Abstract MP39: Modeling the Clinical Implications of Andexanet Alfa in Factor Xa Inhibitor-Associated Intracerebral Hemorrhage

Abstract

\n Background:\n Andexanet alfa was recently approved as a reversal agent for the Factor Xa inhibitors (FXai) apixaban and rivaroxaban, but its impact on long-term outcomes in FXai-associated intracerebral hemorrhage (ICH) is unknown.\n \n \n Methods:\n We performed a simulation study to determine the potential clinical implications of andexanet alfa across a range of possible hemostatic effects using data from a single-center cohort of ICH patients who did not receive the drug. We used this data to determine the baseline probability of insufficient hemostatic efficacy (IHE) across patients with and without FXai use via k-fold cross-validated multivariable regression models, which we aggregated into an IHE propensity score. We then determined the probability of unfavorable 3-month outcome (modified Rankin Scale 4-6) using a model comprised of established clinical predictors and IHE propensity. We applied model parameters from this derivation cohort to simulate a range of IHE reductions and corresponding outcomes, which we used to calculate absolute risk reduction (ARR) and projected number needed to treat (NNT) to prevent one unfavorable outcome.\n \n \n Results:\n Training models using a real-world ICH cohort (n=604 total; 55 FXai patients) had good accuracy in predicting IHE (AUC 0.78) and unfavorable outcome (AUC 0.82). IHE was strongly associated with unfavorable outcome (OR 6.7, 95% CI 3.8-11.8) and occurred in 11.4% of FXai patients. Predicted ARR of unfavorable outcome was 5% (95% CI 3-8%) at one-third reduction of IHE and 8% (95% CI = 4-13%) at 50% IHE reduction, translating to a projected NNT of 20 (cumulative treatment cost $495,000) and 13 ($321,750), respectively.\n \n \n Conclusion:\n Even optimistic simulated hemostatic effects suggest that the costs and potential benefits of andexanet alfa should be carefully considered, and placebo-controlled randomized trials are needed before its use can definitively be recommended.\n