DISCOVERY, REPLICATION AND EXTENSION STUDY OF PLASMA PROTEOMIC BIOMARKERS RELATING TO BRAIN AMYLOID BURDEN (CSF Aβ OR AMYLOID-PET) IN THE EMIF-AD BIOMARKER DISCOVERY COHORT

Abstract

Given that Alzheimer's disease (AD) pathology precedes disease onset, a pathology endophenotype design for biomarker discovery creates the opportunity for detection of much earlier markers of disease. We have previously used a pathology endophenotype approach to discover plasma proteomic correlates of amyloid PET, cerebrospinal fluid (CSF) levels of amyloid-β (Aβ) and tau, brain atrophy and progression from prodromal disease to AD. The study presented (EMIF-AD biomarker discovery) aims to undertake replication of these previously identified candidate biomarkers, and furthermore provides additional discovery and extension studies to identify a plasma proteomic signature related to AD pathology at the preclinical and prodromal stages of disease. Targeted plasma proteomic analyses will be conducted in amyloid positive and negative cognitively healthy elderly subjects (N=400), subjects with mild cognitive impairment (N=400) and in patients with mild AD (N=200), selected from the EMIF catalogue. Outcome measures include amyloid status (CSF Aβ or amyloid-PET) and cognitive decline. Proteomic analyses were conducted using luminex xMAP and ELISA assays, and sought to replicate previously identified biomarker findings such as; a 10 plasma protein panel associated with MCI progression (Hye et al, 2014); fibrinogen γ-chain is associated with neocortical amyloid burden (Ashton et al, 2015); and ficolin-2 is associated with CSF tau/Aβ (Baird et al, unpublished). Untargeted proteomic analyses were conducted in a subset of the 1000 (n=597) and used SomaLogic's SOMAscan assay for biomarker discovery. All 1000 subjects had targeted immunocapture based plasma proteomics and amyloid measures (CSF Aβ or amyloid-PET, n=1000). Of these 1000 subjects there is further multimodal data available including genomics (n=872) CSF proteomics (n=681) structural MRI (n=678) and plasma metabolomics (n=600). 25 proteins were measured by targeted assays and >1000 proteins by the SOMAscan assay. The results of the targeted replication as well as the untargeted SOMAscan will be discussed. A plasma proteomic biomarker signature could be used to enrich populations for clinical trial recruitment, and allow easy repeated testing to monitor drug efficacy and toxicity. Further investigation will determine whether a multimodal biomarker signature has added value in this cohort in comparison to a single-modality plasma proteomic biomarker.