Prognostic plasma protein panel for Aβ deposition in the brain in Alzheimer’s disease

•Existing blood biomarkers for Alzheimer’s disease remain unreliable.•This study used the available amyloid positron emission tomography imaging data of 254 individuals.•We found five plasma biomarker candidates by proteomics and enzyme-linked immunosorbent assay.•Our integrated models were highly predictive of brain amyloid deposition in patients with mild cognitive impairment. Alzheimer’s disease (AD) is the most common age-associated dementia. Many studies have sought to predict cerebral amyloid deposition, the major pathological hallmark of AD, using body fluids such as blood or cerebral spinal fluid (CSF). The use of blood in diagnostic procedures is widespread in medicine; however, existing blood biomarkers for AD remain unreliable. We sought to discover blood biomarkers that discriminate Aβ deposition status in the brain. This study used 107 individuals who were cognitively normal (CN), 107 patients with mild cognitive impairment (MCI), and 40 AD patients with Pittsburg compound B positron emission tomography (PiB-PET) amyloid imaging data available. We found five plasma biomarker candidates via mass spectrometry (MS) based-proteomic analysis and validated these proteins using enzyme-linked immunosorbent assay (ELISA). Our integrated models were highly predictive of brain amyloid deposition, exhibiting 0.871 accuracy with 79% sensitivity and 84% specificity overall, and 0.836 accuracy with 68% sensitivity and 90% specificity in patients with MCI. These results indicated that a combination of proteomic-based blood proteins might be a possible biomarker set for predicting cerebral amyloid deposition.