Increasing the Predictive Accuracy of Amyloid-β Blood-Borne Biomarkers in Alzheimer's Disease

Article type: Research Article

Authors: Watt, Andrew D.^{a; b} | Perez, Keyla A.^{a; b; c} | Faux, Noel G.^c | Pike, Kerryn E.^{c; d} | Rowe, Christopher C.^{d; e} | Bourgeat, Pierrick^f | Salvado, Olivier^f | Masters, Colin L.^c | Villemagne, Victor L.^{c; d; e} | Barnham, Kevin J.^{a; b; c; *}

Affiliations: [a] Department of Pathology, The University of Melbourne, Parkville, Melbourne, Victoria, Australia | [b] Neuroproteomics Platform Bio21 Molecular Science and Biotechnology Institute, The University of Melbourne, Parkville, Melbourne, Victoria, Australia | [c] Mental Health Research Institute, The University of Melbourne, Parkville, Melbourne, Victoria, Australia | [d] Department of Nuclear Medicine and Centre for PET, Heidelberg, Victoria, Australia | [e] Department of Medicine, Austin Health, Heidelberg, Victoria, Australia | [f] CSIRO Preventative Health National Flagship ICTC, The Australian e-Health Research Centre, BioMedIA, Herston, Queensland, Australia

Correspondence: [*] Correspondence to: Kevin J.Barnham, Department of Pathology, The University of Melbourne, Victoria 3010, Australia. Tel.: +61 3 83442555; Fax: +61 3 93476750; E-mail: [email protected].

Abstract: Diagnostic measures for Alzheimer's disease (AD) commonly rely on evaluating the levels of amyloid-β (Aβ) peptides within the cerebrospinal fluid (CSF) of affected individuals. These levels are often combined with levels of an additional non-Aβ marker to increase predictive accuracy. Recent efforts to overcome the invasive nature of CSF collection led to the observation of Aβ species within the blood cellular fraction, however, little is known of what additional biomarkers may be found in this membranous fraction. The current study aimed to undertake a discovery-based proteomic investigation of the blood cellular fraction from AD patients (n = 18) and healthy controls (HC; n = 15) using copper immobilized metal affinity capture and Surface Enhanced Laser Desorption/Ionisation Time-Of-Flight Mass Spectrometry. Three candidate biomarkers were observed which could differentiate AD patients from HC (ROC AUC > 0.8). Bivariate pairwise comparisons revealed significant correlations between these markers and measures of AD severity including; MMSE, composite memory, brain amyloid burden, and hippocampal volume. A partial least squares regression model was generated using the three candidate markers along with blood levels of Aβ. This model was able to distinguish AD from HC with high specificity (90%) and sensitivity (77%) and was able to separate individuals with mild cognitive impairment (MCI) who converted to AD from MCI non-converters. While requiring further characterization, these candidate biomarkers reaffirm the potential efficacy of blood-based investigations into neurodegenerative conditions. Furthermore, the findings indicate that the incorporation of non-amyloid markers into predictive models, function to increase the accuracy of the diagnostic potential of Aβ.

Keywords: Amyloid-β, biomarkers, blood, copper, diagnostics, SELDI-TOF-MS

DOI: 10.3233/JAD-2010-101722

Journal: Journal of Alzheimer's Disease, vol. 24, no. 1, pp. 47-59, 2011