ACUTE ISCHEMIC STROKE: CANCER-RELATED STROKE AND MULTIMODAL ANALYSIS OF THROMBI.

Background Acute ischemic stroke (AIS) usually results from atherosclerosis or cardiac sources, but less frequent mechanisms, such as cancer-related stroke, play a role in some cases. Studying these rarer causes of AIS is crucial for secondary prevention. Thrombi analysis provides a unique window into stroke biology. This thesis investigates (1) clinical features and outcomes of AIS in patients with active cancer (AC) and (2) histopathological and multi-omics profiles of retrieved thrombi from AIS patients with large vessel occlusion. Methods We examined AIS–cancer interactions using the Acute-Stroke-Registry-and-Analysis-of-Lausanne (ASTRAL, 2003–2024) and international collaborations, focusing on outcomes, antithrombotic and anticancer strategies, and the role of newly diagnosed vs known cancer. In parallel, we created the POLIclinico-Stroke-Thrombi-Analysis-Registry (POLISTAR) to prospectively collect thrombi with linked clinical, imaging, histological, and omics data. Histology assessed fibrin/red blood cell content and immune-cell composition. Proteomic and transcriptomic analyses explored convergent pathways and cross-platform signatures. Results Among 464 AIS patients with AC from ASTRAL, anticoagulation at discharge did not reduce recurrent AIS compared with antiplatelets. Newly diagnosed cancer doubled mortality risk (adjusted common HR 2.11, 95% CI 1.38–3.21) compared with having a known AC. Initiating or resuming cancer treatment within 3 months after AIS reduced 3-month mortality both before (aHR 0.39, 95% CI 0.28–0.56) and after (aHR 0.33, 95% CI 0.13–0.80) propensity score matching. Histology of 82 thrombi showed no significant compositional differences by clot permeability or between embolic stroke of undetermined source and cardioembolic subtypes, with 74% fibrin-prevalent. Multi-omics analyses of four thrombi revealed enrichment of platelet activation, neutrophil degranulation, complement cascades, and metabolic rewiring. Integrative proteomic–transcriptomic networks identified central hubs (FN1, GAPDH, HSP90AA1, ACTB, MYH9) linking immune and stress-response modules. Conclusions These studies show how AC, its subtypes and its therapies shape AIS prognosis. Thrombus histology alone cannot discriminate stroke etiologies, while omics highlight convergent immune, platelet, and metabolic programs. Integrating large registries, pathology, and molecular biology may advance precision medicine in AIS, especially in cancer-related and cryptogenic subgroups.