How Clot Composition Influences Fibrinolysis in the Acute Phase of Stroke: A Proteomic Study of Cerebral Thrombi
Émilie Doche, Constance Sulowski, Jean‐Marie Guigonis, Fanny Graslin, Barbara Casolla, Jean‐François Hak, Xavier Carle, Hervé Brunel, Sabine Lindenthal, Jean‐Charles Martin, Thierry Pourcher, Laurent Suissa, Marie‐Christine Alessi, Mickaël Bobot, Abdelmalek Brinet, Sonia Dagnino, Jean-Daniel Dehondt, Silvia Di Legge, Philippe Dory‐Lautrec, Cécile Dulau-Metras, Pierre Durozard, Basile Kerleroux, Nadia Laksiri, Barbara Leclercq, P. Lehmann, Ellen Magoncia, Christine Manzac, Ophélie Osman, Roxane Peres, Sandra Pérez, Caroline Rey, Anthony Reyre, Emmanuelle Robinet-Borgomano, Salomé Snyman, Ljubica Svilar, Catherine Tardivel, Patricia Tourniaire, Sivadji Vingadalassalom
Abstract
BACKGROUND: The dramatic clinical improvement offered by mechanical thrombectomy raised questions about the relevance of prior intravenous thrombolysis in large-vessel occlusion strokes. Hence, studying intravenous thrombolysis susceptibility and its dependence on thrombus composition is crucial. We used an observational proteomic study of whole thrombi retrieved by mechanical thrombectomy to identify factors associated with fibrin content and fibrinolytic activity (FA). METHODS: In 104 stroke patients, the thrombi proteome was established by mass spectrometry coupled to liquid chromatography. FA was estimated in clots both outside (FA out ) by measuring D-dimer levels at the blood-thrombus interface and inside (FA in ) by evaluating the ratio of fibrinogen α to its plasmin-cleaved forms using proteomics coupled with protein electrophoresis. The factors associated with fibrin content, FA in , and FA out were determined by intravenous thrombolysis-adjusted linear regression. RESULTS: FA out ( P <0.0001) and FA in ( P =0.0147) were driven by recombinant tissue-type plasminogen activator (r-tPA) administration (47/104) and thrombus composition. Indeed, FA out was greater with fibrin-rich than erythrocyte-rich thrombi, presumably because of more (r)tPA substrates. Thus, FA out was increased with cardioembolic thrombi (72/104), which are rich in fibrin ( P =0.0300). Opposite results were found inside the thrombus, suggesting that (r)tPA penetrability was hampered by the density of the fibrinous cap. Moreover, blood cells had a strong impact on thrombus structure and susceptibility to (r)tPA. Indeed, fibrin content was negatively associated with erythrocyte-specific proteins in the thrombus, admission hematocrit ( P =0.0139), and hemoglobin level ( P =0.0080), which underlines the key role of erythrocytes in thrombus composition. Also, an increased number of neutrophils impaired FA out ( P =0.0225), which suggests that their aggregation around the thrombus prevented the (r)tPA attack. Only FA out was significantly associated with reduced thrombus weight ( P =0.0310), increased recanalization rate ( P =0.0150), good clinical outcome ( P =0.0480), and reduced mortality ( P =0.0080). CONCLUSIONS: Proteomics can offer new insights into the close relationship between thrombus composition and susceptibility to fibrinolysis, paving the way for new adjuvant therapies.