Abstract DP282: CTA Spot Sign and Response to Andexanet in FXai-associated Intracerebral Hemorrhage: a Secondary Analysis of ANNEXa-I

Introduction: The spot sign identified on CT angiography (CTA), is a well-established predictor of hematoma expansion (HE), but its role in FXai-related intracerebral hemorrhage (ICH) remains poorly explored. We investigated the association between spot sign and HE in patients with acute FXai-associated ICH and to assess whether its presence modified the treatment effect of andexanet alfa. Methods: Participants with a qualifying ICH on baseline brain imaging and available CTA at baseline were included in this post-hoc analysis of the ANNEXa-I trial. Imaging analyses assessed the presence of spot sign and its features (e.g., number, maximum diameter, maximum density, co-localization with hypodensity). HE was defined as a hematoma volume increase of ≥12.5 mL or ≥35% between baseline and 12 hours after randomization. The association between spot sign presence and its characteristics with HE was assessed with logistic regression analysis adjusted for time from symptom onset to scan, baseline hematoma volume, and diastolic blood pressure. The treatment effect modification of andexanet on HE by these imaging variables was assessed with a multivariable regression analysis. Results: Of the 530 patients in the intention-to-treat population, 156 had CTA performed and were included in this study (66 [42.3%] female, mean [ + SD] age 78.5 [±7.8]) ( Table 1 ). HE was observed in 55 (35.7%) patients and 31 (20.1%) patients had a spot sign. Spot sign presence (adjusted odds ratio [aOR] 4.24 [95% CI=1.55-11.56]), as well as some of its features, such as diameter >5 mm (aOR 5.40 [95% CI=1.39-20.97]), central location (aOR 7.76 [95%CI=1.63-37.03]), and co-localization with a hypodensity (aOR 6.56, [95%CI=1.46-29.50]) were associated with HE following multivariable adjustment ( Table 2 ). HE occurred in 36 (29.3%) patients with no spot sign (andexanet 20% vs. usual care 38.1%; NNT with andexanet to prevent HE: 5.6) and in 19 (61.3%) with a spot sign (andexanet 45% vs. usual care 90.9%; NNT with andexanet to prevent HE: 2.2). No significant heterogeneity in the proportional reduction of HE with andexanet was observed based on spot sign presence (p-interaction=0.150) or its characteristics ( Figure 1 ). Conclusions: The presence of the spot sign and its features were significantly associated with HE in FXai-associated ICH. The treatment effect of andexanet was still consistent regardless of the presence of the spot sign and its features