Probert Fay, Yeo Tianrong, Zhou Yifan, Sealey Megan, Arora Siddharth, Palace Jacqueline, Claridge Timothy D W, Hillenbrand Rainer, Oechtering Johanna, Leppert David, Kuhle Jens, Anthony Daniel C
Department of Pharmacology, University of Oxford, Oxford OX1 3QT, UK.
Department of Chemistry, University of Oxford, Oxford OX1 3TA, UK.
Brain Commun. 2021 Apr 19;3(2):fcab084. doi: 10.1093/braincomms/fcab084. eCollection 2021.
Eighty-five percent of multiple sclerosis cases begin with a discrete attack termed clinically isolated syndrome, but 37% of clinically isolated syndrome patients do not experience a relapse within 20 years of onset. Thus, the identification of biomarkers able to differentiate between individuals who are most likely to have a second clinical attack from those who remain in the clinically isolated syndrome stage is essential to apply a personalized medicine approach. We sought to identify biomarkers from biochemical, metabolic and proteomic screens that predict clinically defined conversion from clinically isolated syndrome to multiple sclerosis and generate a multi-omics-based algorithm with higher prognostic accuracy than any currently available test. An integrative multi-variate approach was applied to the analysis of cerebrospinal fluid samples taken from 54 individuals at the point of clinically isolated syndrome with 2-10 years of subsequent follow-up enabling stratification into clinical converters and non-converters. Leukocyte counts were significantly elevated at onset in the clinical converters and predict the occurrence of a second attack with 70% accuracy. Myo-inositol levels were significantly increased in clinical converters while glucose levels were decreased, predicting transition to multiple sclerosis with accuracies of 72% and 63%, respectively. Proteomics analysis identified 89 novel gene products related to conversion. The identified biochemical and protein biomarkers were combined to produce an algorithm with predictive accuracy of 83% for the transition to clinically defined multiple sclerosis, outperforming any individual biomarker in isolation including oligoclonal bands. The identified protein biomarkers are consistent with an exaggerated immune response, perturbed energy metabolism and multiple sclerosis pathology in the clinical converter group. The new biomarkers presented provide novel insight into the molecular pathways promoting disease while the multi-omics algorithm provides a means to more accurately predict whether an individual is likely to convert to clinically defined multiple sclerosis.
85%的多发性硬化症病例始于一种被称为临床孤立综合征的离散发作，但37%的临床孤立综合征患者在发病20年内未出现复发。因此，识别能够区分最有可能发生第二次临床发作的个体与那些仍处于临床孤立综合征阶段的个体的生物标志物，对于应用个性化医疗方法至关重要。我们试图从生化、代谢和蛋白质组学筛查中识别出能够预测从临床孤立综合征到多发性硬化症的临床定义转化的生物标志物，并生成一种基于多组学的算法，其预后准确性高于任何目前可用的检测方法。一种综合多变量方法被应用于分析从54名处于临床孤立综合征阶段的个体采集的脑脊液样本，这些个体随后进行了2至10年的随访，从而能够分层为临床转化者和非转化者。临床转化者发病时白细胞计数显著升高，预测第二次发作的准确率为70%。临床转化者的肌醇水平显著升高，而葡萄糖水平降低，分别以72%和63%的准确率预测向多发性硬化症的转变。蛋白质组学分析确定了89种与转化相关的新基因产物。将所识别的生化和蛋白质生物标志物结合起来，生成一种算法，对于向临床定义的多发性硬化症转变的预测准确率为83%，优于包括寡克隆带在内的任何单独的生物标志物。所识别的蛋白质生物标志物与临床转化者组中过度的免疫反应、能量代谢紊乱和多发性硬化症病理一致。所呈现的新生物标志物为促进疾病的分子途径提供了新的见解，而多组学算法提供了一种更准确地预测个体是否可能转化为临床定义的多发性硬化症的方法。
