Comparison of the pharmacokinetics and pharmacodynamics of apixaban and rivaroxaban in dogs

Abstract

Background: Comparative pharmacokinetics and pharmacodynamics (PK/PD) of apixaban and rivaroxaban have not been studied in dogs and the propensity of these drugs to cause hypercoagulability after discontinuation is unknown.

Hypothesis: Compare the PK/PD of clinical dosing regimens of PO apixaban and rivaroxaban administered repeatedly to healthy dogs and assess the effect of abrupt drug discontinuation on coagulation.

Animals: Six University-owned, purpose-bred, middle-aged, mixed-breed dogs (4 male, 2 female).

Methods: Dogs were given apixaban or rivaroxaban PO at 0.5 mg/kg q12h for 7 days with a 14-day washout period between drugs. Plasma drug concentrations were quantitated, and anticoagulant effects were measured using clotting times, calibrated anti-Xa bioactivity assays, and measurements of thrombin generation. The potential for rebound hypercoagulability was assessed by measuring D-dimers, thrombin-antithrombin (TAT) complexes, and antithrombin activity after drug discontinuation.

Results: Plasma drug concentrations and anti-Xa bioactivities were closely correlated for both drugs, but drug concentrations varied considerably among dogs, despite consistent dose regimens. Thrombin generation variables were significantly correlated with the anti-Xa bioactivity of both drugs and no significant differences in the effects of apixaban and rivaroxaban on thrombin generation were observed. Drug discontinuation had no effect on D-dimer concentrations. The concentration of TAT complexes decreased after apixaban discontinuation and did not change after rivaroxaban discontinuation.

Conclusions and clinical importance: Repeated PO administration of apixaban or rivaroxaban to healthy dogs produced comparable anticoagulant effects measured by inhibition of thrombin formation. Rebound hypercoagulability after drug discontinuation was not observed and weaning of these drugs in clinical patients might not be necessary.

Keywords: DOAC; anticoagulant; anti‐Xa; direct oral anticoagulant; thrombin generation; thromboprophylaxis; thrombosis.

FIGURE 1

The effects of repeated PO apixaban administration on anti‐Xa bioactivity in healthy dogs. Dogs received 0.5 mg/kg q12h for 7 days. (A) Average plasma apixaban concentrations measured by liquid chromatography‐mass spectrometry for the 6 dogs plotted over time. Dots represent the median and error bars represent the IQR. (B) Plasma apixaban concentrations over time for individual dogs. Each symbol represents the measured concentration for each dog at each time point. (C) Average anti‐Xa bioactivity was measured using an apixaban‐calibrated assay for the 6 dogs plotted over time. Dots represent the median and error bars represent IQR. (D) Anti‐Xa bioactivity over time for individual dogs. Each symbol represents the measured anti‐Xa bioactivity, expressed as ng/mL apixaban equivalent for each dog at each time point. (E) Box and Whisker plots of the anti‐Xa data are displayed in panel C where the middle line represents the median, the Box represents the IQR, and the Whiskers represent the minimum and maximum values. Asterisks denote significant differences between the marked timepoint and baseline (Day 0), as follows *<.05, **<.01, ***<.001. (F) Scatterplot of plasma apixaban concentrations (X‐axis) versus anti‐Xa bioactivity (Y‐axis). Nonlinear regression analysis was used to model the relationship between plasma apixaban concentration and the resulting anti‐Xa bioactivity.

FIGURE 2

The effects of repeated PO apixaban administration on clotting time assays in healthy dogs over time. Dogs received 0.5 mg/kg q12h for 7 days. For both clotting time assays, data are displayed as dots plots where dots represent the median and error bars represent the IQR, and as Box and Whisker plots where the middle line represents the median, the box represents the IQR, and the Whiskers represent the minimum, and maximum values. The part figures (A, B) present data for the prothrombin time (PT); (C, D) represent data for the activated partial thromboplastin time (aPTT). For all the panels asterisks denote significant differences between the marked timepoint and baseline (Day 0), as follows *<.05, **<.01, ***<.001. Horizontal dotted lines represent the upper and lower bounds of the laboratory reference intervals.

FIGURE 3

The effects of repeated PO apixaban administration on thrombin generation variables in healthy dogs over time. Dogs received 0.5 mg/kg q12h for 7 days. For all panels, average values for the 6 dogs are plotted over time. Dots represent the median and error bars represent the IQR. (A) The lag time from assay initiation to the onset of thrombin formation; (B) time to peak thrombin formation; (C) peak thrombin formation; (D) endogenous thrombin potential (ETP) derived from the area under the curve and representing the total amount of thrombin generated. For all the panels asterisks denote significant differences between the marked timepoint and baseline (Day 0), as follows *<.05, **<.01, ***<.001.

FIGURE 4

The effects of repeated PO rivaroxaban administration on anti‐Xa bioactivity in healthy dogs. Dogs received 0.5 mg/kg q12h for 7 days. (A) Average plasma rivaroxaban concentrations measured by liquid chromatography‐mass spectrometry for the 6 dogs plotted over time. Dots represent the median and error bars represent the IQR. (B) Plasma rivaroxaban concentrations over time for individual dogs. Each symbol represents the measured concentration for each dog at each time point. (C) Average anti‐Xa bioactivity was measured using a rivaroxaban‐calibrated assay for the 6 dogs plotted over time. Dots represent the median and error bars represent the IQR. (D) Anti‐Xa bioactivity over time for individual dogs. Each symbol represents the measured anti‐Xa bioactivity, expressed as ng/mL rivaroxaban equivalent for each dog at each time point. (E) Box and Whisker plots of the anti‐Xa data are displayed in panel C where the middle line represents the median, the box represents the IQR, and the Whiskers represent the minimum and maximum values. Asterisks denote significant differences between the marked timepoint and baseline (Day 0), as follows *<.05, **<.01, ***<.001. (F) Scatterplot of plasma rivaroxaban concentrations (X‐axis) versus anti‐Xa bioactivity (Y‐axis). Linear regression analysis was used to model the relationship between plasma rivaroxaban concentration and the resulting anti‐Xa bioactivity.

FIGURE 5

The effects of repeated PO rivaroxaban administration on clotting time assays in healthy dogs over time. Dogs received 0.5 mg/kg q12h for 7 days. For both clotting time assays, data are displayed as dot plots where dots represent the median and error bars represent the IQR and as Box and Whisker plots where the middle line represents the median, the Box represents the IQR, and the Whiskers represent the minimum and maximum values. The part figures (A, B) present data for the prothrombin time (PT); (C, D) present data for the activated partial thromboplastin time (aPTT). For all the panels asterisks denote significant differences between the marked timepoint and baseline (Day 0), as follows *<.05, **<.01, ***<.001. Horizontal dotted lines represent the upper and lower bounds of the laboratory reference intervals.

FIGURE 6

The effects of repeated PO rivaroxaban administration on thrombin generation variables in healthy dogs over time. Dogs received 0.5 mg/kg q12h for 7 days. For all panels, average values for the 6 dogs are plotted over time. Dots represent the median and error bars represent IQR. (A) The lag time from assay initiation to the onset of thrombin formation; (B) time to peak thrombin formation; (C) peak thrombin formation; (D) endogenous thrombin potential (ETP) derived from the area under the curve and represent the total amount of thrombin generated. For all panels asterisks denote significant differences between the marked timepoint and baseline (Day 0), as follows *<.05, **<.01, ***<.001.