Department of Biostatistics, Mailman School of Public Health, Columbia University, New York, New York.
Division of Biostatistics, New York State Psychiatric Institute, New York, New York.
Stat Med. 2018 Dec 30;37(30):4721-4742. doi: 10.1002/sim.7951. Epub 2018 Sep 6.
Due to a lack of a gold standard objective marker, the current practice for diagnosing a neurological disorder is mostly based on clinical symptoms, which may occur in the late stage of the disease. Clinical diagnosis is also subject to high variance due to between- and within-subject variability of patient symptomatology and between-clinician variability. Effectively modeling disease course and making early prediction using biomarkers and subtle clinical signs are critical and challenging both for improving diagnostic accuracy and designing preventive clinical trials for neurological disorders. Leveraging the domain knowledge that certain biological characteristics (ie, causal genetic mutation) is part of the disease mechanism, and certain markers (eg, neuroimaging measures, motor and cognitive ability measures) reflect pathological process, we propose a nonlinear model with random inflection points depending on subject-specific characteristics to jointly estimate the changing trajectories of the markers in the same disease domain. The model scales different markers into comparable progression curves with a temporal order based on the mean inflection point and establishes the relationship between the progression of markers with the underlying disease mechanism. The model also assesses how subject-specific characteristics affect the dynamic trajectory of different markers, which offers information on designing preventive therapeutics and personalized disease management strategy. We perform extensive simulation studies and apply our method to markers in neuroimaging, cognitive, and motor domains of Huntington's disease using the data collected from a large multisite natural history study of Huntington's disease, where we assess the temporal ordering of disease impairment between domains. We show that atrophy from certain brain area occurs first, followed by motor and cognitive domain, and show that an average patient has already experienced substantial regional brain atrophy when reaching clinical diagnosis age.
由于缺乏客观的金标准标志物，目前诊断神经疾病的方法主要基于临床症状，而这些症状可能出现在疾病的晚期。由于患者症状的个体内和个体间变异性以及临床医生间的变异性，临床诊断也存在很大差异。使用生物标志物和细微的临床体征来有效模拟疾病进程并进行早期预测，对于提高诊断准确性和设计神经疾病的预防性临床试验至关重要，也极具挑战性。利用某些生物学特征（即因果遗传突变）是疾病机制的一部分，以及某些标志物（如神经影像学测量、运动和认知能力测量）反映病理过程的领域知识，我们提出了一个具有随机拐点的非线性模型，该模型取决于个体特征，以联合估计同一疾病领域中标志物的变化轨迹。该模型根据平均拐点将不同的标志物划分为具有时间顺序的可比进展曲线，并建立标志物与潜在疾病机制之间的关系。该模型还评估了个体特征如何影响不同标志物的动态轨迹，这为设计预防性治疗和个性化疾病管理策略提供了信息。我们进行了广泛的模拟研究，并将我们的方法应用于亨廷顿病的神经影像学、认知和运动领域的标志物，使用来自亨廷顿病大型多中心自然史研究中收集的数据，我们评估了不同领域之间疾病损伤的时间顺序。我们发现某些大脑区域的萎缩首先发生，其次是运动和认知领域，并且表明当达到临床诊断年龄时，平均患者已经经历了大量的区域性脑萎缩。
