Why the study?
Experience with andexanet alfa in patients requiring urgent or emergent procedures is limited, as such patients were excluded from the ANNEXA-4 trial.
Does andexanet alfa provide adequate hemostasis and safety in patients requiring emergent procedures shortly after treatment for life-threatening bleeding?
Does andexanet alfa provide adequate hemostasis and safety in patients requiring emergent procedures shortly after treatment for life-threatening bleeding?
Andexanet alfa may not provide adequate duration of anticoagulation reversal for prolonged or delayed emergent procedures due to its short half-life, highlighting the need for multidisciplinary periprocedural guidance.
Answers are grounded in this paper and cite their sources.
May support selected urgent procedures after FXa-inhibitor reversal but rebound cautions use; leaves open dedicated peri-procedural trials.
Andexanet alfa is the first FDA approved agent (May 2018) for factor Xa inhibitor (FXai) reversal. As previously reported,1 patients with planned surgery within 12 hours (updated from 24 hours amendment 1) of andexanet alfa administration were excluded from the ANNEXA-4 trial, unless minimally invasive diagnostic or therapeutic procedures were required such as endoscopy, bronchoscopy, or central line placement. Experience with use in patients requiring urgent or emergent procedures is limited.2-4 Real-world scenarios lead to situations that fall outside of clinical trial parameters, such as patients who require emergent procedures concurrently, or soon after treatment of uncontrolled bleeding with andexanet alfa, as with trauma patients. Although our institution does not approve the use of andexanet alfa to reverse anticoagulation in the non-bleeding patient requiring emergency surgery, we have had experience with patients requiring emergent surgery after treatment with andexanet alfa for life threatening bleeding.1, 5 As of June 2019, we have administered andexanet alfa to 41 patients, 12 of whom required an emergent procedure within 24 hours (Table 1). Six of the treated patients were receiving rivaroxaban and six were receiving apixaban. All patients had taken the oral FXai within 18 hours of presentation. Nine received the standard dose of andexanet alfa, and three received the high dose. Four patients had pre-andexanet alfa anti-Xa (unfractionated heparin [UFH]/low molecular weight heparin [LMWH] calibrated) levels that ranged from 1.26 to 4.74 units/mL. Post-andexanet, pre-operative anti-Xa levels were available for two patients, one taken 19 hours after andexanet alfa administration, and one taken 8.75 hours after administration, with levels ranging from 1.37 to 3.26 units/mL. Nine patients had andexanet alfa infusing during the procedure, of which two patients' procedures were longer than the andexanet infusion. Three of the 12 patients had procedures several hours after andexanet administration, ranging from 6 to 19 hours. Seven patients presented with intracranial hemorrhage (ICH). Procedures were needed for the index bleeding event in nine patients (75%). In five patients with ICH, two had neurosurgery with external ventricular drain (EVD) placement (patients one and seven) and three required craniotomies (patients six, eight, and 12). In non-ICH reversal indications, four required surgery for life-threatening bleeding associated with the indication for reversal (patients three, four, nine and 11) and three patients treated with andexanet for ICH subsequently required surgery for other indications (patients two, five, and 10). Four were treated with 4 factor prothrombin complex concentrate (4F-PCC) prior to hospital transfer (patients one, six, seven, and 12). Two (patients three and 12) did not achieve hemostasis and required additional intra-operative transfusion with packed red cells, fresh frozen plasma (FFP), cryoprecipitate, and 4F-PCC (see table). Estimated blood loss ranged from 0 mL to 1200 mL. No thrombotic events were observed at 7 days after administration of andexanet alfa. Mortality at discharge was 25%. 17 hours Rivaroxaban 20 mg daily Unknown Bolus started at time of incision, infusion ended after procedure ended 18 hours Rivaroxaban 20 mg daily 315 min Incision at completion of bolus; infusion ended before end of procedure >12 hours Rivaroxaban 20 mg daily 129 min Incision at completion of bolus, infusion ended at time of surgery end time 500 mL Intra-operative: 4 units FFP, 3 units PRBC, 1 unit cryo; 4F-PCC 1500 units 12 hours Apixaban 5 mg twice daily 304 min Patient received infusion at presentation, procedure was delayed. Second infusion pre-operative: incision at completion of bolus, infusion ended before end of procedure 7 hours Apixaban 5 mg twice daily 438 min Incision 19 hours after completion of infusion <18 hours Apixaban 2.5 mg twice daily 113 min Incision at completion of bolus, procedure completed prior to end of infusion 6 hours Rivaroxaban 15 mg twice daily unknown Incision at start of bolus, procedure completed prior to end of infusion 17 hours Apixaban 2.5 mg twice daily 53 min Incision 6 hours after infusion complete 2.5 hours Rivaroxaban 20 mg daily 80 min Incision at completion of bolus, procedure completed prior to end of infusion 14 hours Apixaban 2.5 mg twice daily 129 min Incision 14.5 hours after infusion complete >8 hours Rivaroxaban 20 mg daily 24 min Incision at time of bolus, procedure completed prior to end of infusion >8 hours Apixaban 2.5 mg twice daily 44 min Incision at completion of bolus, procedure completed prior to end of infusion 800 mL Intraoperative: received 3 units PRBC In the ANEXXA-4 trial, patients were not eligible if surgery or a major invasive procedure was planned within 12 hours of administration of andexanet alfa. It is unknown if andexanet alfa provides adequate duration of reversal of anticoagulation in the procedural setting due to its short half-life and limited data in this population. In the ANNEXA A and R trials the duration of suppression of FXai activity lasted approximately 4 hours, with maximal suppression achieved during the infusion. The FXai activity increased over the following 2 hours to the levels observed in volunteers receiving placebo.3 In the ANNEXA-4 trial, similar results were seen; with an observed increase in Xai activity 4 hours after the initiation of andexanet alfa treatment infusion.4 The use of andexanet alfa in patients with decreased renal function has also not been studied. Only 9% of the safety population and 11% in the efficacy population in the ANEXXA-4 trial had creatinine clearance less than 30 mL/min. Poor renal function results in slower FXai clearance; the anticoagulant activity may increase to a clinically relevant level or persist beyond the expected timeframe in these patients.4 The pharmacokinetic properties of andexanet alfa are important for prescribers to understand in the setting of potentially prolonged procedure times or during the postoperative course where rebound anticoagulation could be harmful. There are no data to support the use of andexanet alfa infusions for greater than 2 hours to maintain reversal. In our previous publication, time from order placement to bedside-delivery averaged 43 minutes, and time from order to bolus administration averaged 66 minutes.1 This presents a challenge with emergent procedures, especially due to patient movement among different departments and providers. Unlike for vitamin K antagonists, there are no FDA approved coagulation tests available to measure FXai concentrations. Although anti-Xa assays with specific FXai calibrators are recommended in guidelines, these assays are not FDA approved and can have prolonged turnaround times. Standard coagulation tests such as PT/INR, aPTT, and anti-Xa activity (calibrated to UFH/LMWH) are available but have significant variation in ability to detect clinically relevant FXai concentrations due to differences in reagent sensitivity. Although the PT can be affected by FXai, patients can have a normal PT when taking FXai despite drug levels that can impair coagulation. As the PT and aPTT do not correlate consistently with therapeutic drug levels, they should not be used to interpret degree of anticoagulation with apixaban or rivaroxaban.6-10 The anti-Xa assay (UFH/LMWH) is available at many institutions and can be used to rule out clinically significant anticoagulation to determine need for reversal. A normal result (less than 0.10 units/ml) provides reassurance that there is no clinically relevant FXai concentration, however long turnaround times may limit use in life-threatening bleeding. Elevated values should be considered as a relative measure of activity and not interpreted in a linear fashion to represent drug concentration levels.6-10 Importantly, standard clinical anti-Xa assays, such as Stago STA-Liquid Anti-FXa or HemosIL Liquid Anti-Xa, cannot be used to measure degree or efficacy of reversal during or after administration of andexanet alfa, as sample processing factors result in dissociation of andexanet alfa from FXai.11 Andexanet alfa, the first FDA approved specific reversal agent for oral FXai, is approved for reversal of anticoagulation in the setting of life-threatening bleeding. For patients undergoing surgery the suppression of FXai activity may not be adequate to maintain hemostasis throughout a longer operative procedure, or to perform a new operative procedure hours after the completion of the andexanet alfa infusion. Hematologists, emergency medical providers, surgeons, proceduralists, pharmacists, and nurses should devise multidisciplinary work flow guidance documents to determine periprocedural bleeding risks, as well as timing of andexanet alfa administration in relation to procedures to minimize bleeding risk. As new data become available we expect that our process for use of andexanet alfa in the perioperative and periprocedural setting will continue to evolve. Jean M. Connors is a consultant for Bristol-Myers Squibb, and has participated in scientific advisory boards for Pfizer and Portola. Katelyn W. Sylvester has participated on scientific advisory board for Bristol-Meyers Squibb/Pfizer and as a consultant for Portola. The remaining authors have no conflicts of interest to disclose.
No takes yet. Share an insight, caveat, or question.
Culbreth et al. (2019) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: