Early-onset cognitive impairment diagnosis is often challenging due to the overlapping symptoms between the different degenerative and non-degenerative conditions. We aimed to evaluate the usefulness of cerebrospinal fluid (CSF) biomarkers in early-onset cognitive impairment differential diagnosis, to assess their contribution to the diagnosis of Alzheimer's disease (AD) based on the new National Institute of Aging-Alzheimer's Association (NIA-AA) workgroup's recommendations and their capacity to predict subsequent decline in early-onset mild cognitive impairment (MCI). 37 controls and 120 patients (clinical onset <65 years) with diagnosis based on criteria available in 2009 (51 MCI, 42 AD, 10 frontotemporal dementia (FTD), 3 posterior cortical atrophy, and 14 primary progressive aphasia (PPA)) were included. In addition, all subjects were also reclassified according to the revised criteria for MCI, AD, FTD, and PPA, excluding CSF data. We assessed the impact of adding the CSF data to the subject categorization according to the NIA-AA criteria. After inclusion of CSF results, 90% of amnestic and 82% of the non-amnestic AD presentation could be categorized as "high probability", while 3% of AD patients fit into the category "dementia probably not due to AD". All the 24 MCI patients who progressed to AD dementia and only 1/27 stable MCI presented pathological CSF at baseline. Only 4% of the FTD clinical diagnosis had pathological CSF levels. CSF biomarkers provide high diagnostic accuracy in a clinical context in differentiating AD, frontotemporal lobar degeneration, and controls in presenile subjects and can be used equally in amnestic and non-amnestic AD. Abnormal CSF-AD biomarker levels predict subsequent progression to AD dementia in subjects with early-onset MCI.
Keywords: CSF biomarkers; NIA-AA criteria; early-onset Alzheimer's disease.