Analysis of Thrombi from Acute Ischemic Stroke Patients Using Advanced Imaging Techniques

Abstract

Background: Since cerebral thrombi are the proximate cause of acute ischemic stroke (AIS) and the primary target for current treatments, defining new thrombus characteristics associated with successful recanalization could help us improve treatment outcomes. We utilized Fourier-transform infrared spectroscopy (FTIR) and synchrotron-based X-ray fluorescence imaging (XFI) techniques, in combination with conventional histology and immunohistochemistry to analyze biochemical, elemental and histopathological composition of thrombi, respectively. Methods: AIS patients who underwent endovascular therapy from September 2017 to January 2019 were recruited. One-half of all freshly retrieved thrombi were formalin-fixed paraffin embedded and stained with H&E, Masson’s Trichrome, CD31, vWF and CD68 while the other halves were flash frozen in isopentane and analyzed using FTIR and XFI. Compositional findings were correlated with various clinical and interventional parameters collected prospectively, including National Institute of Health Stroke Scale (NIHSS), Thrombolysis in Cerebral Infarction (TICI) and modified Rankin Scale (mRS). Additionally, permeability to Isovue300 contrast agent was used as a biomarker to analyze thrombus perviousness. Results: Twenty-eight patients were included. Retrospective analysis revealed that histological composition was not associated with pre-treatment NIHSS (p = 0.19), choice of thrombectomy device (p = 0.833), 90-day mRS (p = 0.892) or 90-day NIHSS (p = 0.512). Additionally, stroke etiology did not alter histological composition of thrombi (p = 0.437). Patients in whom near-to-complete recanalization (TICI 2b-3) was achieved had thrombi composed of significantly higher proportions of erythrocytes than fibrin (p = 0.01). The choice of primary thrombectomy device, stentriever or aspiration, used during thrombus retrieval was not associated with recanalization outcomes (p = 536) or 90-day functional outcomes (mRS: p = 0.363; NIHSS: p = 0.869). XFI distribution maps demonstrated K and Fe to co-localize with erythrocytes; whereas, Ca and taurine to localize with fibrin. Lastly, measurement of thrombus permeability revealed that erythrocyte-rich thrombi were less pervious than fibrin-rich thrombi, and thrombus perviousness was associated with improved recanalization outcomes. Conclusion: XFI has served as a novel technique to identify new markers involved in poor clinical outcomes. Characterizing thrombus composition, identifying targets, and correlating these findings with clinical information are key steps in improving our understanding of stroke treatment outcomes.