The direct oral anticoagulants (DOACs) are the treatment of choice for most patients with atrial fibrillation and/or non-cancer associated venous thromboembolic disease. While routine monitoring of these agents is not required, assessment of anticoagulant effect may be desirable in special situations. The objective of this study was to systematically review and summarize current evidence regarding laboratory assessment of the anticoagulant effects of dabigatran, rivaroxaban, apixaban and edoxaban.
Pubmed, Embase, and Web of Science were searched for studies reporting a relationship between drug levels and coagulation assay results.
One hundred and twelve eligible studies were identified, including 35 for dabigatran, 50 for rivaroxaban, nine for apixaban and 13 for edoxaban. The performance of standard anticoagulation tests, including APTT and PT/INR, varied across both DOACs as well as reagents. For most assays, standard anticoagulation tests showed insufficient correlation to provide a reliable assessment of DOAC effects, including exclusion of clinically relevant drug levels. Dilute thrombin time assays demonstrated a high degree of linear correlation (r2 = 0.67-0.99) across a range of expected concentrations of dabigatran, as did ecarin-based assays. Calibrated anti-Xa assays demonstrated a high degree of linear correlation (r2 = 0.78-1.0) across a wide range of drug concentrations for rivaroxaban, apixaban and edoxaban.
Unfortunately, an ideal test, offering both accuracy and precision for measurement of any DOAC is not widely available. We recommend the use of a dilute thrombin time assay or ecarin-based assay for assessment of the anticoagulant effect of dabigatran and anti-Xa assays with drug-specific calibrators for assessment of the anticoagulant effects of direct Xa inhibitors. In the absence of these tests, thrombin time or APTT are recommended over PT/INR for assessment of dabigatran and PT/INR is recommended over APTT for detection of factor Xa inhibitors. Time since last dose, the presence or absence of drug interactions, renal and hepatic function should impact clinical estimates of anticoagulant effect in a patient for whom laboratory test results are not available.