Reversal of warfarin era thinking.

Abstract

Anticoagulation offers benefits in appropriately selected patients; however, all anticoagulants increase the risk of bleeding, which can be, at times, life threatening. For decades, vitamin K antagonists were the only oral anticoagulants available to physicians and their patients. In recent years, we have witnessed the approval of five direct oral anticoagulants (DOACs), which have shifted prior paradigms in anticoagulation such as omitting the need for routine therapeutic monitoring and lessening concerns about food and drug interactions. Now, as two new targeted DOAC reversal agents work their way through the US Food and Drug Administration (one approved, and the other nearing approval), it is important to revisit the philosophy of anticoagulation reversal. We believe that reversal agents for novel anticoagulants rely on data that is insufficient to determine their clinical safety and efficacy compared to available alternatives. Historically, warfarin-associated bleeding was reversed with fresh frozen plasma (FFP) and vitamin K. Warfarin's effects are long-lasting, with a plasma half-life of 36 h, and a much longer biological half-life. Despite the traditional use of FFP in reversing warfarin, randomized studies have shown that quick normalization of the INR (in <3 h) occurs in less than 10% of patients and that on average, it takes 24 h for normalization to occur 1. In the last few years, prothrombin complex concentrate (PCC) has been approved for warfarin reversal after randomized control trials showed PCC was superior to FFP in terms of achieving rapid reversal, favourable surgical outcomes and better outcomes in CNS bleeding events 1, 2. Professional guidelines currently recommend PCC reversal for severe, warfarin-associated bleeding 3. Currently, many providers give PCC attempting to reverse severe DOAC-associated bleeds. However, this off-label use has not been prospectively trialed in humans yet. Data from animal models suggest that PCC improves markers of coagulation and decreases bleeding time in mice and rabbits, including models of intracranial haemorrhage 4. Data on reversing DOACs with PCC in humans are limited to in vitro studies suggesting restoration of thrombin generation and uncontrolled retrospective reports suggesting efficacy 4. Despite these limitations, we believe it is reasonable to use PCCs for severe DOAC-associated bleeds given the high mortality associated with intracranial haemorrhage. Recently, there have been promising trials on two agents designed specifically to reverse DOAC-associated bleeding. The first agent, idarucizumab, is a monoclonal antibody against dabigatran, which was shown to reverse markers of anticoagulation in patients on dabigatran. In the trial which led to idarucizumab approval, 90 patients with severe dabigatran-associated bleeding received idarucizumab, and of those with abnormal markers of coagulation, 88–98% of the patients had immediate normalization of these markers. Despite this, the average time to cessation of bleeding was 11.4 h, and 20% of patients died 5. The second agent, andexanet alfa, is a nonactive, molecular mimic of factor Xa, which transiently binds factor Xa inhibitors (like rivaroxaban and apixaban) and reverses markers of anticoagulation. The preliminary results of an ongoing cohort study (ANNEXA-4) were recently published on 67 patients with acute major bleeding events associated with factor Xa inhibitors. The patients were treated with a bolus of andexanet alfa followed by a 2-h infusion. Amongst the patients with abnormal markers of coagulation at enrolment, levels of anti-Xa activity decreased by 88–93%. After 12 h, 65% of patients were adjudicated as having achieved ‘excellent’ haemostasis, and 12.7% achieved ‘good’ haemostasis. Of note, 15% of patients in the study died and 18% had thrombotic complications 6. Idarucizumab is currently FDA approved for the reversal of dabigatran, and andexanet alfa is likely to be approved for the reversal of rivaroxaban and apixaban soon. Making DOAC reversal agents widespread and available seems to be a logical next step; however, there are unique properties of DOACs and the nature of reversal that must be taken into account. First, DOACs have short half-lives (typically ranging from 7 to 17 h depending on the drug) which is shorter than the average time it takes to reverse warfarin with FFP (~24 h) 1. Whilst warfarin reversal seems like a reasonable intervention given the long half-life of the drug, it must be noted that holding a DOAC at presentation may reverse anticoagulant effects as quickly or quicker than FFP reverses warfarin. Whether use of the DOAC reversal agents is better than simply waiting for the DOAC to metabolize is unknown. Secondly, most patients with DOAC-associated bleeding are close to or have surpassed one drug elimination half-life by the time reversal is administered. The average time from last use of an anticoagulant to time of administration of a reversal agent is approximately 12 h for apixaban and rivaroxaban, and approximately 15 h for dabigatran 5, 6. This likely explains why nearly a quarter of patients in both of the aforementioned studies who received reversal did not have detectable drug levels in their systems (26% and 22%, respectively) 5, 6. Long delays between anticoagulant use and bleeding episode minimize the need for and efficacy of reversal agents. Thirdly, direct evidence that DOAC reversal results in clinically meaningful benefits is lacking. Unlike with warfarin, where randomized trials have shown a benefit with PCC, no controlled studies of DOAC reversal have been reported. Both of the aforementioned DOAC reversal agents have been trialed in single-arm studies, where markers of coagulation were quickly normalized; despite this, it took many hours for clinical haemostasis to be achieved 5, 6. No data currently exist on the average time to haemostasis or other relevant clinical outcomes with simply holding DOACs. This point is particularly important for andexanet alfa, which has an extremely short half-life. The drug quickly neutralizes the effects of factor Xa inhibitors; however, after the infusion, anticoagulation parameters quickly return to the normal decay. Whilst the ability to immediately turn off anticoagulation effects for a few hours is novel and whilst preclinical trial data would suggest that the 2-h infusion of andexanet alfa is sufficient to restore thrombin generation and form a haemostatic plug, there are no controlled data that shows this is sufficient to provide a clinically meaningful benefit in terms of cessation of bleeding, or improved mortality or morbidity 6. Lastly, in pooled analyses comparing major bleeding events in patients on DOACs compared to their warfarin counterparts, there have been no worse outcomes in patients on DOACs compared to warfarin, despite no dedicated DOAC reversal agents being available at the time of these trials. In fact, in some cases, the patients who bled on DOACs have even done better 7. We have entered a new era of anticoagulation, and the current paradigm is shifting. DOACs have many benefits when compared to warfarin, the most relevant of which is improved safety. Consistently, DOACs have been shown to cause less major bleeding than warfarin. Perhaps this is due to shorter half-lives, and the fact that nearly a quarter of patients will have anticoagulation levels below the threshold of detection at presentation. These facts, as well as costs and risks associated with reversal, must be considered when evaluating patients for DOAC reversal 5, 6. Certainly, there are situations in which the benefits of reversal may outweigh the risks, such as in patients with severe trauma or CNS bleeding with recent ingestion of a DOAC, but these patients are not likely to represent the majority of subjects considered for reversal. Point of care testing for DOAC drug levels is sorely lacking, but would be a good first step in preventing unnecessary reversal in patients with undetectable drug levels. Additionally, in patients with detectable drug levels, controlled trials are needed to determine if there is significant benefit to DOAC reversal. Finally, given the high costs of the new reversal agents, it would be proper to compare the efficacy of these novel agents with older ones like prothrombin complex concentrates. No conflict of interests to declare.