579 The Efficacy and Safety of Andexanet Alfa in Patients With Acute Gastrointestinal Bleeding While Taking Factor Xa Inhibitors: An ANNEXA-4 Sub-Analysis

Abstract

INTRODUCTION: Andexanet alfa is a recombinant modified human Factor Xa (FXa) decoy protein developed to reverse the anticoagulant effect of FXa inhibitors. METHODS: ANNEXA-4 was a prospective single-arm study of andexanet alfa in patients with acute major bleeding while taking a FXa inhibitor. Patients with major gastrointestinal (GI) bleeding were eligible if they had (1) signs/symptoms of hemodynamic compromise or (2) bleeding associated with a hemoglobin (Hb) drop ≥2 g/dL or baseline Hb ≤8 g/dL. Upon enrollment, patients were treated with one of two andexanet dosing regimens based on the type, timing, and dose of the most recent inhibitor intake. Efficacy was evaluated in all patients who met the bleeding criteria above and had a baseline anti-FXa activity ≥75 ng/mL (≥0.25 IU/mL for enoxaparin). Co-primary endpoints were the change in anti-FXa activity and the achievement of hemostasis, as adjudicated by an independent committee, at 12 hours after andexanet treatment. Hemostasis was defined as: excellent if the red blood cell transfusion-corrected Hb at 12 hours was ≤10% lower than baseline, good if Hb was 10-20% lower, and poor/none if Hb was >20% lower. RESULTS: Of 352 patients enrolled in the study, 90 had GI bleeding, of whom 62 were included in the efficacy analysis. Median age was 76 years (range 24-95); 51 (57%) were male. Patients were receiving rivaroxaban (n = 44), apixaban (n = 5), and edoxaban (n = 3) at study entry. Bleeding location was identified in 48 patients (upper n=27, lower n = 21) and unknown in 42 patients. 66 (73%) patients had some evidence of hemodynamic compromise, including 38 (42%) patients with hypotension. Mean Hb at the time of dosing was 4.9 ± 1.4 g/dL. In patients taking rivaroxaban, the baseline anti-FXa activity was 267 ± 168.5 ng/mL and the percent reduction with andexanet was 94% (95%CI 89-96). In apixaban patients, the baseline anti-FXa activity was 244 ± 215.0 ng/mL and the percent reduction with andexanet was 92% (95%CI 88-95). Excellent/good hemostasis was achieved in 85% of evaluable patients. 58 (94%) of 62 patients received red blood cell transfusions, 3 (5%) received platelet transfusion, and 2 (3%) received a coagulation factor product. 30-day rates of thrombosis and death were 6% and 13%, respectively. CONCLUSION: In patients with major GI bleeding treated with andexanet, a high rate of effective hemostasis was achieved. The thrombosis and mortality rates are consistent with prior studies in this patient population.

Anticoagulant‐associated gastrointestinal bleeding: Framework for decisions about whether, when and how to resume anticoagulants

“…Overall, there are limited data regarding all‐cause mortality and specific causes of death after major OAC‐associated GI bleeding in contemporary studies. In OAC reversal studies and observational cohorts, 30‐day all‐cause mortality after OAC‐associated major GI bleeding is 11%–13% 6,7,17,41,52 . For example, in the RE‐VERSE AD and ANNEXA‐4 studies, all‐cause mortality among patients presenting with acute major GI bleeding was 11% and 13%, respectively, despite the treatment with agents that reversed anticoagulant effect and reduced levels of active drug 6 …”

“…Among patients receiving idarucizumab for dabigatran reversal in the RE‐VERSE AD study, 30‐day and 90‐day rates of thrombotic events were 2.9% and 4.4%, respectively (Table 3); two thirds of these cases occurred in absence of OAC resumption. In the ANNEXA‐4 study assessing andexanet alfa for reversal of factor Xa inhibitors, the 30‐day risk of thromboembolism was 6% 6 . Further, Qureshi et al reported 17% risk of thromboembolism within 1 year of major GI bleeding in an AF population with median CHADS 2 score of 3 46 .…”

“…6,7,17,41,52 For example, in the RE-VERSE AD and ANNEXA-4 studies, all-cause mortality among patients presenting with acute major GI bleeding was 11% and 13%, respectively, despite the treatment with agents that reversed anticoagulant effect and reduced levels of active drug. 6 Subgroup analyses of the ROCKET AF and ENGAGE TIMI trials reported case-fatality rates (adjudicated as death due to bleeding) of 1.7% and 2.1% among patients receiving rivaroxaban or edoxaban, respectively. 5,25 Although these rates are lower than those reported among non-anticoagulated patients, they likely reflect strict inclusion criteria leading to recruitment of less medically complex patients in clinical trials than those studied in routine practice.…”

“…Nearly half of oral anticoagulant (OAC)‐associated major bleeding arises from the gastrointestinal (GI) tract, 2–5 with a higher rate of major GI bleeds among direct oral anticoagulant (DOAC) users compared to warfarin users in clinical trials and observational studies. Despite a lower case‐fatality rate compared to intracranial hemorrhage, OAC‐associated major GI bleeds are associated with substantial 30‐day mortality even among DOAC‐treated patients receiving specific reversal agents (11%–13%) despite reversal of anticoagulant effect 6,7 …”

“…Despite a lower case-fatality rate compared to intracranial hemorrhage, OAC-associated major GI bleeds are associated with substantial 30-day mortality even among DOAC-treated patients receiving specific reversal agents (11%-13%) despite reversal of anticoagulant effect. 6,7 There is substantial uncertainty about the net clinical benefit of anticoagulation after major GI bleeding episodes and, in particular,…”