using freeze-dried rabbit brain thromboplastin (STA Neoplastin CI Plus®, Diagnostica Stago, Parsippany, NJ)—the same thromboplastin reagent used in preclinical and early phase I clinical pharmacology studies—using an ACL TOP coagulation autoanalyzer; absolute results were reported in seconds. The thromboplastin reagent (STA Neoplastin CI Plus®) had International Sensitivity Index values that varied from 1.21 to 1.25. PiCT was measured using the Pefakit® PiCT® kit (Pentapharm, Basel, Switzerland) using a one-step method15 on an ACL TOP coagulation analyzer. The absolute results were reported in seconds.

Population Pharmacokinetic Modeling

The structural population PK model was based on a previous population PK analysis performed in patients with acute symptomatic DVT, which adequately described the PK of rivaroxaban using an oral one-compartment model, parameterized in terms of apparent oral clearance (CL/F), apparent volume of distribution (V/F), and first-order absorption rate constant (Ka).[12] The structural population PK model was constructed and confirmed across several study populations in which many models were considered and evaluated.[4-7, 12, 23] In the process of constructing the original models, many models and covariates were also tested individually in NONMEM® for significant influences on PK and PK/PD. These covariates included age, body weight, body height, body surface area, lean body mass (LBM), body fat, race, sex, center identification, visit number, serum creatinine (SCr), CrCl, albumin, and concomitant medication. The previously identified covariates that influenced PK parameters of rivaroxaban were included in the population PK model: age and renal function (assessed via SCr) on CL/F, and body weight (expressed via LBM) and age on V/F. No additional covariates were included in the current study and none were tested.

Population Pharmacokinetic/Pharmacodynamic Modeling

The structural population PK/PD models were also based on a previous population analysis performed in patients with acute symptomatic DVT.[12] A conventional direct-effect linear model with a declining exponent was used for correlating rivaroxaban plasma concentration with both PT and PiCT (= Base + Slope × Cp(1 − n × Cp)), in which Base = intercept; Cp = rivaroxaban plasma concentration; n = exponent. The effect of CrCl was included in the intercept (Base) and the exponent (n) terms. Rivaroxaban plasma concentration and factor Xa activity followed an inhibitory maximum-effect (Emax) relationship, where the inhibition of factor Xa activity increases with increasing concentration and reaches asymptotic response for Emax: factor Xa (%) = E0 × (1 − Emax × Cp/(EC50 + Cp)), in which E0 = baseline; Emax = relative maximum level of inhibition; EC50 = concentration of rivaroxaban producing 50% of maximal inhibition.

Confirmation of the Models

Data from the current study were used to confirm the above PK and PK/PD models and estimate values of the models' parameters for the current AF population. Rivaroxaban concentration–time and PK/PD data were analyzed separately using non-linear mixed-effects modeling (initially using NONMEM® V Level 1.1 [University of California, San Francisco, San Francisco, CA, USA] to be consistent with the comparison of the original model results, and then using VERSION 7.1.0 [ICON Development Solutions, Ellicott City, MD] to rerun the model; no major differences were found in the parameter estimates between the two programs). The first-order conditional estimation with interaction method was used for all analyses, producing estimates of population parameters (θ), including inter-individual variability (IIV) parameters (η) of variance–covariance matrix (Ω matrix), as well as estimates of the residual variability (ε) as a part of the population parameter estimation step. Dataset preparation, exploration, and data visualization were performed using S-PLUS 6.2 for Windows (Tibco Software, Inc., Palo Alto, CA).

PK and PK/PD models were developed separately and individual models were developed for each PD endpoint. The ability to use the original population models[12] and the re-estimated model parameters to predict rivaroxaban plasma concentrations and PD biomarkers in this subset of the ROCKET AF population was assessed. The new parameter estimates for the current AF population were used to generate the empirical Bayesian (individual) predictions for all observations. Prediction errors were computed to provide a measure of bias and precision by assessing the differences between the measured and predicted individual observations. Measured rivaroxaban observations were plotted against population and individual predictions. Based on the level of variability that was found in the previous analysis,[12] the model was determined to be acceptably accurate and precise if the median PE% and the median |PE|% are ≤|20|% and 40%, respectively. A visual predictive check was also conducted to determine the predictability of the model. Otherwise, changes to the structural components of the model were to be implemented to address the observed model misspecifications.

The prediction-corrected visual predictive check (pcVPC), suggested by Bergstrand[16] as a new tool for diagnosing non-linear mixed-effect models, was used and implemented in R (www.r-project.org) with XPOSE4 (www.xpose.sf.net) from 1,000 simulations performed in NONMEM7 using PsN-toolkit software (http://psn.sourceforge.net). The XPOSE4 library (version 4.3.2) was implemented in R (version 2.12.2; R Foundation for Statistical Computing, Vienna, Austria).[17] pcVPC figures were generated using Perl-Speaks-NONMEM (PsN version 3.2.12).[18] Diagnostic plots, including standard goodness-of-fit plots and graphical assessment of the distribution of conditional weighted residuals (CWRES) were also employed to establish whether the PK and PK/PD models were adequate for describing the current dataset.

Comparison of Dosing Regimens

In ROCKET AF, renal impairment was assessed at baseline. Patients with either normal renal function (CrCl ≥ 80 mL/min) or mild renal impairment (CrCl 50–80 mL/min) received a rivaroxaban 20 mg q.d. dosing regimen, whereas a reduced 15 mg q.d. dosing regimen was used for patients with moderate renal impairment (CrCl 30–49 mL/min). The resultant PK parameters of this dose reduction in patients with moderate renal impairment were assessed using the final population PK model. Relevant patient variables, such as age, SCr, and LBM were sampled for each population (ie, with or without moderate renal impairment) from the current dataset. The rationale for dose modification according to CrCl was pre-specified to be confirmed if the population-simulated ratio of the means of maximum plasma concentration (Cmax) and area under the concentration–time curve (AUC) over 24 hours (AUC0–24) of patients with moderate renal impairment to those with normal renal function/mild renal impairment was within 0.70–1.43; this range was determined to be appropriate based on simulations using the original PK model and the current study PK sample size. Additionally, a graphical comparison of rivaroxaban exposure between patients with and those without moderate renal impairment at baseline was performed.

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Declaration of Conflicting Interests
  8. Funding
  9. Acknowledgments
  10. References
  11. Supporting Information

Patients

Matched PK and PD samples were collected in selected countries and sites. A total of 161 patients from ROCKET AF were included in the current analysis, including 25 patients with moderate renal impairment (baseline CrCl of 30–49 mL/min). A total of 801 data points were collected for the PK analyses, including 114 samples in patients with moderate renal impairment. Patient demographics are presented in Table 1. Patients with moderate renal impairment (n = 25) were, on average, 10 years older and had approximately 9 kg less LBM compared with patients with either normal renal function or mild renal impairment. In order to maximize the model's predictive accuracy and due to the available number of the PK observations, data splitting for the model validation was not carried out.

Table 1. Patient Demographics and Observations
 ROCKET AF PK/PD subset of patientsOverall ROCKET AF trial patient demographicsa
PopulationAll subjectsModerate renal impairment subjects (15 mg q.d. dose)Normal/mild renal impairment subjects (20 mg q.d. dose)All subjectsModerate renal impairment subjects (15 mg q.d. dose)Normal/mild renal impairment subjects (20 mg q.d. dose)
  1. IQR, interquartile range; PD, pharmacodynamics; PK, pharmacokinetics; q.d., once-daily; SD, standard deviation.

  2. a

    Intent-to-treat population, rivaroxaban-receiving patients only.

  3. b

    n = 7,099.

  4. c

    n = 1,471.

  5. d

    n = 7,099.

Subjects (n)161251367,1311,4795,652
Lean body mass (kg)   
Mean57.550.258.955.948.957.7
±SD9.97.39.710.58.410.3
Age (years)   
Mean65746471.278.169.4
±SD9.57.88.99.56.59.3
Baseline serum creatinine (mg/dL)   
Mean1.091.371.051.08b1.3c1.02d
±SD0.290.380.240.290.350.24
Mean CHADS2 score   
Mean3.683.683.683.483.673.42
±SD0.770.800.800.941.000.91
Median333343
IQR3–43–43–43–43–43–4

Pharmacokinetic Model

Rivaroxaban PK in this study was adequately described by an oral one-compartment model, with moderate IIV in clearance and V/F (coefficients of variation: 35% and 18%, respectively). The residual (unexplained) variability of the model was 48%.

Goodness-of-fit of the PK model was evaluated using a graphic approach (Figure 1). The prediction-corrected visual predictive check plots demonstrate that the model predicted overall concentration data in patients with AF well. The median predictive error (−9.22%) and median absolute prediction error (32.99%) were found to be within the cut-off values (±20% and 40%, respectively; a test statistic determined a priori[19] (based on previous modeling results)[12] for validity. Diagnostic plots that used CWRES, normalized prediction distribution errors, and quantile–quantile plots (Figure 1, Panels A and B) confirmed the validity of the model, as they followed the expected normal distribution.

image

Figure 1. Panel A: Distribution of normalized prediction distribution errors (NPDE) with the density of the standard normal distribution overlaid (dashed lines: median, 2.5%, and 97.5% percentiles). Panel B: Quantile–quantile plot of NPDE versus the expected standard normal distribution. Panel C: Pharmacokinetic modeling predicts rivaroxaban plasma concentrations in patients with AF. The solid red line represents the median prediction-corrected plasma concentration and the semi­transparent red field represents a simulation-based 95% confidence interval (CI) for the median. The observed 5% and 95% percentiles are presented with dashed red lines and the 95% CIs for the corresponding model-predicted percentiles are shown as semi-transparent blue fields. The observed prediction-corrected plasma concentrations are represented by blue circles. Prediction-corrected visual predictive check plots show observed (open circles) and model-predicted plasma concentrations (shaded region: 90% prediction intervals) in all patients receiving rivaroxaban. Panels D–F: The prediction-corrected visual predictive check for rivaroxaban plasma concentration–effect relationships in all patients receiving rivaroxaban for prothrombin time (PT), prothrombinase-induced clotting time (PiCT), and factor Xa activity, respectively. The solid red line represents the median prediction-corrected value and the semi­transparent red field represents a simulation-based 95% confidence interval (CI) for the median. The observed 5% and 95% percentiles are presented with dashed red lines and the 95% CIs for the corresponding model-predicted percentiles are shown as semi-transparent blue fields. The observed prediction-corrected values are represented by blue circles. Prediction-corrected visual predictive check plots show observed (open circles) and model-predicted values (shaded region: 90% prediction intervals).

Download figure to PowerPoint

Patient covariates that influenced PK in this study were age and renal function (when expressed as SCr), which affected apparent CL/F, and LBM and age, which affected apparent V/F. The effects of these factors on PK parameters are summarized in Table 2. The relationships between demographic factors on rivaroxaban PK are presented in Figure S1 of the Online Supplement. Increased SCr and age were associated with reduced rivaroxaban clearance (Supplement Figure S1, Panels A and B, respectively). Increased age was also associated with a reduced apparent V/F of rivaroxaban and increased LBM was associated with an increased apparent V/F of rivaroxaban (Supplement Figure S1, Panels C and D). Apparent CL/F and V/F were estimated to be 6.1 L/hour (IIV = 35%) and 79.7 L (IIV = 18%), respectively. The estimated model condition number was 18 (the ratio of the largest to the smallest eigenvalues of the covariance matrix), demonstrating a low degree of model ill-conditioning. The shrinkage of the variability terms, as defined by Bertrand et al,[20] provides additional information on the properties of the model fit. In the final population PK model, the shrinkage of the IIV (η) for the CL/F and V/F were 7.5% and 53.4%, respectively. The shrinkage of the IIV (ε) was 7.2%. The high shrinkage on the V/F reflects an element of uncertainty in the individual estimates—likely the result of sparse sampling used in this study (subjects with four or five samples). The total shrinkage is considered acceptable because 50% shrinkage is the threshold to potentially impact on covariates testing.[20]

Table 2. Final Parameter Estimates from the Population PK Model in Patients with Atrial Fibrillation, Original DVT Population PK Model,[10] and ACS Population PK Model Showing Typical (Population) Values (% SE)
Parameter (units)DVT (previously reported),[12]a n=ACS (previously reported)[23]aAF (re-estimated with the current dataset)Effect of parameter on PKExample of effect on PK
  1. % CV, coefficient of variation; % SE, percent standard error; CL, clearance; CL/F, oral clearance; CrCl, creatinine clearance; DVT, deep vein thrombosis; Ka, first-order absorption rate constant; LBM, lean body mass; PK, pharmacokinetics; RI, renal impairment; SCrE, effects of serum creatinine; V, volume; V/F, volume of distribution.

  2. CL/F = 7.16 × (1 – 0.00692 × (Age – 65) – 0.2690 × (SCrE – 1.05)).

  3. V/F = 68.69 × (1 – 0.00486 × (Age – 65) + 0.0082 × (LBM – 56.62)).

  4. a

    For 20 mg.

Ka (/h)1.23 (5.0)1.24 (3.3)1.16 (14.1)
CL/F (L/h)7.16 (3.7)9.1 (2.2)6.10 (3.9)
Inter-individual variability in CL/F (% CV)39.9 (7.6)31.3 (4.7)35.2 (14.3)
Age effect on CL/F−0.0069 (14.6)−0.011 (8.8)−0.011 (26.3)1.05% decrease in CL per 1 year increase from median age of 65 yearsApproximately 30% higher exposure for a 90-year-old vs. 65-year-old patient
SCrE on CL/F−0.269 (18.2)−0.151 (20.3)−0.194 (34.0)1.94% decrease in CL per 0.1 mg/dL increase from median SCrE of 1.05 mg/dLApproximately 26% higher exposure for a patient with moderate RI (CrCl ∼30 mL/min) with 2.4 mg/dL SCrE vs. patient with SCrE 1.05 mg/dL
Central V/F (L)68.7 (3.8)81.6 (1.2)79.7 (6.1) 
Inter-individual variability in V/F (% CV)28.8 (11.4)10.0 (3.6)17.6 (61.5)
LBM effect on V/F0.0082 (17.8)0.0083 (13.1)0.0118 (32.4)1.18% increase in V per 1 kg increase from median LBM of 56 kgApproximately 19% higher exposure for patient with LBM 40 kg (body weight ∼45 kg) vs. patient with LBM 56 kg
Age effect on V/F−0.00486 (20.8)−0.00707 (16.3)−0.00133 (187)0.13% decrease in CL per 1 year increase from median age of 65 yearsApproximately 30% higher exposure for a 90-year-old vs. 65-year-old patient
CV residual error (% CV)40.7 (3.2)59 (1.1)47.9 (6.2)

In general, the parameter estimates for the current AF population are similar to those reported for the DVT population used to construct the original PK model[12] (see Table 2). Several of the original structural model parameter values fall within the 95% confidence intervals of the new parameter estimates. The majority of the covariate effect parameters (eg, effect of LBM, age, and SCr) fell inside the intervals and no additional covariates were evaluated in this study. The parameter estimates for the acute coronary syndrome (ACS) population[23] are also shown in Table 2. Oral clearance (CL/F) is around 15% and 30% lower for the AF population compared with the DVT and ACS populations, respectively. On the other hand, the volume of distribution (V/F) is higher by approximately 16% for the AF population compared with the DVT population, but comparable with the ACS population. The differences in the parameter estimates between these populations are likely due to different patient demographics, dissimilar covariate distributions, and/or influences of the underlying disease state on the PK of rivaroxaban. The inter-individual variability on CL/F and V/F and residual variability terms are also similar: 35%, 18%, and 48%, respectively, for the current analysis compared with 40%, 29%, and 41% using the DVT dataset and 31%, 10%, and 59% using the ACS dataset. For the current analysis, although the age effect on V/F showed high standard error, age was kept in the model because it was significant in all models constructed previously.[4, 5, 23] The age-limited effect on V/F is believed to be attributed to a smaller number of subjects in the current AF dataset (161 patients and 801 PK samples) compared with the DVT dataset (870 patients and 4,634 PK samples) and the ACS dataset (1,347 patients and 6,644 PK samples) and not due to patient population characteristics.

Nonetheless, these results provide broad confirmation of the ability of the original model[12] to predict overall exposure to rivaroxaban and the PK parameter estimates in AF patients. Hence, the population model was used to predict the following PK parameters for rivaroxaban: Cmax, minimum plasma concentration (Cmin), and AUC0–24.

Dosing Regimens

Dose modification for patients with moderate renal impairment was confirmed, with population-simulated ratios (patients with moderate renal impairment: patients with mild renal impairment or normal renal function) of the means of Cmax and AUC0–24 of 0.88 and 0.91, respectively. These fell within the pre-specified statistic test range of 0.70–1.43 (Supplement Table S1). PK parameters for patients with moderate renal impairment (15 mg q.d.) were generally in agreement with parameters estimated for patients with either normal renal function or mild renal impairment at baseline and who received rivaroxaban 20 mg q.d. These results are also consistent with the comparable exposures predicted in simulated AF patients with and without moderate renal impairment (Figure 2). The distributions for AUC0–24 and Cmax overlapped considerably (Supplement Figure S2).

image

Figure 2. Plasma concentration–time profiles using the current model for atrial fibrillation population and three virtual patients with different age, renal function, and body weight receiving rivaroxaban 20 or 15 mg once daily. Simulated patients have mean characteristics of: age = 64 years, body weight = 89 kg, and creatinine clearance = 89 mL/min. Simulated patients have mean characteristics of: age = 74 years, body weight = 68 kg, and creatinine clearance = 42 mL/min. !Fluconazole 400 mg1.

Download figure to PowerPoint

Figure 2 also shows several examples of rivaroxaban PK, generated by employing the current PK model, in patients with characteristics (including age, renal function, and co-medications) commonly encountered by clinicians in everyday practice, including those with moderately to severely reduced CrCl.

Population Pharmacokinetic/Pharmacodynamic Model Results

The concentration–PD relationships were consistent with the previous findings and showed rapid, direct responses to rivaroxaban treatment.[12] PT and PiCT were positively correlated and factor Xa activity was negatively correlated with increasing plasma rivaroxaban concentration (Figure 1F). There was a near-linear relationship between rivaroxaban concentration and PT and PiCT values (Figure 1D and E, respectively). Goodness-of-fit graphs and statistics confirmed that for both PT and PiCT, the linear intercept, with a declining exponent on Cp reasonably described the concentration–PT or concentration–PiCT relationships, respectively. A flattening of the linear relationship was observed at higher concentrations (Table 3).

Table 3. Final Parameter Estimates for Population PK–PT and PK–PiCT Structural Models, Original DVT Population PK–PT Model,[12] and ACS PK–PT Model[23] Showing Typical (Population) Values (% SE)
Parameter [units]DVT PT model[12]ACS PT model[23]PT modelPiCT model
  1. % SE, percent standard error of the mean; Cp, rivaroxaban plasma concentration; CrCl, creatinine clearance; % CV, coefficient of variation; DVT, deep vein thrombosis; PiCT, prothrombinase-induced clotting time; PK, pharmacokinetics; PT, prothrombin time.

  2. PT = 11.40 × (1 − 0.000192 × (CrCl – 76)) + 0.0426 × (Cp(1 – 0.0000551 × (1 + 0.0174 × (CrCl – 76)) × Cp)).

  3. PiCT = 7.97 × (1 − 0.0016 × (CrCl − 76)) + 0.0954 × (Cp(1 – 0.000263 × (1 + 0.00293 × (CrCl – 76)) ×Cp)).

  4. a

    Exponent on Cp starts at 1 and declines linearly with increasing Cp (at Cp = 1,000, exponent = 0.95); accounts for flattening of linear relationship observed for high PT/Cp values.

  5. b

    Change per unit CrCl (mL/min) difference to the median (76 mL/min).

Base [s]12.5 (0.7)13.9 (0.3)11.40 (2.0)7.97 (5.2)
Inter-individual variability (% CV)9.7 (21.7)9.32 (19.4)22.6 (33.3)46.2 (24.9)
Slope [s/(µg/L)]0.036 (2.8)0.032 (1.4)0.043 (6.6)0.096 (6.2)
Slope, describing decline of exponent on Cp (n)0.000096 (7.0)0.0000593 (23.3)0.0000551 (55.0)a0.000263 (8.4)
Inter-individual variability (% CV)4.30 (13.2)6.61 (17.9)4.42 (27.1)5.56 (35.6)
CrCl on baseline−0.0004 (23.2)−0.00030 (23.1)0.000192 (169.3)b0.0016 (112.5)
CrCl on decline of exponentb0.0046 (24.4)0.0233 (35.8)0.0174 (100.6)0.00293 (55.6)
CV residual error [% CV]10.3 (8.6)7.6 (12.1)12.9 (18.4)22.1 (21.6)

The parameter values for PT and PiCT models are shown in Table 3. An effect of renal function (expressed as CrCl) was detected in the PT and PiCT models, with a moderate influence on the base and exponent parameters. At a concentration of 250 μg/L, the PT and PiCT values varied approximately 8.5% and 6.1%, respectively, for the CrCl range in our study population.

The baseline for factor Xa activity was 104% (IIV 16.61%), Emax was 107%, and EC50 was 760 μg/L (IIV 5.97%). The residual variability of the model was low at 10.05%. Age had a moderate effect on baseline parameter of the factor Xa activity model; factor Xa activity values varied approximately 30% for the age range included in our study population. In general, observed and predicted values were in good agreement for PT, PiCT, and factor Xa activity (Figure 1, Panels D, E, and F).

PD observations were generally scattered around the line of identity in observed versus predicted goodness-of-fit plots (data not shown), suggesting that the model adequately described PT, PiCT, and factor Xa parameters.

Majority of parameter estimates for the current AF population are similar to those reported for the DVT population, originally used to construct the PT model, and the ACS population (see Table 3). For the current analysis, the estimate for the influence of CrCl on PT baseline exhibits a small effect and has a higher standard error, which might be attributed to the smaller number of subjects in the current database.

In summary, the estimated mean slope (4.3 seconds in the AF patients per 100 µg/L of rivaroxaban plasma concentration) reflecting the sensitivity of this coagulation marker towards increases in rivaroxaban drug exposure is similar as seen in DVT patients (3.6 seconds per 100 µg/L),[12] and ACS patients (3.2 seconds per 100 µg/L).[23] The close-to-linear relationship between rivaroxaban drug exposure and PT response also holds true for this patient population.

The above models were used to generate predicted rivaroxaban plasma concentration–time curves and predicted PT and PiCT profiles (normalized to baseline levels; Figure 3). Notably, detectable levels of rivaroxaban were still present at 24 hours post-dose, as were detectable prolongations of both PT and PiCT.

image

Figure 3. Predicted plasma concentration (Cp)–time curves and predicted normalized prothrombin time (PT) and prothrombinase-induced clotting time (PiCT) profiles over 24 hours for a patient with normal or mildly impaired renal function of 20 mg once daily rivaroxaban.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Declaration of Conflicting Interests
  8. Funding
  9. Acknowledgments
  10. References
  11. Supporting Information

This PK/PD analysis of a subset of patients with non-valvular AF randomized to rivaroxaban in ROCKET AF demonstrated that increasing rivaroxaban plasma concentrations elevated PT and PiCT in a near-linear manner and reduced factor Xa activity. An anticoagulant effect on these parameters was observed through 24 hours following both the 15 and 20 mg doses.

Rivaroxaban PK and PK/PD have been extensively published.[4, 5, 12, 21, 22] Importantly, however, the current work represents the first characterization of rivaroxaban's pharmacology in patients with AF. Samples were obtained during the conduct of ROCKET AF, a trial in over 14,000 patients with non-valvular AF.[3] This patient population is wholly distinct from those previously studied, namely patients with acute DVT,[12] patients undergoing major orthopedic surgery,[21] and patients with acute coronary syndrome.[23] Aside from different diseases, the patient demographics in ROCKET AF were also unique.[3] The median age was 73 years; median duration of treatment exposure was 590 days; 91% of patients had hypertension; 63% had heart failure; 40% had diabetes; and 55% had a history of ischemic stroke, systemic embolism, or transient ischemic attack. In the trial, approximately 20% of patients randomized had a calculated CrCl between 30 and 49 mL/min at baseline.[24] This patient population is highly relevant to the prescribing physician and other interested healthcare providers.

Dose and Dose Regimen Selection

The 20 mg q.d. dose was chosen for ROCKET AF based on data from two phase II dose-ranging studies of rivaroxaban in patients with acute symptomatic DVT.[8, 9] These studies demonstrated no clear efficacy or safety advantage with twice-daily compared with once-daily dosing for the prevention of secondary DVT recurrence, in contrast to resolution of the index DVT.[8, 9] Furthermore, all doses showed similar efficacy and safety, including the lowest total daily dose tested, 20 mg.

For patients with AF and moderate renal impairment, current PK simulations showed that dose reduction to 15 mg q.d. would result in approximately similar drug exposure to patients with AF and normal or mildly impaired renal function receiving 20 mg q.d. For patients with moderate renal impairment at baseline (who received rivaroxaban 15 mg q.d.), the geometric mean Cmax and AUC0–24 were 229 µg/L and 3,249 µg ∙ h/L, respectively. In patients with either normal or mildly impaired renal function (who received rivaroxaban 20 mg q.d.), the geometric mean Cmax and AUC0–24 were 249 µg/L and 3,164 µg ∙ h/L, respectively. The choice of the reduced 15 mg q.d. dose for patients with moderate renal impairment was confirmed because the ratio of the means of Cmax (0.88 [0.81–0.95]) and AUC0–24 (0.91 [0.78–1.04]) were completely within the pre-specified range of 0.70–1.43 for drug exposure equivalence. The current observations corroborate this dose adaptation to 15 mg for patients with moderate renal impairment in ROCKET AF, taking into account this modest increase in rivaroxaban exposure[25] in the context of the higher bleeding risk intrinsic to renal impairment.[26]

A ROCKET AF subanalysis further confirmed the results of this PK/PD model in a clinical trial population.[24] The dose adjustment from 20 mg to 15 mg q.d. in patients with moderate renal impairment yielded efficacy and safety results that were consistent with the overall trial in comparison with dose-adjusted warfarin (P-values for interaction between renal impairment subgroups for the primary efficacy endpoint and principal safety outcome being 0.84 and 0.76, respectively). However, it should be noted that dose reduction from 20 mg to 15 mg q.d. in the ROCKET AF trial was determined by baseline CrCl. No adjustment was made for patients whose CrCl fell below 50 mL/min during the trial and it is conceivable that this accounts for some of the outlier values seen in Supplement Figure S2.

Comparison of Parameter Estimates between Study Populations

The current PK/PD analyses confirm the previous phase II population PK/PD structure models in patients with acute symptomatic DVT.[12] Population mean values of the PK and PK/PD model parameters were generally similar between the DVT, AF, and ACS populations (see Tables 2 and 3). PK of rivaroxaban was adequately described by an oral one-compartment model with IIV in both CL/F and V/F terms. Factor Xa activity correlated inversely with rivaroxaban plasma concentrations following an Emax model, and PT correlated following a simple linear relationship establishing the close correlation between PK and PD with respect to elapsed time since dosing.[5, 12, 14] Essentially PD “tracks” with PK, as might be expected for a rapid-acting small molecule direct inhibitor of a single coagulation factor. In contrast, the PD effects of the indirect-acting Vitamin K antagonists lag markedly behind their PK because time is required to downregulate the synthesis of the Vitamin K-dependent coagulation factors.

Although rivaroxaban does not require routine coagulation monitoring,[1] an assay may be useful in certain circumstances, such as overdose, planned or unplanned invasive procedures, or assessment of patient compliance.[27] Our PK/PD study shows that rivaroxaban, given at the approved doses for patients with non-valvular AF, prolongs the PT and PiCT and reduces factor Xa activity. Despite these clinically relevant observations, they may not reliably reflect the intensity of the anticoagulant effect of rivaroxaban.[27] Preliminary studies[27-30] and small field trials[31] have shown that chromogenic anti-factor Xa assays, calibrated against a standard curve produced with known amounts of rivaroxaban, may provide a suitable coagulation assay for use with rivaroxaban.

Conclusions

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Declaration of Conflicting Interests
  8. Funding
  9. Acknowledgments
  10. References
  11. Supporting Information

Similar structural population PK and PK/PD models have now been developed and confirmed in several populations (DVT, AF, and ACS[23]). The overall PK and PD similarity in many analyses across different populations suggests that rivaroxaban provides anticoagulation in all studied populations that is both consistent and predictable. Reliable estimation of rivaroxaban exposure as a result of this predictability, coupled with low inter-individual variation, should facilitate optimal management of anticoagulation.

The current analyses confirm the established structural PK and PK/PD models for rivaroxaban, the covariates affecting the PK and PK/PD of rivaroxaban, the dosing modification in patients with moderate renal impairment, and the PK/PD parameter estimates in patients with non-valvular AF. These results are reinforced by the renal impairment subgroup efficacy and safety analysis of ROCKET AF; they support use of a 20 mg q.d. rivaroxaban dose in ROCKET AF for the patients with AF and normal or mildly impaired renal function, and the reduced 15 mg q.d. dose for the patients with AF and moderate renal impairment. Because this model has been shown to be valid across different populations, its use in future trials should minimize the number of blood samples required for the prediction of PK and PD parameters.

Declaration of Conflicting Interests

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Declaration of Conflicting Interests
  8. Funding
  9. Acknowledgments
  10. References
  11. Supporting Information

Drs. Girgis, Nessel, and Peters, and Mr. Moore report employment by Janssen Pharmaceuticals Research & Development; Dr. Patel reports receiving consulting fees from Ortho-McNeil-Janssen and Bayer HealthCare and serving on an advisory board for Genzyme; Dr. Mahaffey has received consulting fees from AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Daiichi Sankyo, Eli Lilly, GlaxoSmithKline, Johnson & Johnson, Merck, Novartis, Ortho-McNeil, Pfizer, PolyMedix, and sanofi-aventis, and grant support from AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Daiichi Sankyo, Eli Lilly, GlaxoSmithKline, Johnson & Johnson, Merck, Novartis, Ortho-McNeil, Pfizer, PolyMedix, Pozen, Regado Biotechnologies, and The Medicines Company; Dr. Halperin has received consulting fees from Astellas Pharma, Boehringer Ingelheim, Bristol-Myers Squibb, Daiichi Sankyo, Pfizer, and sanofi-aventis; Dr. Califf reports receiving consulting fees from Kowa, Nile, Orexigen, sanofi-aventis, Novartis, and Xoma, and grant support from Novartis, Merck, and Amylin/Lilly, and having an equity interest in Nitrox; Dr. Fox reports receiving grant support/lecture fees from Eli Lilly, and lecture fees from sanofi-aventis and AstraZeneca; Dr. Becker has received grant support from Johnson & Johnson and AstraZeneca, and consulting fees from Boehringer Ingelheim and Daiichi Sankyo. No other potential conflict of interest relevant to this article was reported.

Funding

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Declaration of Conflicting Interests
  8. Funding
  9. Acknowledgments
  10. References
  11. Supporting Information

The ROCKET AF study was funded by Janssen Research and Development, LLC and Bayer HealthCare Pharmaceuticals.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Declaration of Conflicting Interests
  8. Funding
  9. Acknowledgments
  10. References
  11. Supporting Information

The authors thank Thomas Ortel MD, PhD, at the Duke Comprehensive Center for Hemostasis and Thrombosis, for performing technical aspects of the PD measures, and Abigail Macleod, who provided editorial assistance with funding from Bayer HealthCare Pharmaceuticals AG and Janssen Scientific Affairs, LLC.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Declaration of Conflicting Interests
  8. Funding
  9. Acknowledgments
  10. References
  11. Supporting Information
  • 1
    Bayer Pharma AG. Xarelto® (rivaroxaban) Summary of Product Characteristics. 2013. Available at: http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000944/WC500057108.pdf. Accessed March 1, 2013
  • 2
    Janssen Pharmaceuticals, Inc. Xarelto® (rivaroxaban) Prescribing Information. Available at: http://www.accessdata.fda.gov/drugsatfda_docs/label/2013/022406s004lbl.pdf. Accessed July 23, 2013.
  • 3
    Patel MR, Mahaffey KW, Garg J, et al. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med. 2011; 365:883891.
  • 4
    Mueck W, Becka M, Kubitza D, Voith B, Zuehlsdorf M. Population model of the pharmacokinetics and pharmacodynamics of rivaroxaban—an oral, direct Factor Xa inhibitor—in healthy subjects. Int J Clin Pharmacol Ther. 2007; 45:335344.
  • 5
    Kubitza D, Becka M, Wensing G, Voith B, Zuehlsdorf M. Safety pharmacodynamics, and pharmacokinetics of BAY 59-7939—an oral, direct Factor Xa inhibitor—after multiple dosing in healthy male subjects. Eur J Clin Pharmacol. 2005; 61:873880.
  • 6
    Weinz C, Schwarz T, Kubitza D, Mueck W, Lang D. Metabolism and excretion of rivaroxaban, an oral, direct Factor Xa inhibitor, in rats, dogs and humans. Drug Metab Dispos. 2009; 37:10561064.
  • 7
    Kubitza D, Becka M, Roth A, Mueck W. Dose-escalation study of the pharmacokinetics and pharmacodynamics of rivaroxaban in healthy elderly subjects. Curr Med Res Opin. 2008; 24:27572765.
  • 8
    Agnelli G, Gallus A, Goldhaber SZ, et al. Treatment of proximal deep-vein thrombosis with the oral direct Factor Xa inhibitor rivaroxaban (BAY 59-7939): the ODIXa-DVT (oral direct Factor Xa inhibitor BAY 59-7939 in patients with acute symptomatic deep-vein thrombosis) study. Circulation. 2007; 116:180187.
  • 9
    Buller HR, Lensing AW, Prins MH, et al. A dose-ranging study evaluating once-daily oral administration of the Factor Xa inhibitor rivaroxaban in the treatment of patients with acute symptomatic deep vein thrombosis: the Einstein-DVT Dose-Ranging Study. Blood. 2008; 112:22422247.
  • 10
    The EINSTEIN–PE Investigators. Oral rivaroxaban for the treatment of symptomatic pulmonary embolism. N Engl J Med. 2012; 366:12871297.
  • 11
    The EINSTEIN Investigators. Oral rivaroxaban for symptomatic venous thromboembolism. N Engl J Med. 2010; 363:24992510.
  • 12
    Mueck W, Lensing AW, Agnelli G, Decousus H, Prandoni P, Misselwitz F. Rivaroxaban: population pharmacokinetic analyses in patients treated for acute deep-vein thrombosis and exposure simulations in patients with atrial fibrillation treated for stroke prevention. Clin Pharmacokinet. 2011; 50:675686.
  • 13
    Kubitza D, Becka M, Zuehlsdorf M, Mueck W. Body weight has limited influence on the safety, tolerability, pharmacokinetics, or pharmacodynamics of rivaroxaban (BAY 59-7939) in healthy subjects. J Clin Pharmacol. 2007; 47:218226.
  • 14
    Kubitza D, Becka M, Voith B, Zuehlsdorf M, Wensing G. Safety pharmacodynamics, and pharmacokinetics of single doses of BAY 59-7939, an oral, direct Factor Xa inhibitor. Clin Pharmacol Ther. 2005; 78:412421.
  • 15
    Samama MM, Martinoli JL, Le Flem L, et al. Assessment of laboratory assays to measure rivaroxaban—an oral, direct Factor Xa inhibitor. Thromb Haemost. 2010; 103:815825.
  • 16
    Bergstrand M, Hooker AC, Wallin JE, Karlsson MO. Prediction-corrected visual predictive checks for diagnosing nonlinear mixed-effects models. AAPS J. 2011; 13:143151.
  • 17
    Jonsson EN, Karlsson MO. Xpose—an S-PLUS based population pharmacokinetic/pharmacodynamic model building aid for NONMEM. Comput Methods Programs Biomed. 1999; 58:5164.
  • 18
    Lindbom L, Pihlgren P, Jonsson EN. PsN-Toolkit—a collection of computer intensive statistical methods for non-linear mixed effect modeling using NONMEM. Comput Methods Programs Biomed. 2005; 79:241257.
  • 19
    Sheiner LB, Beal SL. Some suggestions for measuring predictive performance. J Pharmacokinet Biopharm. 1981; 9:503512.
  • 20
    Bertrand J, Comets E, Laffont CM, Chenel M, Mentré F. Pharmacogenetics and population pharmacokinetics: impact of the design on three tests using the SAEM algorithm. J Pharmacokinet Pharmacodyn. 2009; 36:317339.
  • 21
    Mueck W, Eriksson BI, Bauer KA, et al. Population pharmacokinetics and pharmacodynamics of rivaroxaban —an oral, direct Factor Xa inhibitor—in patients undergoing major orthopaedic surgery. Clin Pharmacokinet. 2008; 47:203216.
  • 22
    Mueck W, Borris LC, Dahl OE, et al. Population pharmacokinetics and pharmacodynamics of once- and twice-daily rivaroxaban for the prevention of venous thromboembolism in patients undergoing total hip replacement. Thromb Haemost. 2008; 100:453461.
  • 23
    Xu XS, Moore K, Burton P, et al. Population pharmacokinetics and pharmacodynamics of rivaroxaban in patients with acute coronary syndromes. Br J Clin Pharmacol. 2012; 74:8697.
  • 24
    Fox KAA, Piccini JP, Wojdyla D, et al. Prevention of stroke and systemic embolism with rivaroxaban compared with warfarin in patients with non-valvular atrial fibrillation and moderate renal impairment. Eur Heart J. 2011; 32:23872394.
  • 25
    Kubitza D, Becka M, Mueck W, et al. Effects of renal impairment on the pharmacokinetics, pharmacodynamics and safety of rivaroxaban, an oral, direct Factor Xa inhibitor. Br J Clin Pharmacol. 2010; 70:703712.
  • 26
    DiMarco JP, Flaker G, Waldo AL, et al. Factors affecting bleeding risk during anticoagulant therapy in patients with atrial fibrillation: observations from the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) study. Am Heart J. 2005; 149:650656.
  • 27
    Lindhoff-Last E, Samama MM, Ortel TL, Weitz JI, Spiro TE. Assays for measuring rivaroxaban: their suitability and limitations. Ther Drug Monit. 2010; 32:673679.
  • 28
    Karst A, Bakowski-Enzian B, Perzborn E. Monitoring of rivaroxaban: suitability of a well-established chromogenic anti-Factor Xa assay. J Thromb Haemost. 2010; 7(suppl 2):372. Abstract PP-MO-162.
  • 29
    Samama MM, Le Flem L, Guinet C, Perzborn E, Amiral J, Depasse F. Suitability of chromogenic anti-FXa methods to measure rivaroxaban in human plasma. J Thromb Haemost. 2009; 7(suppl 2):693. Abstract PP-WE-199.
  • 30
    Samama M, Amiral J, Guinet C, Perzborn E, Depasse F. Specific and rapid measurement of rivaroxaban in plasma using a new, dedicated chromogenic assay. Hämostaseologie. 2010; 30:A37. Abstract P01–P17.
  • 31
    Samama MM, Contant G, Spiro TE, et al. Evaluation of the anti-Factor Xa chromogenic assay for measuring rivaroxaban plasma concentrations using calibrators and controls. Blood (ASH Annual Meeting Abstracts). 2010; 116:13631364. Abstract 3330.

Supporting Information

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Declaration of Conflicting Interests
  8. Funding
  9. Acknowledgments
  10. References
  11. Supporting Information

Additional supporting information may be found in the online version of this article at the publisher's website.

FilenameFormatSizeDescription
jcph288-jcph288-sm-0001-SuppData-S1.docx50KSupplementary Figure Legends and References
jcph288-jcph288-sm-0002-SuppFig-S1.tif6124K

Figure S1. Relationships between demographic factors on rivaroxaban pharmacokinetics in patients with non-valvular atrial fibrillation, based on individual post hoc estimates from the current pharmacokinetic model.1 The effects of serum creatinine (SCr) and age on rivaroxaban clearance are shown in (A) and (B), respectively; the effects of lean body mass (LBM) and age on rivaroxaban volume of distribution (V/F) are shown in (C) and (D), respectively. Filled circles represent observed values; straight lines represent trend lines (evaluated using LOWESS).

jcph288-jcph288-sm-0003-SuppFig-S2.tif4622K

Figure S2. Box-whisker plots showing estimated levels of rivaroxaban exposure in patients with normal (20 mg) and moderately impaired renal function (CrCl 30–49 mL/min; 15 mg). Shown are plots for Cmax (A) and AUC0–24 (B). Top and bottom of the shaded box correspond to the upper (75%) and lower (25%) percentiles, respectively. The middle white band represents the median; whiskers correspond to the lowest and highest data points within 1.5× interquartile range, calculated from the center of the data. Points beyond this range are possible outliers, indicated by horizontal lines. AUC0–24, area under the concentration versus time curve from 0 to 24 hours; Cmax, maximum plasma concentration; CrCl, creatinine clearance.

jcph288-jcph288-sm-0004-SuppTab-S1.docx48K

Table S1. Rivaroxaban Exposure Individual Parameter Estimates at Steady State: Cmax and AUC0­24 hours

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.