Portal de Periódicos da CAPES

Intraoperative Off-Label Reversal of Apixaban by Andexanet Alfa while on VA-ECMO Immediately After Emergent Surgery for Acute Type A Aortic Dissection

2020; Elsevier BV; Volume: 35; Issue: 1 Linguagem: Inglês

10.1053/j.jvca.2020.08.017

1532-8422

Matthias Kainz, Paul Bsuchner, Peter Schellongowski, Martin Dworschak,

Cardiac, Anesthesia and Surgical Outcomes

The authors report a case of intraoperative reversal of apixaban with andexanet alfa in a patient supported with venoarterial extracorporeal membrane oxygenation due to low- cardiac-output immediately after surgery for acute type A aortic dissection and massive intraoperative transfusion with administration of procoagulants. In this patient, andexanet alfa's off-label use was not associated with acute thrombotic complications despite being given during extracorporeal life support and after previous administration of prothrombin complex concentrates. The authors report a case of intraoperative reversal of apixaban with andexanet alfa in a patient supported with venoarterial extracorporeal membrane oxygenation due to low- cardiac-output immediately after surgery for acute type A aortic dissection and massive intraoperative transfusion with administration of procoagulants. In this patient, andexanet alfa's off-label use was not associated with acute thrombotic complications despite being given during extracorporeal life support and after previous administration of prothrombin complex concentrates. PRESCRIPTIONS OF direct oral anticoagulants are increasing constantly. It can be estimated that about 3% of patients anticoagulated with direct oral anticoagulants at some point will have to undergo emergent procedures.1Garcia D Alexander JH Wallentin L et al.Management and clinical outcomes in patients treated with apixaban vs warfarin undergoing procedures.Blood. 2014; 124: 3692-3698Crossref PubMed Scopus (114) Google Scholar Andexanet alfa, a recombinant factor Xa without intrinsic activity, was US Food and Drug Administration–approved in 2018 for the reversal of direct oral factor Xa inhibitors (eg, apixaban and rivaroxaban) by high-affinity binding in life-threatening bleeding. Its elimination half-life is 5-to-7 hours, and its volume of distribution approximately 5 L.2Heo YA Andexanet alfa: First global approval.Drugs. 2018; 78: 1049-1055Crossref PubMed Scopus (65) Google Scholar Unfortunately, the andexanet alfa trials that led to the approval of this drug excluded patients requiring urgent or emergent surgery, in which case prothrombin complex concentrates should be considered as first-line agents.3Culbreth SE Rimsans J Sylvester K et al.Andexanet alfa–The first 150 days.Am J Hematol. 2019; 94: E21-E24Crossref PubMed Scopus (18) Google Scholar Some patients, however, already have been treated with andexanet alfa before surgery that could not be delayed.4Culbreth SE Sylvester KW Rimsans J et al.Coordinating emergent procedures after andexanet alfa.Am J Hematol. 2019; 94: E278-E282Crossref PubMed Scopus (4) Google Scholar Still, there is currently no information available on either safety issues or efficacy related to the off-label use of andexanet alfa when given after prior administration of prothrombin complex concentrate, which has been discouraged so far, and in patients on venoarterial extracorporeal membrane oxygenation (VA-ECMO). The authors report a 69-year-old man (body weight: 162 kg, height: 189 cm) with a history of hypertension and chronic atrial fibrillation who had to undergo emergent surgery because of type A aortic dissection while on apixaban (5 mg twice a day), of which the last tablet was taken 6 hours before surgery. Apixaban's time to peak effect is about 1-to-2 hours. The patient reported onset of chest pain and shortness of breath 15 hours before hospital admission and worsening of symptoms 5 hours before he called an ambulance. His computed tomography angiography scan from a peripheral infirmary showed type A aortic dissection with extension to the level of the iliac arteries, whereby the left renal artery was perfused via the false lumen. A transthoracic echocardiograph performed immediately before transfer revealed moderate pericardial effusion. On presentation at the authors' institution, the patient had no apparent neurologic dysfunction. He was alert and fully oriented, with a slightly elevated respiratory rate and an oxygen saturation of 94% on 6 L of O2/min via facemask. His heart rate was 110/min while on 0.25 µg/kg/min of norepinephrine support. Due to the high immediate mortality of untreated type A aortic dissection, surgery could not be delayed to await natural clearance of apixaban.5Bonser RS Ranasinghe AM Loubani M et al.Evidence, lack of evidence, controversy, and debate in the provision and performance of the surgery of acute type A aortic dissection.J Am Coll Cardiol. 2011; 58: 2455-2474Crossref PubMed Scopus (146) Google Scholar During rapid-sequence induction, hemodynamics deteriorated quickly. Pericardial tamponade necessitated emergent sternotomy. Evacuation of the effusion, together with continuous administration of vasopressors, bought time for the surgeon to anastomose a vascular graft onto the subclavian artery for arterial cannulation. Venous access was attained via the right femoral vein. Before going on cardiopulmonary bypass, 47,500 units of unfractionated heparin were administered to achieve an activated clotting time >450 seconds. During 74 minutes of deep hypothermic circulatory arrest with antegrade cerebral perfusion, an ascending aorta and a hemi-aortic arch replacement was carried out. Due to a secondary expansion of the dissection into the aortic root that only became apparent after release of the aortic clamp, a subsequent Bentall procedure had to be performed. A third aortic cross-clamp was necessary for repair of a tear in the pulmonary artery. After 494 minutes on cardiopulmonary bypass, the patient could not be weaned due to low- cardiac-output despite high-dose pharmacologic support. A VA-ECMO with heparin-coated tubing was implanted using the same cannulation sites as for cardiopulmonary bypass, and systemic heparin was reversed with protamine at a 1:1 dose ratio followed by the attempt to correct hemostasis. Despite the patient whose core and peripheral temperature was ranging between 36.5°C and 37°C having received 4.1 L of cell-saver blood, 13 units of packed red blood cells, 17 units of fresh frozen plasma, 4 units of platelet concentrates, 7 g of fibrinogen, 3,500 IU of prothrombin complex concentrate (PCC), 30 µg of desmopressin, and 100 mL of 10% calcium gluconate in total, nonsurgical bleeding persisted to such an extent that andexanet alfa (800 mg continuous infusion over 30 minutes followed by 960 mg over 120 minutes) was administered as a last resort. Within 5 minutes after the initiation of the bolus infusion, coagulation and blood loss, as assessed clinically and semiquantitatively, improved dramatically. The intraoperative time course is shown in Figure 1, and the coagulation profiles before and immediately after surgery are depicted in Table 1. After the 12-hour procedure, the patient was transferred to the intensive care unit with an open sternum, on low-dose inotropic and moderate pressor support, and an ECMO flow of 3.5 L/min. In the immediate postoperative period, the patient showed no signs of bleeding. Within the first 24 hours after surgery only 1 unit of packed red blood cells had to be given, while cumulative drainage from the chest tubes was a mere 360 mL. Low-dose unfractionated heparin was begun on postoperative day (POD) 1. On the same day, hemodiafiltration had to be initiated due to acute renal failure. On POD 3, a sternal vacuum-assisted closure system was implanted. The patient could be weaned from ECMO the following day. Inotropic support was discontinued on POD 10 and the sternum was closed on POD 15, together with insertion of a permanent pacemaker. Due to prolonged respiratory weaning, surgical tracheostomy was performed on POD 18, with vasopressors being discontinued the next day. Postoperative cranial computed tomography scan showed small infarctions in the right frontal lobe and the right cerebellum that were described as rather old injuries. When awake, the patient was not able to move both of his legs and remained paraplegic. A magnetic resonance imaging scan of the spinal cord showed ventral ischemic damage below the level of the seventh thoracic vertebra compatible with an anterior spinal artery syndrome. The patient, who iwas of sound mind, gave written informed consent for the publication of his clinical course.Table 1Coagulation Profile Before Surgery and Postoperatively After Massive Transfusion, Replacement of Coagulation Factors, and Administration of Andexanet AlfaLaboratory Test (unit)PreoperativePostoperativeReference RangePlatelets (G/L)12554150-350PT (%)456370-125INR1.51.2aPTT (sec)386827-41Anti-Xa LMWH (IU/mL)1.870.34<0.10Fibrinogen (mg/dL)302197200-400INTEM CT (sec)188359100-240INTEM CFT (sec)11323430-110INTEM α-angle (°)685444-66Abbreviations: anti-Xa LMWH, anti-factor Xa activity for low–molecular-weight heparin; aPTT, activated partial thromboplastin time; CFT, clot formation time; INR, international normalized ratio; INTEM CT, INTEM clotting time (thromboelastometry study); PT, prothrombin time. Open table in a new tab Abbreviations: anti-Xa LMWH, anti-factor Xa activity for low–molecular-weight heparin; aPTT, activated partial thromboplastin time; CFT, clot formation time; INR, international normalized ratio; INTEM CT, INTEM clotting time (thromboelastometry study); PT, prothrombin time. The authors' observation demonstrated that even after emergent, prolonged, and complicated aortic surgery comprising deep hypothermic cardiac arrest, high-dose andexanet alfa, given after inadequate conventional hemostatic treatment, could reverse a sustained anticoagulant effect immediately, presumably due to prolonged apixaban clearance. Despite earlier administration of PCC, this management caused neither hemolysis nor thrombus formation within the oxygenator and the tubing of the VA-ECMO. The sole replacement of coagulation factors, specifically PCC, which generally is recommended for situations like these, was not capable to reinstitute sufficient hemostasis. In this patient, early reversal, as described by Culbreth et al.,3Culbreth SE Rimsans J Sylvester K et al.Andexanet alfa–The first 150 days.Am J Hematol. 2019; 94: E21-E24Crossref PubMed Scopus (18) Google Scholar was not possible in the hemodynamically unstable situation the authors encountered soon after anesthesia induction requiring stabilizing hemodynamics and going on bypass quickly. Additionally, andexanet alfa and PCC when given preoperatively can lead to an unpredictable coagulation status and heparin resistance due to the fact that andexanet alfa also reverses indirect factor Xainhibitors. This interaction requires excessive doses of heparin to reach the targeted activated clotting time on the heart-lung machine, which has been reported when apixaban reversal was begun at the time of surgical incision.6Flaherty D Connors JM Singh S et al.Andexanet alfa for urgent reversal of apixaban before aortic surgery requiring cardiopulmonary bypass: A case report.A A Pract. 2019; 13: 271-273Crossref PubMed Google Scholar Using a high-dose andexanet alfa regimen in a situation in which approximately 20 U/kg of PCC and other procoagulants already had been administered is, however, not without risks—not to mention the horrendous costs (approximately $50,000) associated with it. Postrepletion hypercoagulation, with consecutive thromboembolic complications, is a serious concern because patients treated with direct factor Xa inhibitors still have normal circulating factor X.7Gibler WB Racadio JM Hirsch AL Roat TW Management of severe bleeding in patients treated with oral anticoagulants: Proceedings monograph from the Emergency Medicine Cardiac Research and Education Group-International Multidisciplinary Severe Bleeding Consensus Panel October 20, 2018.Crit Pathw Cardiol. 2019; 18: 143-166Crossref PubMed Scopus (12) Google Scholar Nevertheless, the authors opted in favor of it because it was considered a rescue therapy during life-threatening bleeding and the risk of insufficient control of bleeding was judged as clearly outweighing the risk of thrombosis at that point. The authors further assumed that apixaban clearance might be impeded due to insufficient kidney perfusion and previous deep hypothermia. Postoperatively, the anti-Xa level (calibrated for low–molecular-weight heparins) still was elevated but had dropped significantly (Table 1), while plasmatic coagulation still was not normalized completely. Unfortunately, spinal infarction is a feared complication of aortic dissection and surgery of the thoracic aorta.5Bonser RS Ranasinghe AM Loubani M et al.Evidence, lack of evidence, controversy, and debate in the provision and performance of the surgery of acute type A aortic dissection.J Am Coll Cardiol. 2011; 58: 2455-2474Crossref PubMed Scopus (146) Google Scholar Paraplegia in this patient, with pronounced atherosclerosis, most likely was attributable to impaired spinal perfusion. However, the authors cannot rule out that it may have been aggravated by embolic or thrombotic luminal narrowing of spinal arteries under pre-existing flow limitations, which could result from the administration of andexanet alfa, together with other procoagulants.8Andexxa - An antidote for apixaban and rivaroxaban.JAMA. 2018; 320: 399-400Crossref PubMed Scopus (8) Google Scholar In a porcine polytrauma model, however, andexanet alfa merely antagonized the anticoagulant effect of apixaban without causing overactivation of coagulation and fibrinolysis.9Grottke O Braunschweig T Rossaint R et al.Transient or extended reversal of apixaban anticoagulation by andexanet alfa is equally effective in a porcine polytrauma model.Br J Anaesth. 2019; 123: 186-195Abstract Full Text Full Text PDF PubMed Scopus (4) Google Scholar Anticoagulation, with direct factor Xa inhibitors, does not require coagulation monitoring. However, as viscoelastic bedside tests and anti-Xa studies calibrated for low–molecular-weight heparins provide only limited information about the degree of anti-Xa activity, specific anti-Xa assays still should be available routinely to optimize treatment of patients having to undergo emergent surgery or those with major spontaneous bleeding.7Gibler WB Racadio JM Hirsch AL Roat TW Management of severe bleeding in patients treated with oral anticoagulants: Proceedings monograph from the Emergency Medicine Cardiac Research and Education Group-International Multidisciplinary Severe Bleeding Consensus Panel October 20, 2018.Crit Pathw Cardiol. 2019; 18: 143-166Crossref PubMed Scopus (12) Google Scholar None.