Background: The direct, oral, factor Xa inhibitor, rivaroxaban, is increasingly used to provide effective anticoagulation in atrial fibrillation and venous thromboembolism. Whilst rivaroxaban does not require therapeutic monitoring there are situations when it is useful to estimate the anticoagulant effect of the drug such as during bleeding episodes or before emergency surgery. It has previously been shown that conventional coagulation tests can provide a crude estimation of anticoagulant effect of rivaroxaban if a sensitive reagent is used. In this study we explore whether rotational thromboelastometry (ROTEM) would provide a more accurate measure of rivaroxaban effect.

Methods: Peak serum rivaroxaban levels were taken 3-5 hours post-dose in 121 consecutive patients established on a once-daily anticoagulation regime with rivaroxaban. Conventional coagulation tests, prothrombin time (PT) and activated partial thromboplastin time (APTT) were performed alongside rivaroxaban level and rotational thromboelastometry (ROTEM, TEM Ltd, Germany) on native citrated whole blood. PT and APTT used HemosIL Recombiplastin 2G and SynthasIL reagents (Instrumentation Laboratory (IL), USA) respectively. Rivaroxaban levels were measured using HemosIL Liquid Anti Xa kit (IL, USA) with rivaroxaban calibrators from Hyphen Biomed, France on an ACL TOP 700 coagulometer (IL, USA)). Demographic and biochemical data was collected on each patient. Results were analysed to determine if ROTEM can be used to assess the anticoagulant effect of rivaroxaban in real-world patients with different demographics and organ function.

Results: Significant positive correlation was seen between rivaroxaban level and prothrombin time (PT) (R=0.796, Pearson's correlation coefficient). Weaker correlation was observed between rivaroxaban level and activated partial thromboplastin time (APTT) (R=0.425). There was modest positive correlation between the clotting time (CT) parameter using ROTEM and rivaroxaban level (R=0.328). However, when grouped into low (<200ng/ml), intermediate (200-300ng/ml) and high (>300ng/ml) rivaroxaban levels, the CTs show no meaningful association and therefore cannot be used as a surrogate marker to predict anticoagulant effect. There is no significant difference between the mean rivaroxaban levels for patients on 15mg rivaroxaban, those on 20mg with creatinine clearance <60ml/min and those on 20mg with creatinine clearance >60ml/min (Analysis of Variance, n=121, F=2.009, P=0.159), suggesting that with dose adjustment a similar anticoagulant effect is achieved in patients with different renal function.

Conclusion: Our data suggests that the correlation between rivaroxaban levels and ROTEM CT parameter is not sufficiently strong to reliably predict the anticoagulant effect of rivaroxaban and does not confer any advantage over conventional clotting tests.

Disclosures

Chowdary:Baxalta: Honoraria, Membership on an entity's Board of Directors or advisory committees; Biogen: Honoraria, Membership on an entity's Board of Directors or advisory committees; CSL Behring: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Sobi: Honoraria; Bayer: Honoraria; Pfizer: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Novo Nordisk: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding. Drebes:Bayer: Consultancy; Bayer: Other: Educational Grant.

Topics:
anticoagulants, rivaroxaban, activated partial thromboplastin time measurement, anticoagulation, blood coagulation tests, creatinine clearance, prothrombin time, antithrombin iii, atrial fibrillation, emergency surgical procedure

Author notes

*

Asterisk with author names denotes non-ASH members.

© 2016 by The American Society of Hematology
2016
Sign in via your Institution