Predictive modeling of acute rejection in pediatric heart transplant recipients

项目概要/摘要 尽管在儿科心脏移植（PHTx）患者的护理方面取得了重大进展，但急性排斥反应（AR） 仍然是导致死亡的主要原因之一。心导管检查和肌内膜活检（活检）是 诊断AR的护理标准，并在临床怀疑AR时或在常规治疗期间进行 监视不幸的是，活检是侵入性的，并与潜在的风险相关，包括： 麻醉或镇静、瓣膜损伤、传导系统损伤、血管损伤或闭塞，以及 心脏穿孔这些潜在并发症在儿科人群中被放大。无创 检测AR的方法，如血液生物标志物和心脏磁共振成像（CMR）， 减少活检的频率。血液生物标志物，如供体部分无细胞DNA和microRNA， 显示出诊断AR的潜力，但尚未在PHTx中获得广泛采用。先进 CMR参数标测序列量化心肌纤维化和水肿，我们的初步数据 提示这些序列诊断AR潜力。虽然CMR参数映射具有显著的 承诺，简单地关注整个左心室平面或区域的平均特性忽略了 疾病的空间模式，导致信息丢失和使用成像数据的能力受损， 直接护理。在这里，我们提出了比平面分析更精细的高级图像分析方法， 包括纹理分析，作为客观分析心肌疾病的不同模式的手段， 开发预测模型，以改进临床决策。这个问题的核心假设是 格兰特认为，非侵入性心脏磁共振和血液生物标志物可以检测心肌 与儿科心脏移植受者的急性排斥反应一致的异常，可以预测需要 用于心内膜心肌活检。为了解决这一假设，目标1将制定和验证一个全面的 用于识别疑似AR并需要心导管插入术的PHTx接受者的预测模型。 目标2将评估血液生物标志物是否改善目标1中开发的CMR模型。子目标将 包括成本评估，以确定最具成本效益的筛选方案。Aim 3将扩展建模 确定组织学定义的AR严重程度。本多PI提案是一项前瞻性、多中心研究 对同时接受临床活检的PHTx伴和不伴AR患者进行CMR。的创新之处 研究是使用先进的CMR，纹理分析和血液生物标志物进行非侵入性检测， AR.该提案充分利用了美国儿科学会先天性/儿科研究委员会的支持， 心血管磁共振（SCMR）。将这些数据应用于临床实践可以提高 生活质量和降低相关的发病率，确保只有高概率的患者， 排异进行活检。

PROJECT SUMMARY/ABSTRACT Despite significant advances in the care of pediatric heart transplant (PHTx) patients, acute rejection (AR) remains one of the leading causes of death. Cardiac catheterization with endomyocardial biopsy (biopsy) is the standard of care for diagnosing AR and is performed when there is a clinical suspicion for AR or during routine surveillance. Unfortunately, biopsy is invasive and associated with potential risks, including: complications from anesthesia or sedation, valve damage, injury to the conduction system, vascular damage or occlusion, and cardiac perforation. These potential complications are magnified in the pediatric population. Non-invasive methods of detecting AR, such as blood biomarkers and cardiac magnetic resonance imaging (CMR), could decrease the frequency of biopsy. Blood biomarkers, such has donor fraction cell-free DNA and microRNA, have shown potential for diagnosis of AR but have not yet gained widespread adoption in PHTx. Advanced CMR parametric mapping sequences quantify myocardial fibrosis and edema, and our preliminary data suggest a potential for these sequences to diagnose AR. While CMR parametric mapping has significant promise, focusing simply on the average properties across an entire left ventricular plane or region ignores the spatial patterns of disease, resulting in a loss of information and an impaired ability to use the imaging data to direct care. Here we propose advanced image analysis methods that are more granular than plane analysis, including texture analysis, as a means for objectively analyzing different patterns of myocardial disease and developing predictive models that would allow improved clinical decision making. The central hypothesis of this grant is that non-invasive cardiac magnetic resonance and blood biomarkers can detect myocardial abnormalities consistent with acute rejection in pediatric heart transplant recipients and can predict the need for endomyocardial biopsy. To address this hypothesis, Aim 1 will develop and validate a comprehensive predictive model for identifying PHTx recipients having suspected AR and requiring cardiac catheterization. Aim 2 will evaluate whether blood biomarkers improve the CMR model developed in Aim 1. SubAims will include assessment of cost to determine the most cost-efficient screening protocol. Aim 3 will expand modeling to determine severity of AR as defined histologically. This multi-PI proposal is a prospective, multicenter study to perform CMR in PHTx with and without AR who are also undergoing clinical biopsy. The innovation of this study is the use of advanced CMR, texture analysis, and blood biomarkers for the non-invasive detection of AR. This proposal leverages the support of the Congenital/Pediatric Research Committee within the Society of Cardiovascular Magnetic Resonance (SCMR). Application of these data to clinical practice could improve quality of life and decrease associated morbidity by ensuring that only patients with a high probability of rejection undergo biopsy.