Abstract
BACKGROUND: Factor Xa inhibitor-associated intracerebral hemorrhage (ICH) requires rapid anticoagulation reversal. Although andexanet alfa, a specific FXaI antidote, demonstrated efficacy in andexenet alfa for acute intracerebral hemorrhage (ANNEXA-I) trial, it was associated with a high thromboembolic rate. Consequently, 4-factor prothrombin complex concentrate (4F-PCC) is widely used, though real-world data remain limited. OBJECTIVE: To assess the hemostatic effectiveness and safety of 4F-PCC for reversal of oral factor Xa inhibitors in patients with acute intracerebral hemorrhage. METHODS: We conducted a single-center, retrospective observational study of consecutive patients with FXaI-associated ICH treated with 4F-PCC between January 2017 and May 2025. The primary endpoint was hemostatic efficacy according to ANNEXA-I criteria: hematoma expansion < 35%, National Institutes of Health Stroke Scale (NIHSS) score increase of <7 points, and absence of rescue therapy within 12 hours. The secondary endpoint was a stable neurological status (no worsening of the NIHSS score) at 48 hours. Safety outcomes included 30-day thromboembolic events and mortality. RESULTS: Fifty-two patients (median age, 81 years; IQR, 75-87; 61.5% male) were included. Apixaban was the most frequent FXaI (86.6%), with atrial fibrillation as the main indication (94.3%). The median baseline hematoma volume was 5.45 mL (IQR, 2-21), and the NIHSS score was 4.5 (IQR, 1-6). The primary endpoint was achieved in 39 patients (75.0%; 95% CI, 61.1%-86.0%). Stable neurological status at 48 hours occurred in 37 patients (71.2%; 95% CI, 56.9%-82.9%). One thromboembolic event (deep vein thrombosis) occurred (1.9%; 95% CI, 0.0%-10.3%), and 12 patients (23.1%; 95% CI, 12.5%-36.8%) died within 30 days. CONCLUSION: 4F-PCC achieved high hemostatic efficacy and low thromboembolic risk in FXaI-associated ICH. Mortality was comparable to ANNEXA-I, but thrombotic events were markedly lower, supporting current guideline recommendations for 4F-PCC use in this setting.