Abstract Background and aims Preventing long-term recurrence after primary intracerebral hemorrhage (ICH) remains challenging, as recurrence may be ischemic (ISR) or hemorrhagic (HSR). Population-based data to guide risk stratification remain limited. We aimed to quantify long-term risks of ISR and HSR and identify clinical and radiological predictors of recurrence. Methods We analyzed data from the population-based Brest Stroke Registry (2008–2020), including patients with primary ICH. Recurrent strokes were classified as ischemic or hemorrhagic. Annual recurrence rates were estimated according to index ICH location (lobar vs non-lobar). Cox models were used to identify predictors of recurrence in the overall cohort and in a subgroup with MRI, incorporating standardized small-vessel disease markers. Results Among, 806 survivors, 118 patients (14,6%) experienced recurrence (62 ISR, 55 HSR). The annual recurrence rate was 3.2%. Recurrence patterns differed by index ICH location: lobar ICH was associated with higher HSR risk (2.2% per year) and lower ISR risk (1.1%), whereas non-lobar ICH showed higher ISR risk (2.1%) and lower HSR risk (1.0%). Cardioembolism was the leading ISR mechanism (31%). Independent predictors of recurrence were age, emboligenic cardiac arrhythmia, CT-detected leukoaraiosis, and prior carotid surgery. In the MRI subgroup, centrum semiovale perivascular spaces (HR 4.11) and cortical superficial siderosis (HR 2.61) independently predicted HSR, while lobar microbleeds (HR 2.26) and lacunes (HR 2.53) predicted ISR. Stroke recurrence after ICH is strongly influenced by index hematoma location. Integrating clinical factors with CT and MRI biomarkers enables stratification of ischemic versus hemorrhagic recurrence risk, supporting a personalized approach to secondary prevention. Conflict of interest TREGUER Olyvia : nothing to disclose
Building similarity graph...
Analyzing shared references across papers
Loading...
Treguer Olyvia
Debonnaire Mathilde
Julien Ognard
European Stroke Journal
Hôpital Maison Blanche
Building similarity graph...
Analyzing shared references across papers
Loading...
Olyvia et al. (Fri,) studied this question.
www.synapsesocial.com/papers/69fd7f86bfa21ec5bbf0806f — DOI: https://doi.org/10.1093/esj/aakag023.1274