背景:由于缺乏疾病特征,阿尔茨海默病(AD)和额颞叶痴呆(FTD)的准确诊断代表了健康问题。我们评估了SIMOA小组在43名AD和33名FTD患者的脑脊液(CSF)中的表现,并结合了人口统计学-临床特征。
方法:136名受试者(AD:n=43,FTD:n=33,对照:n=60)参与。单分子阵列(SIMOA),胶质纤维酸性蛋白(GFAP),神经丝光(NfL),TAU,用多重neuro4plex试剂盒分析CSF中的泛素羧基末端水解酶L1(UCH-L1)。受试者工作特征(ROC)曲线分析比较曲线下面积(AUC),而稀疏偏最小二乘判别分析(sPLS-DA)的原理用于加强自信疾病集群的识别。
结果:CSF显示与对照相比,AD中所有SIMOA生物标志物的水平增加(AUC:分别为0.71、0.86、0.92和0.94)。在具有NfL的FTD中观察到类似的模式,TAU,和UCH-L1(AUC:0.85、0.72和0.91)。sPLS-DA揭示了两个成分,解释了19%和9%的数据集变异。
结论:CSF数据在AD中提供了很高的诊断准确性,FTD,控制歧视。人口统计学-临床特征和生物标志物浓度的亚组强调了组合不同类型的数据以成功和更有效的队列聚类的潜力。
BACKGROUND: Accurate diagnosis of Alzheimer\'s disease (AD) and frontotemporal dementia (FTD) represents a health issue due to the absence of disease traits. We assessed the performance of a SIMOA panel in cerebrospinal fluid (CSF) from 43 AD and 33 FTD patients with 60 matching Control subjects in combination with demographic-clinical characteristics.
METHODS: 136 subjects (AD: n = 43, FTD: n = 33, Controls: n = 60) participated. Single-molecule array (SIMOA), glial fibrillary acidic protein (GFAP), neurofilament light (NfL), TAU, and ubiquitin carboxy-terminal hydrolase L1 (UCH-L1) in CSF were analyzed with a multiplex neuro 4plex kit. Receiver operating characteristic (ROC) curve analysis compared area under the curve (AUC), while the principal of the sparse partial least squares discriminant analysis (sPLS-DA) was used with the intent to strengthen the identification of confident disease clusters.
RESULTS: CSF exhibited increased levels of all SIMOA biomarkers in AD compared to Controls (AUCs: 0.71, 0.86, 0.92, and 0.94, respectively). Similar patterns were observed in FTD with NfL, TAU, and UCH-L1 (AUCs: 0.85, 0.72, and 0.91). sPLS-DA revealed two components explaining 19% and 9% of dataset variation.
CONCLUSIONS: CSF data provide high diagnostic accuracy among AD, FTD, and Control discrimination. Subgroups of demographic-clinical characteristics and biomarker concentration highlighted the potential of combining different kinds of data for successful and more efficient cohort clustering.