Dabigatran and Rivaroxaban have both been approved in Canada for the prevention and treatment of thrombosis. Although they have different mechanisms of action; they both share the same advantage over warfarin that is their predictable anticoagulation effect, which precludes the need for routine monitoring. However, there are special situations (emergent surgery, bleeding, therapeutic failure) where dosing would be clinically relevant. A major concern is that patients treated with these drugs may present themselves in non-tertiary centers where the availability of specialized coagulation assays is usually limited, especially in an emergency setting. So, although specific monitoring assays are under development for dabigatran and rivaroxaban, we sought to investigate the widely available coagulation assays prothrombin time (PT), international normalized ratio (INR), activated partial thromboplastin time (aPTT) and thrombin time (TT) as means to estimate plasma levels of these new anticoagulants.
Our study aimed at emulating real life situations. We purposely recruited academic (15) and non-academic centers (26), to reflect the diversity of the laboratories in the province of Quebec, Canada.
Solutions containing increasing concentrations of dabigatran (from 0,01 μmol/mL to 0,48 μmol/mL) and rivaroxaban (from 0,012 μmol/mL to 0,490 μmol/mL) were prepared. Frozen aliquots of each solution were shipped to each of the 41 participating center (blinded for concentration). Each laboratory then performed automated assays for PT, aPTT and TT, according to each local protocol.
The TT was exquisitely sensitive to dabigatran but was not available in every hospital (available in 25 out of 41). All values were three times above the normal value even for low concentrations (> 0,03μg/mL). With regards to aPTT, we were able to demonstrate a difference in sensitivity of different cephalins for low concentrations of dabigatran. The value was abnormal in 74% centers for a concentration of 0,01 μg/mL. Whereas no abnormal value was obtained with APTT-SP, all values were above normal for CK-PREST. As for high concentrations, normal values were obtained in 5% of centers for a concentration of 0,06 μg/mL. An aPTT value over 80 was obtained for 25% of centers testing the 0,26 μg/mL concentration and of 68% centers with the 0,48 μg/mL samples. We thus have shown that a normal aPTT could be obtained, although rarely, with concentrations that are within or above the therapeutic range. Accordingly with what has already been published, our study demonstrated that INR was less sensitive to dabigatran than aPTT.
The TT was not significantly prolonged with therapeutic concentrations of rivaroxaban (at the highest concentrations we observed a slight prolongation with a ratio on normal value < 1,5). INR prolongation with rivaroxaban was linear and we observed a variation in sensitivity with the 5 studied thromboplastins. Neoplastin C1 and Neoplastin C1 plus showed the greatest sensitivity while Thromborel S showed the least. We were able to demonstrate a statically significant difference between high ISI (>1,15) and low ISI (<1,15) thromboplastins, the latter being less sensitive. In our study, aPTT showed a slightly greater sensitivity to rivaroxaban than INR. Both PT and aPTT were systematically abnormal for concentrations above therapeutic range. At concentrations nearing Cthrough however, a significant proportion of the centers obtained normal values, which demonstrates the lack of sensitivity of both assays for detecting a residual effect of medication.
Dabigatran and Rivaroxaban represent a new challenge for hemostasis laboratories. Our study reflects the diversity of coagulation assays used throughout the province of Quebec. The variation in sensitivity that was observed between laboratories makes it difficult to provide broad recommendations on the use of common coagulation assays as a dosing method for dabigatran or rivaroxaban. We suggest that each center provide recommandations based on a calibrator generated dose response curve before using these assays in a clinical setting.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.