Abstract Number: Esoc2026a405 Systematic Online Living Evidence Summary of Research on Cerebral Small Vessel Disease, Lacunar Infarct, and Vascular Cognitive Impairment and Dementia (Laci-Soles)

Abstract Background and aims Cerebral small vessel disease (CSVD) is increasingly recognised as a major topic in cerebrovascular research, but disease-specific treatments are currently lacking. We developed a systematic online living evidence summary of research on CSVD (LACI-SOLES), which continuously identifies and summarises relevant literature using automated techniques to provide a pipeline of potential interventions for further shortlisting and testing. Methods We developed comprehensive search strategies to identify studies on any presentation of CSVD and potential interventions across the clinical and preclinical literature. We conducted initial searches in March 2025 across PubMed, Web of Science and Scopus and established weekly search updates to continuously retrieve new records. Our automated workflow screens studies using machine-learning classifiers, which were trained on 8,200 human screening decisions and achieved specificity 80% and sensitivity 95% in the validation set. We use text mining to extract information on populations, interventions and outcomes from each study, and annotate studies for risk of bias metrics and data-sharing practices using natural language processing. The findings are presented in an interactive web application, updated weekly as new records are retrieved. Results As of 14 December 2025, our workflow retrieved 144,336 unique records, identified 9,672 relevant animal and 14,399 clinical studies, and annotated their key characteristics. Our approach identified 1,835 potential interventions, with top usage hits including aspirin, exercise, alteplase, donepezil, clopidogrel, galantamine, warfarin, memantine, nimodipine and acupuncture. Conclusions LACI-SOLES provides a living overview of CSVD-related research, including novel treatment targets and intervention candidates for clinical trials, reducing human time required for literature evaluation. Conflict of interest This project is funded by Race Against Dementia and UK Dementia Research Institute. The authors declare no conflicts of interest.