Predictors of Intracranial Hemorrhage Volume Expansion in Patients Receiving Factor Xa Inhibitors in ANNEXA-4: Time and Severity Matter Most

BACKGROUND: Andexanet alfa, a specific reversal agent for factor Xa inhibitors, resulted in effective hemostasis in 79% of patients with intracranial bleeding in the ANNEXA-4 (Andexanet Alfa, a Novel Antidote to the Anticoagulation Effects of Factor Xa Inhibitors) trial (NCT02329327). However, little is known about predictors associated with hematoma expansion in patients with factor Xa inhibitor-associated intracranial hemorrhage (ICrH) receiving andexanet alfa. METHODS: The ANNEXA-4 trial was a prospective, single-arm, open-label study of andexanet alfa in patients with acute major bleeding within 18 hours after taking a factor Xa inhibitor. Hematoma volumes at baseline and 12 hours after andexanet alfa treatment were measured using a computerized-assisted volumetric method. Univariable and multivariable logistic regression analyses of clinical and nonclinical parameters were performed to identify factors predictive of different volumes of hematoma expansion. To this end, an ICrH Expansion Scale was developed. RESULTS: Overall, 305 ANNEXA-4 study patients with baseline and follow-up imaging were included, 15.7% of whom showed evidence of any ICrH expansion >= 6-mL. Patients with >= 6-mL ICrH expansion had a significantly (P<0.05) higher proportion of ICrH with multiple compartment involvement (36% versus 14.3%); shorter times from symptom onset to baseline computed tomography (median, 1.6 hours [interquartile range (IQR), 1.2-4.3 hours] versus 3.7 hours [IQR, 1.6-7.0 hours]); lower Glasgow Coma Scale scores (14 [IQR, 12-15] versus 15 [IQR, 14-15]); higher systolic blood pressure 15 minutes before andexanet alfa bolus (mean, 151.6 mm Hg [SD, 24.1 mm Hg] versus 143.3 mm Hg [SD, 22.3 mm Hg]); and larger median baseline ICrH volumes (29.3 mL [IQR, 13.3-50.8 mL] versus 8.6 mL [IQR, 2.1-22.4 mL]). Multivariable analysis confirmed shorter symptom onset-to-computed tomography time and larger ICrH volume as independent predictors of >= 6-mL growth and ICrH Expansion Scale change. Lower Glasgow Coma Scale showed a trend (P = 0.06) as an independent predictor of >= 6-mL growth but was an independent predictor of ICrH Expansion Scale change. CONCLUSION: Shorter time from symptom onset to computed tomography, larger hematoma volumes, and lower Glasgow Coma Scale score at presentation increased the risk of ICrH expansion in patients with factor Xa inhibitor-associated ICrH treated with andexanet alfa.