ABSTRACT NUMBER: ESOC2026A1291 LEPTOMENINGEAL COLLATERAL STATUS IS ASSOCIATED WITH BASELINE HEMATOMA VOLUME IN SUPRATENTORIAL INTRACEREBRAL HEMORRHAGE

Abstract Background and aims Intracerebral hemorrhage (ICH) is associated with high morbidity and mortality. Hematoma expansion (HE) is a major contributor to poor outcome and represents a potentially modifiable target. Indirect evidence suggests that leptomeningeal collateral status may influence HE. Methods Retrospective single-center cohort study of adult patients (≥18 years) with supratentorial ICH and available admission single-phase CT angiography treated at a tertiary care hospital between 2014 and 2023. Primary outcomes were baseline and follow-up hematoma volumes; HE was defined as an absolute (≥ 6mL) or relative (≥ 33%) increase on CT imaging within 48 hours. Secondary outcomes included neurological status at 24 hours (NIHSS) and functional outcome at 3 months (mRS). Associations between collateral status and outcomes were assessed using univariable and multivariable linear and logistic regression. Results Among 287 patients, 67.6% had good collateral status. Better LMCS was independently associated with smaller baseline hematoma volume (adjusted mean difference -13.7 mL, 95%CI -22.5 to -5.0), corresponding to an approximately 50% lower volume in log-transformed analyses. LMCS was not independently associated with binary hematoma expansion after multivariable adjustment, although continuous expansion measures suggested numerically smaller growth with better collateralization. Furthermore, LMCS was not independently associated with neurological status at 24 hours or functional outcome at 3 months after adjustment for baseline prognostic factors. Conclusions This study provides the first systematic evaluation of leptomeningeal collateral status in patients with supratentorial intracerebral hemorrhage. Leptomeningeal collateral status was independently associated with baseline hematoma volume but not with hematoma expansion or clinical outcomes after adjustment for established prognostic factors. Conflict of interest All authors: nothing to disclose Figure 1 - belongs to Methods Table 1 - belongs to Results Table 2 - belongs to Results Table 3 - belongs to Results