3D Echocardiography to Improve Clinical Outcomes After Surgery for Ischemic Mitral Regurgitation

缺血性二尖瓣返流（IMR）是心脏瓣膜置换术后左心室（LV）重构不良的结果。 心肌梗死（MI）。由于缺乏关于最佳手术入路（瓣膜）的结论性数据， 修复与置换）心胸外科试验网络（CTSN）进行了两个多中心， 随机试验，以评估这两种手术方法的IMR的相对好处。可惜 CTSN IMR试验未确定最佳手术入路。CTSN重度IMR试验的结果 结果表明，修复组和置换组之间的左心室逆向重塑无差异。然而，在这方面， 亚组分析强调了复发性IMR的负面影响。IMR复发的频率更高 在修复组中，导致更多的心力衰竭相关不良事件。重要的是，修复患者 复发性IMR患者的LV体积没有减少，而未复发的修复患者的LV体积 降低程度上级于瓣膜置换术患者。这些结果有力地表明，一个病人- 根据术前影像学风险分层指导手术治疗的具体方法， 预测复发IMR将有助于优化手术结果。在本项目的初始资助期内 在一个项目中，我们在宾夕法尼亚大学（Penn）的研究小组证明， 术前3D超声心动图（3DE）和自定义瓣膜建模算法准确预测了 瓣膜修复后IMR复发。这次竞争性更新的目标是提供确凿的证据，证明前- 使用3DE进行基于手术风险的修复/置换分层可显著降低复发性IMR， 重要的是，改善了IMR患者的LV重塑、长期临床结局和生存率。我们提出 使用两个现有的数据集，以实现我们的预期目标迅速和有限的费用：（1）作为初步的一部分， 在一个受资助的项目中，我们招募了85名IMR患者，这些患者接受了修复前3DE检查，并接受了随访 前瞻性评估IMR复发。我们建议继续在宾夕法尼亚大学进行这种招募，以扩大我们的队列， 120例患者，允许进一步开发和验证MV后复发性IMR的最佳预测算法 修复;（2）CTSN IMR试验数据库，包括551例IMR患者，随机接受MV修复（n=276），MV 置换术（n=125）或单纯CABG（n=150）; CTSN队列中有180例接受了术前3DE。所有CTSN患者 还进行了广泛的超声心动图和长期临床随访，目前正在进行中。在目标1中，我们将建立 从候选算法开发的最佳基于3DE的预测算法的复发IMR 继续在Penn招募IMR修复患者。在目标2中，我们将评估使用理想的 目标1中关于CTSN中复发性IMR发生率和长期临床结局的预测算法 IMR试验数据库。最后，在目标3中，我们将开发一种自动3D分割技术， 二尖瓣和LV的几何建模，以允许基于风险的实时修复/置换 在手术室中对接受IMR手术的患者进行分层。

Ischemic mitral regurgitation (IMR) is a consequence of adverse left ventricular (LV) remodeling after myocardial infarction (MI). As a result of a lack of conclusive data regarding the best surgical approach (valve repair vs. replacement) the Cardiothoracic Surgical Trials Network (CTSN) conducted two multicenter, randomized trials to evaluate the relative benefits of these two surgical approaches to IMR. Unfortunately, the CTSN IMR trials did not establish the optimal surgical approach. Results of the CTSN Severe IMR trial demonstrated no difference in LV reverse remodeling between repair and replacement groups. However, subgroup analysis highlighted the negative implications of recurrent IMR. IMR recurred much more frequently in the repair group, resulting in more heart-failure related adverse events. Importantly, repair patients with recurrent IMR had no reduction in LV volume, while repair patients without recurrence experienced LV volume reduction that was superior to patients having valve replacement. These results strongly suggest that a patient- specific approach to surgical treatment guided by preoperative imaging-based risk stratification that is predictive of recurrent IMR would be useful for optimizing surgical results. During the initial funding period of this project, our group at the University of Pennsylvania (Penn) demonstrated that measures of mitral leaflet tethering derived from pre-operative 3D echocardiography (3DE) and a custom valve modeling algorithm accurately predicted the recurrence of IMR after valve repair. The goal of this competitive renewal is to provide conclusive evidence that pre- operative risk-based repair/replacement stratification using 3DE significantly reduces recurrent IMR and, more importantly, improves LV remodeling, long-term clinical outcomes and survival for patients with IMR. We propose to use two existing data sets to achieve our intended goal expeditiously and at limited expense: (1) as part of the initially funded project we have recruited 85 patients with IMR that have had pre-repair 3DE and have been followed prospectively to assess for recurrence of IMR. We propose to continue this recruitment at Penn to enlarge our cohort to 120 patients to allow further development and validation of an optimal predictive algorithm for recurrent IMR after MV repair; (2) the CTSN IMR trials data base which includes 551 IMR patients randomized to either MV repair (n=276), MV replacement (n=125) or CABG alone (n=150); 180 of the CTSN cohort have had pre-operative 3DE. All CTSN patients also have extensive echocardiographic and long-term clinical follow-up, which is ongoing. In Aim 1 we will establish the optimal 3DE-based predictive algorithm for recurrent IMR from candidate algorithms developed from continued recruitment of IMR repair patients at Penn. In Aim 2 we will assess the benefit of using the ideal predictive algorithm from Aim 1 on the incidence of recurrent IMR and long-term clinical outcomes in the CTSN IMR Trials data base. Finally, in Aim 3 we will develop a technique for automatic 3D segmentation and geometric modeling of the mitral valve and LV to allow for real-time risk-based repair/replacement stratification in the operating room for patients having surgery for IMR.