Identification of CSF biomarkers for cerebral amyloid angiopathy using a targeted proteomics approach

Abstract

AbstractBackgroundCerebral amyloid angiopathy (CAA) is characterized by progressive vascular deposition of amyloid‐β (Aβ), and can cause cognitive decline in the elderly. Currently, the diagnosis of CAA is based on neuroimaging and includes the identification of lobar (micro‐)bleeds and cortical superficial siderosis. There is, however, a need for alternative methods as these neuroimaging methods are not optimal for detection of CAA during life. Moreover, CAA is strongly associated with the risk to develop edematous or hemorrhagic amyloid‐related imaging abnormalities (ARIA) as a consequence of anti‐Aβ immunotherapy (e.g. aducanumab and lecanemab) for Alzheimer’s disease (AD). There is a need for biomarkers that can provide an accurate and early diagnosis of CAA, and additionally help in the identification of AD patients with a high burden of CAA, who are at risk of developing ARIA due to immunotherapy. We aimed to identify novel biomarker candidates for CAA in cerebrospinal fluid (CSF) using a targeted proteomics approach.MethodWe analyzed a panel of 367 proteins in CSF samples of 34 CAA patients and 51 age‐ and sex‐matched controls using the Olink Explore multiplex immunoassay platform (Neurology panel; Olink, Uppsala, Sweden). Proteins were considered measurable when their levels were above the limit of detection of the assay in at least 70% of the samples.ResultFor 15 of the 263 proteins that could be measured in CSF, we identified significantly different (p<0.05) levels in CSF of CAA patients as compared to controls. The strongest difference was observed for neurofilament light chain (NFL). Among these 15 proteins, there were two proteins (milk fat globule‐epidermal growth factor 8 (MFG‐E8/lactadherin) and urokinase‐type plasminogen activator (uPA)) for which we previously observed a significant changed level in CSF of CAA patients compared to controls 1,2.ConclusionConfirmation of previously identified CAA biomarker candidates shows the validity and strength of our biomarker identification study using a targeted proteomics approach. The 15 candidates identified in our study, therefore, hold promise for development into biomarkers that may improve the detection of CAA.