Affiliations: [a] Section for Dementia Research, Hertie-Institute of Clinical Brain Research and Department of Psychiatry and Psychotherapy, University of Tübingen, Germany | [b] DZNE, German Center for Neurodegenerative Diseases, Tübingen, Germany | [c] Department of Psychiatry and Psychotherapy, University of Tübingen, Germany | [d] NMI Natural and Medical Sciences Institute at the University of Tübingen, Reutlingen, Germany | [e] Department of Neurodegeneration, Center of Neurology, Hertie Institute for Clinical Brain Research, University of Tübingen, Germany | [f] Department of Information and Cognitive Sciences, Center for Bioinformatics, University of Tübingen, Tübingen, Germany
Correspondence:
[*]
Correspondence to: Prof. Dr. Christoph Laske, M.D., Section for Dementia Research, Hertie-Institute of Clinical Brain Research and Department of Psychiatry and Psychotherapy, University of Tübingen, Calwer Street 14, D-72076 Tübingen, Germany. Tel.: +49 7071 2983444; Fax: +49 7071 294141; E-mail: christoph.laske@med.uni-tuebingen.de.
Abstract: Background:Alzheimer’s disease (AD) has been linked to a state of cerebral and systemic inflammation. The objective of the present study was to determine whether singular markers or a set of inflammatory biomarkers in peripheral blood allow discrimination between AD patients and healthy controls at the individual level. Methods:Using bead based multiplexed sandwich immunoassays, 25 inflammatory biomarkers were measured in 164 serum samples from individuals with early AD and age-matched cognitively healthy elderly controls. The data set was randomly split into a training set for feature selection and classification training and a test set for class prediction of blinded samples (1 : 1 ratio) to evaluate the chosen predictors and parameters. Multivariate data analysis was performed with use of a support vector machine (SVM). Results:After selection of sTNF-R1 as most discriminative parameter in the training set, the application of SVM to the independent test dataset resulted in a 90.0% correct classification for individual AD and control subjects. Conclusions:We identified sTNF-R1 from a marker set consisting of 25 inflammatory biomarkers, which allowed SVM-based discrimination of AD patients from healthy controls on a single-subject classification level comparably well as biomarker panels with a clinically relevant accuracy and validity. Although larger sample populations will be needed to confirm this diagnostic accuracy, our study suggests that sTNF-R1 in serum—either as singular marker or incorporated into a biomarker panel—could be a powerful new biomarker for detection of AD. In addition, selective inhibition of TNF-R1 function may represent a new therapeutic approach in AD.
