This study focuses on direct factor Xa inhibitors (apixaban, edoxaban, rivaroxaban) and the thrombin inhibitor dabigatran, commonly used for stroke prevention in atrial fibrillation. Despite lower intracranial bleeding risks with these drugs, around 0.2-1.0% of patients annually experience intracranial hemorrhage (ICH), predominantly intracerebral. Treatment options for factor-Xa inhibitor-associated ICH, such as prothrombin complex concentrate (PCC) and andexanet alfa, lack direct comparison evidence except for ongoing trials like ANNEXA-I. This trial assesses hemostatic efficacy and 30-day functional outcomes but leaves gaps regarding anticoagulant activity's role and long-term effects, especially in patients presenting late after drug intake. The measurement of anti-FXa levels helps guide decisions, yet their link to hematoma expansion remains unknown. Efforts to streamline measurement within 30 minutes for acute decisions have shown variability in levels, with some patients exhibiting high levels even beyond 12 hours post-intake. This lack of data poses challenges, particularly for patients potentially benefiting from treatment beyond the current strict time window. Early hematoma expansion strongly predicts poor outcomes, but preventing it faces challenges like recurrent events (up to 5% by 3 months) and rehabilitation intensity, potentially negating its benefits. The ANNEXA-I trial evaluates short-term outcomes, highlighting the need for additional data to comprehend long-term ICH prognosis. The study's objectives involve linking hematoma expansion to anti-FXa levels, determining late-presenting patients' risk of expansion, and identifying predictors of favorable outcomes at 3, 6, and 12 months. Primary endpoints include functional outcomes, while secondary ones encompass expansion rates, anticoagulant activity, and various events at 12 months. This research aims to bridge gaps in understanding factor-Xa inhibitor-related ICH, addressing both immediate and prolonged outcomes to enhance clinical decision-making.