Lower mortality with andexanet alfa vs 4-factor prothrombin complex concentrate for factor Xa inhibitor-related major bleeding in a U.S. hospital-based observational study
Paul P. Dobesh, Gregory J. Fermann, Mary J. Christoph, Bruce Koch, Eva Lesén, Hungta Chen, Belinda Lovelace, Theresa Dettling, Mark D. Danese, Julie Ulloa, Sherry Danese, Craig I Coleman
Abstract
Background: Well-designed studies with sufficient sample size comparing andexanet alfa vs 4-factor prothrombin complex concentrate (4F-PCC) in routine clinical practice to evaluate clinical outcomes are limited.Objectives: To compare in-hospital mortality in patients hospitalized with rivaroxabanor apixaban-related major bleeding who were treated with andexanet alfa or 4F-PCC.Methods: An observational cohort study (ClinicalTrials.govidentifier: NCT05548777) was conducted using electronic health records between May 2018 and September 2022 from 354 U.S. hospitals.Inclusion criteria were age 18 years, inpatient admission with diagnosis code D68.32 (bleeding due to extrinsic anticoagulation), a record of use of the factor Xa inhibitors rivaroxaban or apixaban, andexanet alfa or 4F-PCC treatment during index hospitalization, and a documented discharge disposition.Multivariable logistic regression on in-hospital mortality with andexanet alfa vs 4F-PCC was performed.The robustness of the results was assessed via a supportive propensity score-weighted logistic regression.Results: The analysis included 4395 patients (andexanet alfa, n = 2122; 4F-PCC, n = 2273).There were 1328 patients with intracranial hemorrhage (ICH), 2567 with gastrointestinal (GI) bleeds, and 500 with critical compartment or other bleed types.In the multivariable analysis, odds of in-hospital mortality were 50% lower for andexanet alfa vs 4F-PCC (odds ratio [OR], 0.50; 95% CI, 0.39-0.65;P < .01)and were consistent for both ICH (OR, 0.55; [0.39-0.76];P < .01)and GI bleeds (OR, 0.49 [0.29-0.81];P = .01).Similar results were obtained from the supporting propensity scoreweighted logistic regression analyses. Conclusion:In this large observational study, treatment with andexanet alfa in patients hospitalized with rivaroxaban-or apixaban-related major bleeds was associated with