Selection of Flow Modulation Protocols for Patients on Continuous Flow Ventricular Assist Devices (CF-VADs)

项目总结 连续流量式心室辅助装置(CF-VAD)的一个主要问题是由此产生的非生理性 脉动性减弱的血流已被证明是发展动静脉的主要危险因素 畸形(AVM)和胃肠道(GI)出血。为了解决这个问题，通过快速的流量调制 改变泵的叶轮速度被认为是一种引入“人工脉动”的技术。然而， 鉴于重建CF-VAD相关非手术出血事件的大型动物模型的不足， 目前尚不清楚人工脉动是否可以预防这些不良事件，也不清楚人工脉动的水平是多少 这是必要的。为了评估脉动性的影响并确定有前景的流量调节方法，我们开发了 血管脉搏灌流模型在体外培养人主动脉内皮细胞中的应用 指正常的搏动性或搏动性减弱的血流(CF-VAD支持)。我们为动脉建模的基本原理 血管是因为脉搏主要影响循环系统的动脉侧，它的影响是 由排列在大动脉血管内皮细胞转导。VPPM被验证为相关模型 通过与使用和不使用CF-VADs的患者的主动脉样本进行直接比较。我们公布的数据还显示 这种脉动性的丧失与促血管生成/炎症细胞因子的产生增加有关。这个 这些结果的相关性通过来自经历动静脉动静脉畸形和 胃肠道出血事件(与CF-VAD相关和由于其他情况)显示类似的PRO水平升高 血管生成/炎性细胞因子。因此，VPPM为评估人工智能提供了一个强大的模型 在CF-VAD流量调制的背景下的脉动性，并确定是否恢复脉压和/或脉搏 频率可以减少非手术出血事件。基于最近的研究表明，脉压 35毫米汞柱是发生胃肠道出血的主要危险因素，我们假设“搏动性减弱 与‘CF-VAD支持’相关的结果是内皮功能障碍和促炎症/促血管生成溶解 要素生产。这些变化可以通过引入使用流量调节的人工脉动来缓解 脉压保持在35毫米汞柱以下的策略“。AIM1将评估患者衍生的反应 VPPM内的内皮细胞向CF-VAD流动及血管生成/炎性可溶性因子的定量 生产，AIM2将跟踪患者长达36个月，以评估血清促血管生成/亲血管生成水平 炎性细胞因子和非手术出血事件，然后将与体外结果进行比较 VPPM和Aim3中的研究将评估不同的流量调节策略，使用患者派生的 血管内皮细胞通过比较血流动力学来确定最有希望的患者特异性方法 使用深度学习方法的概况和细胞因子生物标记物。这个项目的成功完成将 能够识别基于设备的策略，以防止接受CF-VAD支持的患者发生非手术出血。

PROJECT SUMMARY A major concern with continuous flow ventricular assist devices (CF-VADs) is the resulting non-physiological flow with diminished pulsatility which has been shown to be a major risk factor for development of arteriovenous malformations (AVMs) and gastrointestinal (GI) bleeding. To address this issue, flow modulation via rapid changes in pump impeller speed has been proposed as a technique to introduce ‘artificial pulsatility’. However, given the inadequacy of large animal models with recreating CF-VAD associated non-surgical bleeding events, it is still unclear if artificial pulsatility can prevent these adverse events or what level of artificial pulsatility is even necessary. To evaluate the effects of pulsatility and identify promising flow modulation approaches we developed a vascular pulse perfusion model (VPPM) to culture Human Aortic Endothelial Cells (HAECs) under conditions of normal pulsatile flow or flow with diminished pulsatility (CF-VAD support). Our rationale for modeling arterial vessels is because pulsatility primarily affects the arterial side of the circulatory system and its effects are transduced by endothelial cells that line the large arterial vessels. The VPPM was validated as relevant model via direct comparison with aortic samples of patients with and without CF-VADs. Our published data also shows that loss of pulsatility is associated with an increase in production of pro-angiogenic/inflammatory cytokines. The relevance of these results is further strengthened by supporting data from patients that experience AVMs and GI bleeding events (both CF-VAD related and due to other conditions) showing similar elevated levels of pro- angiogenic/inflammatory cytokines. The VPPM therefore provides a powerful model to evaluate artificial pulsatility in the context of CF-VAD flow modulation and determine if restoring pulse pressure and/or pulse frequency can mitigate non-surgical bleeding events. Based on recent studies that suggest that pulse pressure < 35 mmHg is a major risk factor for development of GI bleeds, we hypothesize that “Diminished pulsatility associated with ‘CF-VAD support’ results in endothelial dysfunction and pro-inflammatory/pro-angiogenic soluble factor production. These changes can be mitigated via introduction of artificial pulsatility using flow modulation strategies where pulse pressure is preserved at > 35 mmHg”. Aim1 will evaluate response of patient derived endothelial cells within the VPPM to CF-VAD flow and quantify angiogenic/inflammatory soluble factor production, Aim2 will follow patients for up to 36 months to evaluate serum levels of pro-angiogenic/pro- inflammatory cytokines and non-surgical bleeding events which will then be compared to results from in-vitro studies within the VPPM and Aim3 will evaluate different flow modulation strategies using patient-derived endothelial cells to determine most promising patient-specific approaches via comparison of hemodynamic profiles and cytokine biomarkers using deep learning approaches. Successful completion of this project will enable identification of device-based strategies to prevent non-surgical bleeding in patients on CF-VAD support.