Central Slab versus Whole Brain to Measure Brain Atrophy in Multiple Sclerosis
Central Slab versus Whole Brain to Measure Brain Atrophy in Multiple Sclerosis
复制标题
DOI：
10.1159/000495798
复制
发表时间：
2018-01-01
影响因子：
2.4
通讯作者：
Yaldizli, Ozgur
中科院分区：
文献类型：
--
作者：
Ruberte, Esther;Sinnecker, Tim;Yaldizli, Ozgur
背景：使用萎缩标准化的结构图像评估（SIENA）用于测量多发性硬化（MS）的脑萎缩。然而，大脑提取容易在大脑的上部和下部产生伪影。为了克服这些缺点，一些关键的MS试验使用中央板而不是整个大脑作为SIENA的输入。本研究的目的是比较这两种方法之间萎缩测量的内部一致性和统计离散度，与临床结局的相关性以及临床试验中所需的样本量。研究方法：使用SIENA评估了119例MS患者的脑体积变化，并对3D T1加权磁化制备快速梯度回波数据集进行了5年随访，使用全脑或中央板，范围为-10至+60 mm蒙特利尔神经学研究所图谱坐标。统计学分析包括四分位离散系数、与临床结局的偏相关性和样本量计算。临床结局指标包括扩展残疾状态量表、MS功能复合量表和符号数字模式测试。结果如下：使用中央脑片作为SIENA输入的年化脑萎缩率高于全脑（每年-0.51 +/- 0.49 vs. -0.37 +/- 0.39%，p < 0.001）。在5年随访时，中枢和全脑体积变化显示出相当的统计学离散度，并且与临床结局相似。样本量计算估计，在SIENA中使用中央板而不是全脑选项时，检测给定治疗效果所需的患者减少14%。结论：中央板和全脑SIENA产生了相当的统计离散度，与临床结局具有相似的相关性。(C)2019 S. Karger AG，巴塞尔
Background: Structural Image Evaluation using Normalization of Atrophy (SIENA) is used to measure brain atrophy in multiple sclerosis (MS). However, brain extraction is prone to artefacts in the upper and lower parts of the brain. To overcome these shortcomings, some pivotal MS trials used a central slab instead of the whole brain as input for SIENA. The aim of this study was to compare the internal consistency and statistical dispersion of atrophy measures, associations with clinical outcomes and required sample sizes in clinical trials between these two approaches. Methods: Brain volume change was assessed using SIENA in 119 MS patients with 5-years follow-up on 3D T1-weighted Magnetization Prepared Rapid Gradient Echo datasets using the whole brain or a central slab ranging from -10 to +60 mm Montreal Neurological Institute atlas coordinates. The statistical analysis included the quartile coefficient of dispersion, partial correlations with clinical outcomes and sample size calculations. Clinical outcome measures comprised the Expanded Disability Status Scale, MS Functional Composite and Symbol Digit Modalities Test. Results: Annualized brain atrophy rates were higher using central slab than whole brain as input for SIENA (-0.51 +/- 0.49 vs. -0.37 +/- 0.39% per year, p < 0.001). Central and whole brain volume change showed comparable statistical dispersion and similarly correlated with clinical outcomes at 5-years follow-up. Sample size calculations estimated 14% fewer patients required to detect a given treatment effect when using the central slab instead of the whole brain option in SIENA. Conclusion: Central slab and whole brain SIENA produced comparable statistical dispersion with similar associations to clinical outcomes. (C) 2019 S. Karger AG, Basel