Development and Validation of a Flow-Dependent Endothelialized 3D Model of Intracranial Atherosclerotic Disease

Translational Stroke Research https://doi.org/10.1007/s12975-024-01310-4 RESEARCH Development and Validation of a Flow‑Dependent Endothelialized 3D Model of Intracranial Atherosclerotic Disease Grace Prochilo1 · Chuanlong Li1 · Eleni Miliotou1 · Russell Nakasone1 · Alissa Pfeffer1 · Charles Beaman1,2 · Naoki Kaneko2 · David S. Liebeskind1 · Jason D. Hinman1,3 Received: 26 July 2024 / Revised: 31 October 2024 / Accepted: 7 November 2024 © The Author(s) 2024 Abstract Intracranial atherosclerotic disease (ICAD) is a major cause of stroke globally, with mechanisms presumed to be shared with atherosclerosis in other vascular regions. Due to the scarcity of relevant animal models, testing biological hypotheses specific to ICAD is challenging. We developed a workflow to create patient-specific models of the middle cerebral artery (MCA) from neuroimaging studies, such as CT angiography. These models, which can be endothelialized with human endothelial cells and subjected to flow forces, provide a reproducible ICAD model. Using imaging from the SAMMPRIS clinical trial, we validated this novel model. Computational fluid dynamics flow velocities correlated strongly with particle-derived flow, regardless of stenosis degree. Post-stenotic flow disruption varied with stenosis severity. Single-cell RNA-seq identified flowdependent endothelial gene expression and specific endothelial subclusters in diseased MCA segments, including upregulated genes linked to atherosclerosis. Confocal microscopy revealed flow-dependent changes in endothelial cell proliferation and morphology in vessel segments related to stenosis. This platform, rooted in the specific anatomy of cerebral circulation, enables detailed modeling of ICAD lesions and pathways. Given the high stroke risk associated with ICAD and the lack of effective treatments, these experimental models are crucial for developing new ICAD-related stroke therapies. Keywords Stroke · Intracranial atherosclerosis · Endothelia · Blood flow · Computational fluid dynamics Introduction Intracranial atherosclerosis disease (ICAD) is a highly prevalent chronic disease that is a leading cause of stroke worldwide [1, 2]. Mechanical stenting of ICAD lesions has repeatedly failed clinical trials [3–5], indicating that mechanical disruption of lesions is insufficient to reduce stroke risk. Despite its high prevalence and causative role in ischemic stroke, the pathophysiologic mechanisms of ICAD are incompletely * Jason D. Hinman 1 Department of Neurology, David Geffen School of Medicine, Gordon Neuroscience Research Building, The University of California, 635 Charles E. Young Dr. South, Room 415, Los AngelesLos Angeles, CA, USA 2 Department Radiology, David Geffen School of Medicine, The University of California, Los Angeles, Los Angeles, CA, USA 3 Department of Neurology, Department of Veterans Affairs Medical Center, Los Angeles, CA, USA understood. Parallels to atherosclerosis in other circulatory systems may be relevant. However, drawing these parallels makes several assumptions that disregard the distinct anatomy and rates of blood flow in the proximal cerebral vessels that create a unique inter-relationship between the biophysical forces of flow and the endothelial surface in ICAD lesions. This is clear when considering the differential patterns of recurrent stroke in ICAD, including distal cerebral hypoperfusion accounting for 22.7% of stroke recurrence as indicated by a border zone infarct pattern while distal thromboembolic strokes account for 27.3% as indicated by a cortical or territorial pattern [6]. Moreover, the rates of ICAD progression and impact of the degree of stenosis do not drive rates of ischemic events as in the coronary or carotid circulations [7, 8]. Therefore, the development of experimental models of ICAD that can recapitulate the anatomic features of human ICAD lesions with the flow characteristics driven by the lesion and endothelial cell biology can inform the underlying vascular biology that may advance this field beyond stenting. Various models animal and in vitro models of atherosclerosis exist [9]. All existing models have limitations that range Vol.:(0123456789) Translational Stroke Research from the lack of characteristic atherosclerotic plaques in rodent models to the use of artificial stenoses in vitro that do not mimic the anatomic features of ICAD lesions seen in patients. We previously developed an endothelialized 3D flow model of intracranial stenosis using patient-derived angiographic imaging and used it to identify the flow-responsive nature of angiotensin-converting enzyme 2 (ACE2) in the context of binding to Sars-CoV2 spike protein [10]. While this model provides a unique platform to study how stenosis-related flow forces drive focal changes in endothelial biology, the selection of patient-derived images as a source for the model can be biased. To establish a true representative experimental model of ICAD, case selection should include lesions that have a proven role in driving stroke. This has been a key limitation of animal models of ICAD that often develop fatty deposits and, in some cases, characteristic atherosclerotic lesions but infrequently develop localized thrombosis or embolism leading to stroke, as in ICAD patients [9]. The Stenting and Aggressive Medical Management for Preventing Recurrent Stroke in Intracranial Stenosis (SAMMPRIS) trial included patients with symptomatic intracranial atherosclerosis and tested the role of mechanical stent placement to reduce stroke risk [3]. The design of SAMMPRIS labels ICAD lesions as symptomatic and, when considering the MCA, enables a withinsubject comparison vessel that is asymptomatic regardless of whether it harbors a stenosis. We sought to take advantage of the SAMMPRIS imaging dataset to determine if our endothelialized 3D flow model can represent flow features of ICAD as suggested by computational fluid dynamics (CFD) and expand the biologic measures available in this model system. Here, we show the successful development and validation of an experimental model of intracranial atherosclerosis derived from the definitive clinical trial in ICAD. In MCA segments from SAMMPRIS, we mapped CFD flow profiles and experimentally validated these features in vitro. Additionally, we demonstrate the applicability of single-cell RNA-sequencing (scRNA-seq) as well as confocal microscopy to illustrate the unique endothelial cell biology driven by ICAD-related flow disruptions in this 3D flow system. While we sought here only to introduce and validate this model of ICAD, this powerful approach may be used across the variety of ICAD lesions to establish novel biophysical relationships between flow and endothelial biology relevant to stroke risk due to ICAD. Methodology Source Imaging All source imaging was derived from the SAMMPRIS trial [3]. Baseline computed tomography angiography (CTA) images from the aggressive medical management (MM) and aggressive (...truncated)