201 Outcomes Associated With Andexanet Use in Patients With Factor Xa Inhibitor-Associated Intracranial Hemorrhage: A One-Year Single-Center Analysis

Abstract

Patients presenting with intracranial hemorrhage in the setting of oral factor Xa (FXa) inhibitor use have a high inpatient and 30-day mortality. Andexanet received accelerated approval from the FDA in May 2018 for reversing life-threatening or uncontrolled bleeding associated with the use of rivaroxaban or apixaban. Our institution approved andexanet to formulary in June 2018 with restricted criteria, including intracranial hemorrhage. The purpose of this study is to describe patient outcomes during our first year of andexanet use for intracranial hemorrhage. This IRB-approved, retrospective, observational study included all patients admitted between June 2018 and May 2019 with a diagnosis of intracranial hemorrhage who received andexanet for FXa inhibitor reversal. Patients with spontaneous or traumatic bleed were included if they were taking apixaban, rivaroxaban or edoxaban. The primary outcome was hemostatic efficacy on repeat head computerized tomography (CT) defined as either excellent/good or poor based on the rating system adapted from the ANNEXA-4 trial. Other outcomes included time of presentation to andexanet administration, in-hospital mortality, and requirement of surgical intervention for the treatment of intracranial hemorrhage. Safety outcomes evaluated were inpatient mortality, incidence of thrombotic events during hospitalization and 30-day readmission due to thrombotic event. Twenty-seven patients were treated with andexanet for ICH during the study time period. The majority of patients were male (74%) with a median age of 81 years and a baseline GCS of 14. Apixaban was the most common FXa inhibitor reversed (70.3%), followed by rivaroxaban (26%) and edoxaban (3.7%). The majority of patients took their last FXa inhibitor dose either 8-18 hours (51.9%) or less than 8 hours (33.3%) from time of bleed identification and most patients were treated with low dose andexanet (88.8%). Median (IQR) baseline hematoma volume for intracerebral hemorrhage (ICH) was 7.4 mL (1 - 30.8), while baseline thickness for subdural hemorrhage (SDH) and subarachnoid hemorrhage (SAH) was 13 mm (7-21). Eleven (40.7%) patients experienced a spontaneous bleed [ICH 10 (91%); SDH 1 (9%)]. Sixteen (59.3%) patients experienced intracranial hemorrhage secondary to trauma [SDH/SAH 12 (75%); ICH 4 (25%)]. The median door-to-needle time was 1.24 hours. Three patients underwent hematoma evacuation prior to repeat head CT and one patient was discharged home prior to repeat CT scan; these patients were subsequently excluded from the primary outcome analysis. Of the remaining 23 patients, repeat head CT demonstrated excellent/good hemostatic efficacy in 19 (82.6%) patients and poor hemostatic efficacy in 4 (16.7%) of patients. A total of 15% of patients either expired (n=3) or were discharged to hospice (n=1). No patients experienced a thrombotic event by hospital discharge and no patients were readmitted within 30 days due to a thrombotic event. Excellent or good hemostatic efficacy was observed in 82.6% of patients treated with andexanet, which is comparable to the ANNEXA-4 trial. There were no thrombotic events and in-patient mortality was lower than reported data.