Four direct oral anticoagulants (DOACs) have been approved by the Food and Drug Administration (FDA) since October 2010: one direct thrombin inhibitor, dabigatran (Pradaxa, Boehringer Ingelheim Pharmaceuticals); and three direct factor Xa (FXa) inhibitors, rivaroxaban (Xarelto, Janssen Pharmaceuticals), apixaban (Eliquis, Bristol-Myers Squibb), and edoxaban (Savaysa, Daiichi Sankyo).1–4 As with any anticoagulant, bleeding is the most concerning adverse event that can occur with DOAC use. Although there is no established reversal agent for the DOACs, bleeding-related outcomes in clinical trials of DOACs have been favorable. Meta-analyses have shown that compared with warfarin therapy in patients with atrial fibrillation, DOAC agents lowered rates of intracranial hemorrhage, major bleeding, major bleeding–related mortality, and all-cause mortality.5,6 In addition, the case fatality rate for DOAC-associated major bleeding was shown to be lower than the rate with warfarin use in the initial phase of venous thromboembolism treatment.7 Observational studies have been undertaken to confirm the safety of DOACs in routine practice, where their use occurs outside the domain of strict clinical trial inclusion criteria (i.e., in patients with a high risk of bleeding) and with a less structured monitoring plan.8–10 Several ongoing registries, such as the Global Anticoagulant Registry in the FIELD (GARFIELD) and the Global Registry on Long-Term Oral Antithrombotic Treatment in Patients with Atrial Fibrillation (GLORIA-AF),11,12 are assessing DOAC safety, and the majority of data from large observational studies so far support the safety of DOAC agents used in routine care.10,13,14
The lack of a specific reversal agent is often cited as a disadvantage to DOAC use and may prompt some pre-scribers to preferentially use vitamin K antagonists such as warfarin.10,15–18 In 2014, FDA, in conjunction with the Cardiac Safety Research Consortium (CSRC), held a think tank to discuss the need for DOAC reversal agents.19 Although the frequency of need for such agents was considered very low, CSRC suggested that their availability could improve provider and patient confidence in DOAC use and promote an increase in the appropriate use of anticoagulant therapy for stroke prevention. Idarucizumab (Boehringer Ingelheim), a humanized monoclonal antibody fragment that binds to dabigatran, was approved by FDA in October 2015 for use in patients treated with dabigatran when reversal of the anticoagulant effects of dabigatran is needed (i.e., for emergency surgery or other urgent procedures or if there is life-threatening or uncontrolled bleeding).20 There are two other DOAC reversal agents in development: andexanet alfa, or PRT064445 (r-Antidote, Portola Pharmaceuticals, Inc., San Francisco, CA), an FXa decoy that binds to direct and indirect FXa inhibitors; and ciraparantag, or PER977 (Perosphere Inc., Danbury, CT), a potential reversal agent for both direct and indirect FXa inhibitors as well as factor IIa (FIIa) inhibitors (Table 1). Andexanet alfa received the breakthrough therapy designation from FDA in November 2013 and orphan drug status in February 2015.21,22 Portola Pharmaceuticals filed a biologics license application for andexanet alfa with FDA in December 2015.23 Ciraparantag received fast-track status from FDA in April 2015.24 The twofold goal of this article is to review these DOAC reversal agents and to assist pharmacists in planning for the entry of these agents into the marketplace.
Idarucizumab is a humanized monoclonal antibody fragment (Fab) that binds dabigatran and rapidly neutralizes the anticoagulant effect of dabigatran.25 To generate the antibody, a dabigatran-derived hapten coupled to carrier proteins was injected into mice to produce a monoclonal antibody. The sequence and clones of the antibody and the Fab were established and then humanized. The Fab was then recombinantly produced in Chinese hamster ovary cells. Idarucizumab does not contain the antibody fragment crystallizable (Fc) domain and does not bind to Fc receptors, proteins commonly found on the surface of cells. The primary dabigatran binding site of idarucizumab is a benzamide moiety located within a concave region at the interface of the heavy and light chain variable domains. Binding occurs through hydrophobic interactions, hydrogen bonds, and a salt bridge.25 As a result of the binding, the affinity of idarucizumab for dabigatran is approximately 350-fold stronger than the affinity of dabigatran for thrombin. Due to this strong affinity, idarucizumab is highly specific for dabigatran, does not bind to other thrombin substrates (factors V, VIII, or XIII, fibrinogen, von Willebrand factor, protease-activated receptor 1, and protein C) and has no effect on platelet aggregation. A single bolus dose of idarucizumab (0.3 μmol/kg) given to rats after a 20-minute continuous infusion of dabigatran (0.3 μmol/kg by bolus and then 0.1 μmol/kg/hr) completely reversed activated partial thromboplastin time (aPTT) and thrombin time (TT) prolongation within 1 minute despite continued dabigatran infusion.25
Idarucizumab was compared with prothrombin complex concentrate (PCC) in vitro for the reversal of the effect of concentrations of dabigatran of 184 ng/mL (representative of the maximum concentration [Cmax] with administration of dabigatran 150 mg twice daily) in human blood. In contrast to PCC, idarucizumab fully reversed all changes in coagulation parameters, thrombin generation, and thromboelastography parameters.26 In a porcine trauma model, the ex vivo effect of idarucizumab on anticoagulation reversal was compared with the effects of PCC, activated PCC (aPCC), and recombinant factor VIIa (rFVIIa).27 Dabigatran was administered at a dose of 30 mg/kg orally twice daily for 3 days and by i.v. infusion on day 4 to produce supratherapeutic anticoagulation (a mean ± S.D. dabigatran concentration of 1423 ± 423 ng/mL, as measured by dilute TT [dTT] assay). After induction of blunt liver injury resulting in shock, PCC, aPCC, rFVIIa, or idarucizumab were added to ex vivo samples to produce plasma concentrations equivalent to those produced by PCC doses of 30 and 60 units/kg, rFVIIa doses of 90 and 180 μg/kg, and idarucizumab doses of 30 and 60 mg/kg. Both PCC and aPCC reversed the effects of dabigatran on thromboelastography parameters and prothrombin time (PT), but dabigatran concentrations remained elevated, with a less than 50% reduction; rFVIIa had no effect on dabigatran concentrations or any coagulation parameter. Only idarucizumab resulted in unmeasurable dabigatran levels and normalized PT, aPTT, and thromboelastography parameters.
The effectiveness of idarucizumab (30, 60, or 120 mg/kg) in reversing bleeding associated with supratherapeutic dabigatran levels was also assessed using the porcine trauma model.28 Blood loss was measured after induction of a blunt liver injury. Idarucizumab reduced blood loss in a dose-dependent manner, with the highest dose (120 mg/kg) reducing blood loss at 4 hours to a mean ± S.D. volume of 1140 ± 109 mL, as compared with 2977 ± 316 mL in animals not receiving idarucizumab. Mortality was reduced from 100% in pigs not receiving idarucizumab to 0% in animals randomized to receive 60 or 120 mg/kg of the specific reversal agent.
Phase I studies
Glund and colleagues29,30 conducted a randomized, double-blind, placebo-controlled Phase I study (ClinicalTrials.gov identifier, NCT01688830) of idarucizumab in healthy male volunteers. This study comprised two parts: a single-rising-dose assessment of the safety, tolerability, and pharmacokinetics of idarucizumab alone (part 1) and a dose-finding, proof-of-concept investigation of idarucizumab in individuals pretreated with dabigatran etexilate (part 2). The idarucizumab doses studied took into account the large molecular-weight difference between dabigatran (471.5 daltons) and idarucizumab (47,800 daltons) (Figure 1).31
No clinically relevant differences were observed in terms of the overall incidence of adverse events between the placebo and idarucizumab groups.29 Adverse events that were considered to be drug related by the investigators occurred in five individuals (n = 2 in the placebo group, n = 3 with idarucizumab use). These drug-related adverse effects were mostly mild and, in idarucizumab-treated individuals, comprised headache (n = 2), erythema (n = 1), and migraine (n = 1). None of the adverse events resulted in drug discontinuation, and there was no development of persistent new antibodies against idarucizumab. Patient-reported local infusion-site reactions were limited to individuals in the 1-hour infusion group and included swelling (n = 1) and redness and pain (n = 1). Idarucizumab was associated with a transient, dose-dependent increase in urine protein and low-molecular-weight protein excretion, which was thought to be secondary to saturation of the renal tubular reuptake process for small proteins. No corresponding changes suggesting acute tubular injury or loss of function were noted, and values of urinary protein excretion normalized within 4–12 hours after the infusion.
Peak idarucizumab concentrations were dose dependent, occurred after the end of infusions (thus ensuring immediate availability for dabigatran binding in the plasma), and followed a monophasic-to-biphasic decline.29 Idarucizumab 5 g has a mean initial half-life of 47 minutes (geometric coefficient of variation [gCV], 11.4%).20 The geometric mean terminal elimination half-life and volume of distribution at steady state are 10.3 hours (gCV, 18.9%) and 8.9 L (gCV, 24.8%), respectively. The idarucizumab–dabigatran complex is eliminated, with elimination presumed to occur primarily through renal catabolism. In the study of Glund et al.,29 with increasing idarucizumab doses there was a corresponding dose-dependent increase in the amount of unchanged drug excreted in the urine, suggesting saturation of the renal tubule receptors responsible for reuptake of filtered proteins. Idarucizumab did not bind to other, structurally similar thrombin targets, as there was no procoagulant or anticoagulant impact on aPTT, dTT, ecarin clotting time (ECT), or activated clotting time (ACT) and no effect on endogenous thrombin potential.
In part 2 of the Phase I study of Glund et al.,30 the safety, tolerability, and efficacy of idarucizumab for reversing the anticoagulant effect of dabigatran were assessed. Dabigatran was administered to achieve steady state (220 mg twice daily for seven doses) in 47 healthy volunteers prior to administration of an i.v. placebo or short (5-minute) infusions of idarucizumab 1, 2, or 4 g, or two short infusions (doses of 5 and 2.5 g) given 1 hour apart. Mild adverse effects considered to be drug related occurred in two participants receiving idarucizumab: one experienced infusion-site erythema and hot flushes and another had epistaxis. Idarucizumab produced immediate, complete, and dose-dependent anticoagulant reversal, as measured by dTT, ECT, aPTT, and TT. Sustained reversal for 72 hours was evident with doses greater than 1 g. As demonstrated via dTT assay, the mean ratio of the area under the effect curve from 2 to 12 hours (AUEC2–12) on day 4 versus day 3 was 1.01 with placebo use, 0.26 with administration of the 1-g dose (a 74% reduction in dTT), 0.06 with the 2-g dose (a 94% reduction), 0.02 with the 4-g dose (a 98% reduction), and 0.01 with the two-dose (5 and 2.5 g) regimen (a 99% reduction). Endogenous thrombin potential and thrombin lag time were normalized in a dose-dependent manner to close to baseline values 0.5 hour after idarucizumab administration.
After idarucizumab administration, there was an immediate increase in the total dabigatran plasma concentration (bound and unbound dabigatran, including active metabolites) accompanied by a decrease in the unbound concentration of dabigatran (i.e., dabigatran not bound to idarucizumab or plasma proteins and being an approximate measure of pharmacologically active dabigatran). The increase in total dabigatran concentration was not associated with anticoagulant activity and thus likely represented inactive dabigatran bound to idarucizumab. The increase in total plasma dabigatran concentration is not clinically relevant and occurs as a result of the redistribution of dabigatran from tissues into the plasma.30 Idarucizumab has a small volume of distribution (at steady state, 8.9 L) and thus mostly stays within the blood compartment.20 Dabigatran, by contrast, has a large volume of distribution (at steady state, 50–70 L)1 and thus distributes between the blood compartment and the tissues to maintain an equilibrium between these compartments. After an idarucizumab infusion, dabigatran entering the blood compartment is quickly neutralized via idarucizumab binding, lowering the unbound serum dabigatran concentration. Dabigatran is then drawn from the tissues to the blood compartment to maintain the equilibrium, and this process is repeated as long as unbound idarucizumab is present. In the study of Glund et al.,30 after infusions of idarucizumab of 2 g or more, unbound serum dabigatran concentrations remained below 10 ng/mL over the 72-hour observation period and correlated most strongly with dTT (R2 = 0.86 in the presence of idarucizumab). Sustained inhibition of the anticoagulant effect is a result of the tight binding of dabigatran to idarucizumab. The idarucizumab–dabigatran complex is considered almost irreversible in preventing the release of active dabigatran.30
Although the effect of idarucizumab on bleeding could not be measured in the study, a potential marker was evaluated. Specifically, the effect of idarucizumab on restoring fibrin generation at a wound site was evaluated by measurement of fibrinopeptide A (FPA) levels. As FPA is released upon conversion of fibrinogen to fibrin and FPA levels are raised at wound sites, the FPA concentration represents an indirect measure of local activation of coagulation. Idarucizumab restored FPA concentrations at the wound site in a dose-dependent manner, with a 4-g dose producing a return of FPA concentrations to 95% of levels before dabigatran administration at 30 minutes after an infusion of idarucizumab.30
In a Phase Ib randomized, double-blind, two-way cross-over, placebo-controlled study, the effect of idarucizumab on reversing dabigatran-induced anticoagulation was assessed in 46 participants, including healthy volunteers 45–80 years of age and individuals with preexisting mild (creatinine clearance [CLcr] of 60 to <90 mL/min) and moderate (CLcr of 30 to <60 mL/min) renal impairment.33 Dabigatran therapy (220 mg twice daily for healthy volunteers and 150 mg twice daily for participants with renal impairment) was administered over four days. After the last dose of dabigatran, idarucizumab was administered in 5-minute infusions at single doses of 1, 2.5, or 5 g or via a split-dosing regimen consisting of two 2.5-g doses administered one hour apart.
Idarucizumab was well tolerated. Anticoagulation reversal, assessed through measurement of unbound dabigatran concentrations, occurred immediately at the end of an infusion. Values for aPTT, dTT, and ECT were all normalized by the end of the infusion, with sustained reversal over 24 hours evident after idarucizumab doses of 2.5 g or higher. Partial return of dabigatran anticoagulation occurred 2–4 hours after infusion of the 1-g dose. These data support the continued clinical evaluation of idarucizumab.
Phase III trial
The RE-VERSE AD (Reversal Effects of Idarucizumab on Active Dabigatran) trial (NCT02104947) is an ongoing Phase III multicenter, prospective cohort study of dabigatran-treated patients with overt, uncontrollable, or life-threatening bleeding serious enough that the treating physician has determined that a specific reversal agent is needed (group A) or requiring urgent surgery (i.e., a procedure that cannot be delayed by at least eight hours) or an invasive procedure necessitating normal hemostasis (group B).34 Unlike other trials of anticoagulant-reversal agents,35 this trial has included patients with acute trauma, as well as those with a very short life expectancy or requiring urgent surgery in the next eight hours. The lack of a standard-of-care approach to major bleeding episodes in dabigatran-treated patients, site variation in the approach to bleeding management, and ethical concerns about withholding a specific reversal agent supported a cohort trial design.31 The idarucizumab dose selected for the RE-VERSE AD study was derived from results of Phase I studies and in vitro data showing that an equimolar concentration (1:1 stoichiometry of dabigatran and idarucizumab) was needed for complete anticoagulation reversal. Using concentration data from the RE-LY trial, it was determined (based on trough levels) that a 5-g dose of idarucizumab would have been required to reverse the total body load of dabigatran in 99% of RE-LY trial participants with moderate renal dysfunction.34,36 Idarucizumab was administered as a 5-g dose given as two 50-mL bolus infusions of 2.5 g each within 15 minutes; this split-dosing regimen allowed for blood draws to be performed between the doses. The primary endpoint was the maximum percentage reversal of dabigatran anticoagulant effect, as assessed by either dTT or ECT assessment performed at a central laboratory (not available to treating physicians) within 4 hours after the second vial.
Numerous secondary and safety endpoints were assessed, including time to bleeding cessation (group A); intraoperative hemostasis (group B); duration of anticoagulation reversal; normalization of dTT and ECT values; time to achieve complete, at least 80%, and 50% reversal (based on dTT and ECT); minimum unbound dabigatran concentrations; mortality; and adverse events.31
Preliminary results for the first 90 patients in the RE-VERSE AD trial have been published34; patient characteristics are shown in Table 2. The majority (96%) were taking dabigatran for stroke prevention for atrial fibrillation. Just less than 65% were receiving the lower dose of dabigatran (110 mg twice daily). Approximately 35% of patients had a CLcr of <50 mL/min. In group A, gastrointestinal bleeding was the most common bleeding type (39%), followed by intracranial bleeding (35%). In group B, the median patient-reported time since the last dabigatran dose was 16.6 hours, and the most common urgent procedures were bone fractures, acute cholecystitis, and catheter placement for acute kidney injury.34
At baseline, 11 patients in each group (24.4% of all patients) had a normal dTT, with nine of these patients having a normal ECT. The median (range) unbound dabigatran concentration at baseline appeared to be similar between groups A and B: 84 (3–641) and 76 (4–2880) ng/mL, respectively. The median (95% confidence interval) maximum percentage anticoagulation reversal in both groups was 100% (100–100%), as indicated by both dTT and ECT values after the first idarucizumab dose. Normalization of the dTT occurred in 98% and 93% of evaluable group A and group B patients, respectively, and normalization of the ECT occurred in 89% and 88%, respectively (Figure 2). At 12 and 24 hours, the dTT was below the upper limit of normal in 90% of group A patients and 81% of group B patients; the corresponding ECT values were 72% and 54%, respectively. Normalization of aPTT and TT values, as assessed by the central laboratory, followed a similar pattern. At 4 hours after idarucizumab administration, the geometric mean plasma idarucizumab concentration had decreased by 80%. The concentration of unbound dabigatran was at or near the lower limit of detection immediately after idarucizumab administration in all but one patient and remained below 20 ng/mL in 93% of patients at 12 hours and in 79% of patients at 24 hours. At 12 and 24 hours, an increase in unbound dabigatran concentration corresponded to an increase in clotting time in six (6.7%) and 16 (17.8%) patients, respectively. The median investigator-reported time to bleeding cessation in group A was 11.4 hours. Normal intraoperative hemostasis was reported in 92% of group B patients. Four patients (n = 2 in each group) received aPCC, and 23 patients received fresh frozen plasma (group A, n = 14; group B, n = 9) as part of standard care.34 The overall mortality rate was 20%, with deaths deemed to be attributable to either the index event or coexisting conditions. Thrombotic events occurred in five patients while they were not receiving anticoagulation therapy, with one patient experiencing a thrombotic event within the first 72 hours after idarucizumab administration. Other serious adverse events included gastrointestinal hemorrhage (n = 2), postoperative wound infection, pulmonary edema, right ventricular failure, and delirium (n = 1 for each). In a subsequent interim analysis of RE-VERSE AD trial data on 123 patients, outlined in the idarucizumab prescribing information, the following adverse reactions were each reported in at least 5% of patients: hypokalemia, delirium, constipation, pyrexia, and pneumonia.20
Idarucizumab safely and rapidly reversed anticoagulation in this clinical setting. The history of dabigatran intake was provided by the patient or a physician. The median maximum percentage anticoagulation reversal was 100%. Almost a quarter (24.4%) of patients enrolled based on clinical assessment were subsequently found to have a normal dTT. However, in group A the proportion of patients with intracranial bleeding was higher among the 11 patients with normal clotting test results at baseline than among the 40 individuals with elevated results at baseline (64% and 28%, respectively); this suggests that patients can develop a life-threatening intracranial hemorrhage despite having normal coagulation tests. Nevertheless, the use of idarucizumab solely on the basis of clinical assessment may lead to overuse of this costly agent. More detailed information on the subgroup of patients with a normal dTT from the completed trial may provide additional insight to guide patient selection.
The degree of anticoagulation neutralization provided by idarucizumab decreases over time. In the RE-VERSE AD study, any increase in the unbound dabigatran concentration at 12 or 24 hours was associated with prolongation of dTT and ECT.34 This phenomenon may reflect redistribution of dabigatran from tissue stores back into the plasma and highlights the potential value of a readily accessible coagulation assay for dabigatran. ECT and dTT data were not available between 4–12 hours after idarucizumab administration; thus, the return of anticoagulation within that time frame may have occurred. In practice, there may be a desire to readminister idarucizumab, such as in the case of ongoing clinically relevant bleeding together with elevations in coagulation parameters. Additional safety and efficacy data regarding repeat idarucizumab administration are needed.
Based on current data, the potential immunogenicity of idarucizumab does not appear to be a concern. Approximately 15% of normal persons have naturally occurring Fab antibodies, although experience with other therapeutically used Fab fragments suggests that these naturally occurring antibodies are unlikely to affect idarucizumab’s efficacy.16 In a Phase I study, preexisting idarucizumab antidrug antibodies were detected at baseline in 10.8% of individuals receiving idarucizumab, and most of these patients continued to have idarucizumab antidrug antibodies through to three months.29 Antibodies were directed against the C-terminus of the Fab and not the dabigatran binding site. Importantly, idarucizumab did not result in the formation of new persistent antibodies, and there were no adverse events consistent with immunogenic reactions.29 The RE-VERSE AD protocol includes a 90-day follow-up period during which antibodies against idarucizumab will be assessed (data not yet available). In the presence of preexisting antibodies, there is a theoretical concern for decreased renal elimination of complexed dabigatran, idarucizumab, and anti-Fab antibody, as the size of the complex may be too large to be filtered. In such cases, reanticoagulation with dabigatran could be difficult to establish. Glund and associates31 however, demonstrated that dabigatran anticoagulation could be successfully reestablished 24 hours after idarucizumab use (n = 12). Anti-idiotype antibodies (antibodies directed against an epitope of the variable region of another antibody) present a theoretical risk in that they could potentially inactivate idarucizumab with repeated patient exposure to the drug.16 Glund et al.31 demonstrated that repeat exposure to idarucizumab two months after initial exposure in six individuals was safe and effective; there was no impact on the pharmacokinetics or reversal effect of idarucizumab or on the frequency of hypersensitivity reactions in treated subjects.
Lastly, information on the potential to use coagulation laboratory testing to monitor idarucizumab therapy is needed. A commercially available dTT test is under FDA review.37 If the test is approved and made readily available, its potential role in guiding idarucizumab patient selection needs further evaluation. Additional detailed information from the completed RE-VERSE AD trial on the ability to use the centralized aPTT for monitoring idarucizumab therapy is also needed.
Andexanet alfa is a modified recombinant version of FXa that is a potential specific reversal agent for anticoagulants that primarily target FXa. This bioengineered variant has also been referred to in the literature as r-Antidote.38 The modified protein serves as a decoy for small-molecule direct FXa inhibitors and is produced in Chinese hamster ovary cells. The molecule contains three structural modifications that distinguish it from native FXa. First, the amino acid serine is replaced by alanine at position 419, removing any catalytic activity. This modification, along with deletion of a 34-residue fragment of γ-carboxy-glutamic acid (GLA), also allows for binding to direct FXa inhibitors and heparin–antithrombin complexes without initiating enzymatic activity. This latter modification also prevents binding of andexanet alfa to phospholipid surfaces of cells and potential competition with native active FXa in the active clotting cascade. As a result, activation of endogenous FX to FXa by the complex of tissue factor and FVIIa is not interfered with, allowing endogenous hemostasis to occur (Figure 3).38,39
Andexanet alfa binds to FXa inhibitors with affinity similar to that of native FXa; reported mean ± S.D. binding affinities for rivaroxaban and apixaban are 1.53 ± 0.22 and 0.58 ± 0.02 nM, respectively.38 Andexanet alfa has the potential for reversing indirect FXa inhibitors by serving as a decoy for activated antithrombin. As it only competes with FXa for antithrombin binding, andexanet alfa may potentially be more effective at reversing fondaparinux than low-molecular-weight heparins (LMWHs), given the partial antithrombin activity of LMWHs.43
Preclinical studies of rats demonstrated that andexanet alfa (a 4- or 6-mg i.v. bolus plus a 4-, 6-, or 9-mg/hr 90-minute infusion, depending on the anticoagulant used) rapidly and completely normalized the whole blood International Normalized Ratio (INR) after anticoagulation with rivaroxaban (0.25 mg per kilogram of body weight per hour), betrixaban (1 mg/kg/hr), or apixaban (0.5 mg/kg/hr).38 The correction of the whole blood INR was directly correlated with the decrease in the free fraction of the FXa inhibitor, thereby validating the mechanism of reversal. In addition, research using animal models of bleeding demonstrated the ability of andexanet alfa to stem bleeding.38 In rabbits given rivaroxaban and subsequently exposed to liver laceration, bleeding was stemmed by 85% after andexanet alfa administration. In rat-tail transection models, andexanet alfa significantly reduced blood loss when animals were exposed to enoxaparin and fondaparinux. These animal studies formed the basis to pursue investigations of andexanet alfa in humans involving multiple direct and indirect FXa inhibitors.
Phase I and Phase II studies
In a Phase I, first-in-man study of healthy volunteers (n = 32), individuals were randomly assigned to receive a single i.v. bolus of andexanet alfa (30, 90, 300, or 600 mg) or a placebo. Participants were not treated with rivaroxaban, but anti-FXa activity was measured ex vivo after the administration of rivaroxaban 50 ng/mL to plasma samples.41 No thrombotic adverse events or deaths occurred. One serious event (pneumonia), as well as three nonserious infusion-related reactions (n = 2 with the 90-mg dose, n = 1 with placebo use), was reported. Andexanet alfa had a terminal half-life of 6 hours, restored thrombin generation, and reversed the anti-FXa activity of rivaroxaban in a dose-dependent manner. Prothrombin fragment 1+2, thrombin–antithrombin complex, and d-dimer levels transiently increased with andexanet alfa dose, while other coagulation parameters (PT, aPTT, and ACT) and platelet activity did not change. In an additional preliminary study, administration of an i.v. bolus of andexanet alfa was demonstrated to produce a rapid and dose-dependent restoration of thrombin generation, with sustained reversal after the addition of a slow infusion of andexanet alfa in healthy volunteers pretreated with apixaban 5 mg twice daily, rivaroxaban 20 mg daily, or enoxaparin 40 mg daily for 6 days. Andexanet alfa was well tolerated; the most prominent safety signal was mild-to-moderate infusion reactions that were not dose-limiting and mostly did not require interventions.44
Subsequent to preclinical efforts, multiple Phase II trials with various FXa inhibitors (rivaroxaban, apixaban, edoxaban, and betrixaban) have been conducted to further establish the role of andexanet alfa as a specific reversal agent. Results from these investigations have been made available primarily in abstract form and presented at various international meetings. The first available Phase II study results for andexanet alfa demonstrated that in healthy volunteers (n = 9) who had received apixaban 5 mg twice daily for 6 days and subsequently received a single i.v. bolus dose of andexanet alfa 90 mg, anti-FXa activity was reduced by an average of 65% within two minutes. Treatment with andexanet alfa was associated with a transient reduction in TFPI activity and an increase in prothrombin fragments F1 + F2 but no change in d-dimer concentration.40 Andexanet alfa was well tolerated in all study participants; mild adverse events were reported in five individuals.
Additional published results have since emerged from this study, in which healthy volunteers pretreated with apixaban 5 mg twice daily for six days were randomly assigned to placebo use or six different doses of andexanet alfa (a 90-mg i.v. bolus, a 210-mg i.v. bolus, a 420-mg i.v. bolus, a 420-mg i.v. bolus plus a 45-minute 4-mg/min infusion, a 420 mg i.v. bolus plus a two-hour infusion of 480 mg, or a 420 mg i.v. bolus plus a repeat bolus at 45 minutes). Two minutes after administration of the 210-and 420-mg i.v. bolus doses, anti-FXa activity was reduced by 80% and 90%, respectively, which corresponded to reductions in free plasma concentrations of apixaban. In addition, the administration of a 420-mg i.v. bolus followed by a 4-mg/min infusion resulted in complete restoration of thrombin generation throughout the administration of the specific reversal agent. No thrombotic or bleeding events or immunologic reactions to andexanet alfa were reported.45
In a similar double-blind, placebo-controlled study, healthy individuals were pretreated with rivaroxaban 20 mg daily for 6 days and then randomly assigned to receive different doses of andexanet alfa (a 210-mg i.v. bolus, a 420-mg i.v. bolus, a 600-mg i.v. bolus, or a 720-mg i.v. bolus followed by a 240-mg infusion over one hour; n = 6 per cohort) or an i.v. placebo (n = 3 per cohort). Andexanet alfa was given three hours after the last dose of rivaroxaban on day 6, corresponding to the peak effect of the drug. For the 210- and 420-mg cohorts (total n = 12), anti-FXa activity decreased by 20% and 53%, respectively, from values before administration of andexanet alfa and returned to levels associated with placebo use 2 hours after treatment; unbound plasma concentrations of rivaroxaban were also decreased (by 32% and 51%, respectively). Inhibition of thrombin generation was partially reversed by andexanet alfa. The reversal agent was well tolerated, with no thrombotic or serious adverse events noted. Adverse events occurring in andexanet alfa or placebo recipients included mild infusion-related reactions (n = 3) and postprocedural hematoma, headache, or postural dizziness (n = 2 for each).42 Additional data on recipients of the 600- and 720-mg doses of andexanet alfa demonstrated reductions in anti-FXa activity of 70% and 81%, respectively, with reductions in unbound plasma rivaroxaban concentrations of 75% and 70%.45
Similar results have been reported with the use of andexanet alfa for reversal of the anticoagulant effect of edoxaban in healthy individuals (n = 18). Participants receiving edoxaban 60 mg once daily for 6 days were randomly assigned to placebo use or a 600-mg i.v. bolus of andexanet alfa or an 800-mg bolus followed by an 8-mg/min infusion for one hour three hours after the last edoxaban dose on day 6 (n = 18). Anti-FXa activity was reduced by 52% and 73% at 2 minutes after administration of the 600- and 800-mg andexanet alfa doses, respectively, and remained constant during i.v. infusion. Anti-FXa levels returned to baseline two hours after treatment. Andexanet alfa was well tolerated, with no thrombotic events or serious or severe adverse events reported. These data are consistent with data from previous investigations conducted with other direct FXa inhibitors.46
Lastly, in a Phase II study, healthy individuals pretreated with enoxaparin 40 mg subcutaneously once daily for 6 days subsequently received a single i.v. bolus of andexanet alfa of 210 or 420 mg or a placebo. Two minutes after bolus administration of andexanet alfa, anti-FXa activity was reduced below the therapeutic anticoagulation threshold of enoxaparin, and thrombin generation was reversed to baseline levels and maintained for two to three hours. Adverse events related to andexanet alfa or placebo use and occurring in at least 10% of participants consisted of mild infusion-related reactions (n = 4).47
Phase III/registration trials
Three separate Phase III investigations were designed and initiated to further evaluate the role of andexanet alfa in reversing anticoagulation with FXa inhibitors. Results from two of these studies, which evaluated andexanet alfa in healthy older adults, have now been published together.48 Both trials tested the effects of andexanet alfa as a specific reversal agent in healthy volunteers 50–75 years of age who received either rivaroxaban 20 mg daily for 4 days (the ANNEXA-R trial, NCT02220725) or apixaban 5 mg twice daily for 3.5 days (the ANNEXA-A trial, NCT02207725). Participants were randomly assigned in a double-blind, placebo-controlled fashion to either andexanet alfa or placebo use in a 3:1 ratio (ANNEXA-A) or a 2:1 ratio (ANNEXA-R). Both trials were conducted in two phases, with each phase investigating a different dosing scheme for andexanet alfa. In phase 1 of the ANNEXA-A trial, a single 400-mg i.v. bolus dose of andexanet alfa was evaluated, with phase 2 evaluating a 400-mg i.v. bolus followed by a 4-mg/min infusion for two hours. Phase 1 of the ANNEXA-R study evaluated an 800-mg i.v. bolus of andexanet alfa, and phase 2 an 800 mg i.v. bolus followed by an 8-mg/min i.v. infusion for two hours. Study participants were housed at the study site for 8 days, and all subjects had safety outcomes assessed at days 15, 36, and 43 after study drug administration.
Participants in the ANNEXA-A trial received 5 mg of apixaban orally twice daily for 3.5 days in order to achieve steady-state levels prior to administration of a specific reversal agent. Three hours after the last dose of apixaban on day 4, andexanet alfa or a placebo was administered to study subjects. In the ANNEXA-R study, participants received rivaroxaban 20 mg orally once daily for 4 days in order to achieve steady state. Four hours after the last dose of rivaroxaban on day 4, andexanet alfa or a placebo was administered to study subjects. Doses of andexanet alfa were derived from Phase II investigations and designed to reverse the effects of each agent at steady state taking into account the difference in peak plasma concentration for each agent, as well as the volume of distribution for each agent (larger for rivaroxaban). The primary outcome in both studies was the percentage change in anti-FXa activity from baseline to nadir after andexanet alfa or placebo administration. Secondary endpoints included the proportion of patients with a reduction in anti-FXa activity of ≥80% from baseline, changes in unbound concentrations of rivaroxaban and apixaban from baseline, and changes in thrombin generation from baseline. The primary efficacy analysis was conducted in a modified intention-to-treat format that included all participants who underwent randomization or received any amount of andexanet alfa or placebo and in whom data on anti-FXa activity were available at baseline and after administration of andexanet alfa or the placebo.
Baseline characteristics for all study participants can be viewed in Table 3 and were balanced between groups. In both phase 1 and phase 2 components of the ANNEXA-A and ANNEXA-R trials, anti-FXa activity was reduced rapidly and to a greater extent in participants receiving andexanet alfa than in placebo recipients (Table 4). All participants treated with andexanet alfa achieved an 80% or greater reversal of anti-FXa activity except for 1 patient who did not receive the full dose of andexanet alfa due to an error with the i.v. infusion device. Thrombin generation was also rapidly restored upon administration of andexanet alfa, and unbound plasma concentrations of either apixaban or rivaroxaban were reduced to a greater degree with andexanet alfa treatment versus placebo use (Table 4). In addition, mean plasma concentrations of unbound apixaban or rivaroxaban after andexanet alfa administration were below levels considered to have significant clinical effects (3.5 ng/mL for apixaban, 4.0 ng/mL for rivaroxaban). In participants receiving andexanet alfa, reductions in anti-FXa activity, plasma concentrations of apixaban or rivaroxaban, and increases in thrombin generation were observed within 2–10 minutes and maintained for two hours after the end of either bolus (phase 1) or infusion (phase 2) dosing. An important observation was that plasma concentrations of unbound apixaban or rivaroxaban returned back to placebo-associated levels two hours after the end of the bolus or infusion; these concentrations may well be clinically significant and exert anticoagulant activity, as evidenced by results of anti-FXa activity testing. No serious adverse events were observed, and no thrombotic events were reported. Neutralizing antibodies to andexanet were not detected, although nonneutralizing antibodies were observed in 17% of the participants receiving andexanet alfa. Additional observations included transient increases in d-dimer levels and levels of prothrombin fragments 1 and 2, with a return to normal ranges in 24–72 hours.
There are important questions about andexanet alfa still to be answered, including the appropriate dose and duration of therapy for andexanet alfa, depending on the oral FXa inhibitor targeted. Results from the ANNEXA-A and ANNEXA-R studies have identified what appears to be a successful dose of andexanet alfa for reversal of anticoagulation with apixaban or rivaroxaban, but these results were achieved in healthy volunteers, and plasma concentrations of those anticoagulants returned to clinically significant levels two hours after the administration of either an i.v. bolus or infusion of andexanet alfa. The duration of administration to successfully manage a patient on chronic therapy with apixaban or rivaroxaban who is bleeding still needs to be defined and may very well differ from patient to patient depending on the total body load of these agents and underlying renal function. Doses and durations of therapy used for potential reversal of the effects of LMWHs or fondaparinux are also likely to be different and still need to be defined. Whether or not the timing of the last FXa inhibitor dose has a role in andexanet alfa dosing requires further exploration. Without a readily available laboratory test to quantitate the level of FXa anticoagulation, guidance will be needed on how to determine the duration of any continuous infusion of andexanet alfa. Additional issues that still have to be resolved include establishing when anticoagulation could be reinstated, particularly if the patient needs an emergent rescue (e.g., extracorporeal life support where immediate heparinization is needed), and whether alternative strategies such as bivalirudin are needed in these situations. Transient increases in d-dimer levels and levels of prothrombin fragments 1 and 2 were observed in the ANNEXA trials, although this did not result in significant increases in thrombin generation or clinical thrombotic events. Nevertheless, safety data from patients with bleeding complications will be needed to exclude any potential for a prothrombotic effect. As with all structurally modified proteins, immunogenicity may become an issue and will need to be carefully monitored in clinical practice. Finally, an ongoing trial (NCT02329327) assessing andexanet alfa for the management of patients with bleeding associated with direct and indirect FXa inhibitors is actively recruiting patients.49 Inclusion and exclusion criteria are available from the ClinicalTrials.gov website. Results from this trial may be able to address some of the important issues outlined above.
Ciraparantag, or PER977, is a synthetic small, water-soluble molecule designed specifically as an i.v. specific reversal agent targeting DOACs and heparins.50 Ciraparantag exerts its anticoagulant reversal effect through strong, physical, noncovalent bonds (Figure 4). It has been shown to form a complex with large molecules such as unfractionated heparin and LMWH, as well as with smaller molecules such as fondaparinux, apixaban, edoxaban, rivaroxaban, and dabigatran.51 Once these complexes are formed, subsequent interactions with antithrombin and activated clotting factors are inhibited (Figure 5). Ciraparantag does not reverse warfarin activity. While ciraparantag bonds with argatroban, this interaction does not interfere with the binding of argatroban to FIIa.52 Ciraparantag exhibits no binding to serum albumin or other plasma coagulation factors and, therefore, has no procoagulant effect. Ciraparantag has progressed through the early stages of clinical development and is now being investigated in Phase II human trials (Table 5).52,53
Ciraparantag halted tail bleeding in rats that were overdosed with rivaroxaban, edoxaban, dabigatran, or apixaban and restored PT (in rats that received a direct FXa inhibitor), aPTT (in rats that received dabigatran), and thromboelastography (in rats that received endoxaban) assays to baseline levels within 20 minutes of i.v. administration.54 In toxicology trials, ciraparantag was metabolized to 1,4-bis(3-aminopropyl)piperazine, which was rapidly renally eliminated, and showed no neurologic, cardiovascular, or respiratory effects or genotoxicity.55
In the in vitro setting, platelet-poor plasma from normal volunteers was spiked with rivaroxaban or apixaban and then treated with ciraparantag. FXa activity was measured with a chromogenic anti-FXa assay. Ciraparantag completely reversed rivaroxaban and apixaban anti-FXa activity in a dose-dependent fashion.56
Phase I and Phase II studies
Ciraparantag has been investigated in healthy human volunteers. In a Phase I/II dose-ranging cohort trial, 40 healthy volunteers were treated with enoxaparin 1.5 mg/kg subcutaneously followed by ciraparantag acetate 100, 200, or 300 mg (a single dose) i.v. or an i.v. placebo 4 hours later (the time to enoxaparin Cmax [Tmax]). After administration of enoxaparin, whole blood clotting time (WBCT) increased by 28.5% over baseline values. Ciraparantag reversed enoxaparin’s impact on WBCT within 20 minutes of administration of a 100-mg dose and within 5 minutes of a 200-mg dose. There was no rebound anticoagulation or signs of procoagulant effect, as measured by d-dimer, prothrombin fragment 1.2, and TFPI levels. The adverse events were limited to transient temperature sensations and flushing.57 Ciraparantag was investigated in a double-blind, placebo-controlled, dose-escalation trial involving 80 healthy volunteers. In period 1, participants were randomly assigned to cohorts to receive single-dose i.v. ciraparantag 5, 15, 25, 50, 100, 200, or 300 mg or an i.v. placebo. In period 2, at a minimum of 3 days later, study participants received oral edoxaban 60 mg followed by single-dose i.v. ciraparantag (at the above doses) or a placebo 3 hours later (at edoxaban Tmax). Edoxaban administration increased WBCT by 37% over the baseline value. In those participants receiving ciraparantag 100, 200, or 300 mg, WBCT decreased to within 10% above the baseline value in 10 minutes or less, suggesting a full reversal of anticoagulant effects, and remained in that window for 24 hours. In the placebo group, the time to reach a WBCT decline to within 10% above the baseline value was significantly longer, ranging from 12 to 15 hours. In addition, blood clots from the WBCT were visually evaluated using scanning electron microscopy. Edoxaban reduced the mean fibrin–fiber diameter from baseline; the mean diameter returned to normal 30 minutes after a ciraparantag 100–300 mg administration. Ciraparantag appeared to have no procoagulant effect, as measured by d-dimer, prothrombin fragment 1.2, and TFPI levels. Adverse events with ciraparantag use were mild and limited to transient perioral and facial flushing, taste distortion, and headache.58
Interim data (n = 50) Phase-II, single-blind, placebo-controlled trial was recently presented. This study evaluated escalating doses of ciraparantag in subjects administered edoxaban 60 mg daily for 2 days to attain steady state. Ciraparantag was administered in single doses of 25, 50, 100, 300, or 600 mg on day 3, with anticoagulation reversal measured by WBCT at 3 hours after dose administration. Anticoagulation was then reinitiated with edoxaban on day 4 at the time of the next scheduled dose. A second reversal, with the same ciraparantag doses, was completed to ensure that reanticoagulation was not compromised. Ciraparantag at doses of 100 mg or higher completely reversed the anticoagulation produced by steady-state doses of edoxaban within 60 minutes, as shown by a decrease in the mean WBCT to pre-edoxaban levels. The mean WBCT following reanticoagulation with edoxaban was similar to that observed after the first anticoagulation period. The impact of the second reversal with ciraparantag 100 mg was similar to that of the first. Similar to findings in previous trials, a single ciraparantag administration sustained reversal of edoxaban without the need for prolonged infusion and did not interfere with reanticoagulation the following day.59
Future ciraparantag development
Ciraparantag has generated a great deal of excitement as a broad-spectrum specific reversal agent for both DOACs and LMWH. Ciraparantag is also being studied as a replacement for protamine use with unfractionated heparin. It has the ideal characteristics of single-dose administration, rapid onset of action, 24-hour duration, and no rebound anticoagulation requiring accompanying infusions. Furthermore, human studies suggest that WBCT can be successfully used as a biomarker to measure anticoagulant effect and its reversal. WBCT shows low variability and high reproducibility and correlates well with plasma concentrations of direct FXa inhibitors and LMWH pharmacokinetics. However, WBCT is not currently a commonly used laboratory test and will need to be automated if it is to be used routinely in the clinical arena. The role of ciraparantag in the management of anticoagulation-related bleeding and anticoagulant reversal will unfold with the completion of planned Phase III clinical trials.
One potential limitation of a universal reversal agent is that it may interfere with the emergent use of an anticoagulant during a procedure, such as the use of heparin in patients requiring extracorporeal life support or cardiopulmonary bypass support. Research is also required to clarify when anticoagulation can be restarted after the administration of this universal reversal agent. The availability of a universal reversal agent still requires knowledge of the doses needed to reverse the effects of different anticoagulants and requires that the clinician avoid making an assumption that it will counter the effects of all antithrombotics.
Preparing for the arrival of specific reversal agents
When novel anticoagulants became available, several experts advised practitioners and hospitals to devise comprehensive management plans for prompt reversal of their effects in patients with bleeding and those requiring urgent surgical intervention.60–62 The focus of these recommendations was on the use of appropriate laboratory assays for measuring DOAC plasma concentrations and on blood product management. Now clinicians will be faced with incorporating specific rapid-acting reversal agents into their treatment plans, which will generate a new series of considerations.
Ordering, storage, and preparation
Hospital pharmacies will be charged with the responsibility of setting up the supply chain management of these agents. The simplest questions center on which agents to purchase and how much of each agent should be maintained in the inventory. These questions may, in part, be answered by the medical staff’s prescribing patterns and patient utilization of the agents. A busy tertiary care facility with many referrals would be expected to ensure the availability of all of the agents, whereas a community hospital with a single formulary agent may only carry a specific reversal agent whose properties mirror institutional utilization. However, other factors may need to be taken into account when deciding which agent to stock. For example, a remote critical access hospital may require all specific reversal agents to adequately manage urgent cases of bleeding in patients receiving anticoagulants, as patient transfer to another facility may take some time. Furthermore, approved indications for the individual specific reversal agents are likely to differ according to differences in the patient populations evaluated in the Phase III trials: Idarucizumab has been evaluated both in patients with serious bleeding and in those requiring emergency surgery, whereas andexanet alfa is being evaluated only in those with life-threatening bleeding episodes.
Since idarucizumab and andexanet alfa are protein products, hospitals must have adequate storage resources for refrigeration, as well as appropriate tubing and inline filters that may be required for their administration. Moreover, all of the agents discussed here are administered intravenously, with andexanet alfa, in particular, requiring delivery via infusion pump. This may, therefore, require programming in “smart” infusion device libraries. Other agents may be administered as an i.v. bolus dose without the use of an infusion device. Preparation needs for the agents will differ. Idarucizumab is available as a single package consisting of two 50-mL vials (each containing 2.5 g of the drug). Reconstitution is not required, and the contents of each of the two vials is administered as consecutive infusions over 5–10 minutes.63 This can be accomplished by i.v. push administration using a syringe or via infusion by hanging the vials. No administration rate is included in the manufacturer labeling. A reasonable approach is to set device libraries at a target rate of 600 mL per hour, infusing each vial over approximately 5 minutes. For those devices with limits, a maximum infusion rate of 900 mL per hour and a minimum infusion rate of 300 mL per hour will ensure idarucizumab administration ranging from 3 to 10 minutes. Andexanet alfa is expected to require reconstitution prior to administration.
While patients needing urgent anticoagulant reversal can be expected to present to emergency departments for their care, patients receiving anticoagulant therapy are often transitioned to procedure areas for cardiac catheterization or electrophysiology testing. Since specific reversal agents must be rapidly available and drug preparation can be time-consuming, it is logical to determine in advance the most suitable location for this task. As ciraparantag is not a biological agent and is stable at room temperature, it has the potential for out-of-hospital use.
Patient selection and monitoring
The DOACs have been shown to induce fewer bleeding episodes than vitamin K antagonists.5,64 Furthermore, their pharmacokinetic profiles allow bleeding episodes to be managed with a conservative “watchful waiting” strategy rather than aggressive intervention. Restoring coagulation may actually be harmful in many settings.65 Consequently, the role of specific reversal agents will be limited to specific clinical presentations such as overdose, life-threatening bleeding, intracranial or retroperitoneal hemorrhage, imminent emergency surgery, and decreased clearance of the anticoagulant due to renal failure, as well as situations in which all other interventions have failed.61,66 Although the specific reversal agents represent a major breakthrough, important questions remain regarding their optimal use (Table 6). While clinical outcomes are more important than measuring anticoagulant effect, a monitoring plan specifying which (and at what frequencies) laboratory tests are to be performed may be helpful in standardizing DOAC reversal strategies. The primary endpoints of the Phase III clinical trials for idarucizumab and andexanet alfa required the use of laboratory tests that were not routinely available. The ability of clinicians to recognize a return to an anticoagulated state will be limited by the current routinely available screening tests, such as those measuring aPTT (for use in idarucizumab monitoring) or PT (for use in monitoring the reversal of certain direct FXa inhibitors). Establishing a monitoring plan will be important and may require individualization. The plan should include an evaluation of the need for repeat diagnostic testing to verify cessation of bleeding until patient stabilization occurs, consideration of the utility and value of objective testing (e.g., repeat aPTT in patients with known elevated aPTT values prior to idarucizumab administration) as well as an assessment for thrombotic events. Since multiple or repeat doses of specific reversal agent may be required, documentation through care transitions or patient transfers must be in place to ensure that all doses are administered correctly. Because specific reversal agents will be employed in life-threatening situations and are expected to carry a high cost, it makes sense to incorporate their use into a health system’s monitoring or surveillance program. Antithrombotic stewardship programs have been shown to improve patient outcomes, decrease cost, and support the development and implementation of institution-specific protocols and guidelines.67,68 Once approved, each institution should establish a protocol for the use of DOAC reversal agents in the event of life-threatening bleeding or situations involving a need for urgent or emergent surgical or invasive intervention requiring normal hemostasis. For the idarucizumab protocol, clinicians should be aware that the inhibition of dabigatran by idarucizumab does not appear to be influenced by volume replacement or administration of coagulation factor concentrates.20 In terms of patient selection, the risks versus benefits of idarucizumab should be assessed in patients with hereditary fructose intolerance and in those who have received sorbitol, as serious adverse reactions to the sorbitol excipient can occur. An order set based on the list of anticoagulant agents to be targeted for reversal should be crafted for those hospitals with a computerized order-entry system or an electronic health record. This will be especially important in the setting of the emergency department. A common framework of care and supportive measures ensures that the necessary products are available in emergency situations without delay or debate. This may require redefining or changing previous recommendations surrounding PCC and aPCC and their role in life-threatening bleeding or imminent surgical intervention. While usually supported by an interdisciplinary team, the program steward—an attending hematologist or cardiologist—can serve as the contact for resolving conflicts regarding patient and treatment selection or the laboratory parameters to target for reversal.68,69 A stewardship program is probably not feasible at a small rural hospital, but that does not preclude planning and preparation—whether the specific reversal agents are available or not. Identification of a local expert, creation of a treatment plan, duplication of a treatment algorithm from a referral center, incorporation of specific reversal agents into disaster-recovery planning, and simulation workshops or drills will help facilitate patient management.
Once a patient with spontaneous bleeding is treated with a specific reversal agent, it makes sense to identify the patient as having a “high” bleeding risk while on anticoagulant therapy. Many patients are unaware of the attributes of their medications and the potential risks.70 For treated patients, it is prudent to reinforce education surrounding anticoagulant use and recognition of bleeding signs and symptoms.
The most frequently cited cause of medication prescribing errors is inadequate knowledge of the medication or the patient’s condition.71 Despite available electronic resources, there will be knowledge gaps with respect to drug information on specific reversal agents and their use in minor and life-threatening complications, as well as routine issues surrounding correct agent selection, dosing, and therapeutic response. National guidelines are unlikely to be updated in time for FDA approval. Because reversal agent administration is expected to be an infrequent event, each episode represents a learning opportunity for the practitioners involved. A system review or root-cause analysis of each event provides an opportunity to troubleshoot or identify areas or opportunities for improvement. Health systems are encouraged to prepare a checklist (Table 7) of the steps that can be accomplished before the specific reversal agents arrive and when they become established into routine use. Advance planning will help ensure proper supply management, patient selection, provider education, and treatment follow-up and focus quality improvement efforts. Since the emergency department represents a major patient interaction point and is the most common portal of hospital entry, pharmacists in this environment are uniquely poised to improve patient selection, diminish the risks of adverse events, and provide pharmacologic information to staff on the specific reversal agents. Emergency department pharmacists can serve as the catalyst for organizing activities surrounding these agents. They should seek education, or educate themselves, and be prepared to broaden their role in anticoagulant-related bleeding management.