Data_Sheet_1_Analysis of Cardiac Amyloidosis Progression Using Model-Based Markers.pdf

Deposition of amyloid in the heart can lead to cardiac dilation and impair its pumping ability. This ultimately leads to heart failure with worsening symptoms of breathlessness and fatigue due to the progressive loss of elasticity of the myocardium. Biomarkers linked to the clinical deterioration can be crucial in developing effective treatments. However, to date the progression of cardiac amyloidosis is poorly characterized. There is an urgent need to identify key predictors for disease progression and cardiac tissue function. In this proof of concept study, we estimate a group of new markers based on mathematical models of the left ventricle derived from routine clinical magnetic resonance imaging and follow-up scans from the National Amyloidosis Center at the Royal Free in London. Using mechanical modeling and statistical classification, we show that it is possible to predict disease progression. Our predictions agree with clinical assessments in a double-blind test in six out of the seven sample cases studied. Importantly, we find that multiple factors need to be used in the classification, which includes mechanical, geometrical and shape features. No single marker can yield reliable prediction given the complexity of the growth and remodeling process of diseased hearts undergoing high-dimensional shape changes. Our approach is promising in terms of clinical translation but the results presented should be interpreted with caution due to the small sample size.

心脏内淀粉样蛋白沉积可引发心脏扩张，并损害其泵血功能。随着心肌弹性进行性丧失，最终将导致心力衰竭，患者会出现呼吸困难、疲乏等进行性加重的症状。与临床恶化相关的生物标志物，对于开发有效治疗方案至关重要。然而迄今为止，心脏淀粉样变性的疾病进展特征仍未得到充分阐明，当前亟需明确疾病进展及心脏组织功能的关键预测因子。在本概念验证研究中，我们基于常规临床磁共振成像（magnetic resonance imaging）及伦敦皇家自由医院国家淀粉样变性中心的随访扫描图像所构建的左心室数学模型，估算得到一组新型标志物。通过力学建模与统计分类方法，我们证实可实现疾病进展的预测。在针对7例样本开展的双盲试验中，我们的预测结果与临床评估结果在6例样本中相符。重要的是，我们发现分类任务需结合多种因素，包括力学特征、几何特征与形态特征。鉴于患病心脏经历高维形态变化且其生长与重塑过程极为复杂，单一标志物无法提供可靠的预测结果。本研究方法在临床转化方面颇具前景，但由于样本量较小，解读本次研究结果时需谨慎。