Abstract Number: Esoc2026lb103 Clinical and Imaging Determinants of Neurosurgical Intervention in Acute Intracerebral Hemorrhage: A Fastest Cohort Analysis

Abstract Background and aims Clinical and radiographic severity of patients with spontaneous intracerebral hemorrhage (ICH) drive neurosurgical intervention. This analysis sought to identify clinical and imaging factors associated with the decision for neurosurgical intervention in the FASTEST trial. Methods Baseline clinical and CT imaging features at the time of surgical decision-making were evaluated. Multivariable logistic regression were performed to identify predictors of neurosurgical intervention, with secondary analyses assessing rFVIIa as a modifier of surgical likelihood and post-surgical outcomes. 90-day functional outcome was assessed using the modified Rankin Scale. Results Neurosurgical intervention occurred in 54/626 (8.6%) participants. Compared with non-surgical participants, those undergoing surgery were younger (57 vs 61 years, p=0.014), had greater neurological severity (median NIHSS 17 vs 12.5, p0.001), larger baseline ICH volume (35 vs 15 mL, p0.001), and more frequent mass effect (87% vs 67%, p0.001) and midline shift (70% vs 33%, p0.001). In multivariable analysis, independent predictors of neurosurgical intervention were baseline hematoma volume (adjusted OR 2.06 per 10 mL increase, 95% CI 1.62–2.62, p0.0001) and presence of midline shift (adjusted OR 2.59, 95% CI 1.05–6.37, p=0.039). Surgical rates did not differ between rFVIIa and placebo groups (9.1% vs 8.2%, p=0.71). Post-surgical 90-day functional outcomes were similar between treatment groups (interaction p=0.41) after adjustment for baseline clinical and imaging severity. Conclusions Neurosurgical intervention for acute ICH was driven by markers of structural severity, particularly hematoma volume and midline shift. Ultra-early rFVIIa did not influence surgical decision-making or outcomes, supporting the feasibility of integrating early hemostatic therapy with standard neurosurgical care. Conflict of interest Figure 1 - belongs to Methods Figure 2 - belongs to Results