PharmacoMetrica , La Fouillade , France.
Amyotroph Lateral Scler Frontotemporal Degener. 2014 Mar;15(1-2):119-29. doi: 10.3109/21678421.2013.838970. Epub 2013 Sep 26.
Our objective was to develop: 1) a longitudinal model to describe amyotrophic lateral sclerosis (ALS) disease progression using the revised Amyotrophic Lateral Sclerosis Functional Rating Scale (ALSFRS-R); and 2) a probabilistic model to estimate the presence of clusters of trajectories in ALS progression over 12 months of treatment. Three hundred and thirty-eight patients treated with placebo from the PRO-ACT database were included in the analyses. A non-linear Weibull model best described the ALS disease progression, and a stepwise logistic regression approach was used to select the variables predicting a slow or fast disease progression. Results identified two clusters of trajectories: 1) slow disease progressors (46% of patients with a change from baseline of 13%); 2) fast disease progressors (54% of patients with a change from baseline of 49%). ROC curve analysis estimated the optimal cut-off for classifying patients as slow or fast disease progressors given ALSFRS-R measurements at 2-4 weeks. Results showed that the degree of ALS disease progression quantified by the ALSFRS-R symptomatic change on placebo is highly heterogeneous. In conclusion, this finding indicates the potential interest of disease progression models for implementing a population enrichment strategy to control the level of heterogeneity in the patients included in new trials.
1）一种纵向模型，用于使用修订后的肌萎缩侧索硬化（ALS）功能评定量表（ALSFRS-R）描述肌萎缩侧索硬化症（ALS）疾病进展；2）一种概率模型，用于估计在 12 个月的治疗期间 ALS 进展中存在轨迹簇的情况。来自 PRO-ACT 数据库的 338 名接受安慰剂治疗的患者被纳入分析。非线性 Weibull 模型最能描述 ALS 疾病进展，采用逐步逻辑回归方法选择预测疾病进展缓慢或快速的变量。结果确定了两个轨迹簇：1）疾病进展缓慢者（基线变化 13%的患者中占 46%）；2）疾病进展迅速者（基线变化 49%的患者中占 54%）。ROC 曲线分析估计了基于 ALSFRS-R 在 2-4 周时的测量值将患者分类为疾病进展缓慢或快速进展者的最佳截断值。结果表明，基于安慰剂的 ALSFRS-R 症状变化定量的 ALS 疾病进展程度高度异质。总之，这一发现表明疾病进展模型对于实施人群富集策略具有潜在的意义，以控制新试验中纳入患者的异质性水平。
