ABSTRACT NUMBER: ESOC2026A861 COGNITIVE TRAJECTORIES AFTER SMALL VESSEL DISEASE-ASSOCIATED INTRACEREBRAL HEMORRHAGE Authors: Martina Goeldlin, Christina Marti-Leistner, Noah Vonlanthen, Greta Sökeland, Elias Auer, Lorin Begré, Piotr Radojewski, David Seiffge, Janis Rauch Published: 2026 ABSTRACT AbstractBackground and aimsSmall vessel disease (SVD), the predominant cause for non-traumatic intracerebral hemorrhage (ICH), is associated with chronic, and event-related cognitive impairment. Information about cognitive trajectories after ICH is scarce, but invaluable for prognosis and care planning.MethodsIn our longitudinal, prospective, single-centre cohort including persons with SVD-associated ICH, we described cognitive function trajectories by underlying SVD subtype (cerebral amyloid angiopathy [CAA], deep perforator arteriolopathy [DPA], mixed SVD) using the CADMUS classification. We compared them with descriptive statistics and linear mixed-effects models including a random intercept for each individual and random slopes for follow-up time using Montreal Cognitive Assessment (MoCA) as primary outcome. We adjusted for follow-up time, age, sex, years of education, hypertension, and previous stroke or ICH. We performed sensitivity analyses using MoCA z-scores, and estimated group trajectories using fractional polynomials based on individuals’ trajectories.ResultsWe included 176 patients (median age 73 years (IQR 66-77), 48.3% female) with 430 available MoCAs (median 2 MoCAs per patient, IQR 1-3) over a median follow-up of 791 days (IQR 377-1165). In comparison to CAA (n = 74 patients), DPA (n = 80 patients) was associated with significantly higher MoCA scores (aCoef. 2.92, 95%-CI 1.07-4.78) and z-scores (aCoef. 0.73, 95%-CI 0.20-1.27), but mixed SVD (n = 22 patients) was not (aCoef. 1.82, 95%-CI -0.58-4.21, aCoef. z-score 0.16, 95%-CI -0.53-0.86). Individual trajectories were heterogeneous (figure 1).ConclusionsCognitive trajectories post-ICH differ according to the underlying SVD phenotype with significant inter-individual heterogeneity that should prompt further mechanistical investigations.Conflict of interestDr Goeldlin reports grants from the Swiss National Science Foundation, the Bangerter-Rhyner-Foundation, the Swiss Stroke Society, and the Insel Gruppe AG for the present study. Dr Begré’s work is supported by a Postdoc.Mobility fellowship from the Swiss National Science Foundation [P500PM_230336]; he reports unrestricted research grants from Gilead Sciences and Roche Diagnostics, and speaker honoraria from Roche, all paid to his institution and outside of the submitted work. All other authors have nothing to disclose.Figure 1 - belongs to Conclusions