Comparison of ultrasensitive and mass spectrometry quantification of blood-based amyloid biomarkers for Alzheimer’s disease diagnosis in a memory clinic cohort

BACKGROUND: Alzheimer’s disease (AD) is a complex neurodegenerative disorder with β-amyloid pathology as a key underlying process. The relevance of cerebrospinal fluid (CSF) and brain imaging biomarkers is validated in clinical practice for early diagnosis. Yet, their cost and perceived invasiveness are a limitation for large-scale implementation. Based on positive amyloid profiles, blood-based biomarkers should allow to detect people at risk for AD and to monitor patients under therapeutics strategies. Thanks to the recent development of innovative proteomic tools, the sensibility and specificity of blood biomarkers have been considerably improved. However, their diagnosis and prognosis relevance for daily clinical practice is still incomplete. METHODS: The Plasmaboost study included 184 participants from the Montpellier’s hospital NeuroCognition Biobank with AD (n = 73), mild cognitive impairments (MCI) (n = 32), subjective cognitive impairments (SCI) (n = 12), other neurodegenerative diseases (NDD) (n = 31), and other neurological disorders (OND) (n = 36). Dosage of β-amyloid biomarkers was performed on plasma samples using immunoprecipitation-mass spectrometry (IPMS) developed by Shimadzu (IPMS-Shim Aβ(42), Aβ(40), APP(669–711)) and Simoa Human Neurology 3-PLEX A assay (Aβ(42), Aβ(40), t-tau). Links between those biomarkers and demographical and clinical data and CSF AD biomarkers were investigated. Performances of the two technologies to discriminate clinically or biologically based (using the AT(N) framework) diagnosis of AD were compared using receiver operating characteristic (ROC) analyses. RESULTS: The amyloid IPMS-Shim composite biomarker (combining APP(669–711)/Aβ(42) and Aβ(40)/Aβ(42) ratios) discriminated AD from SCI (AUC: 0.91), OND (0.89), and NDD (0.81). The IPMS-Shim Aβ(42/40) ratio also discriminated AD from MCI (0.78). IPMS-Shim biomarkers have similar relevance to discriminate between amyloid-positive and amyloid-negative individuals (0.73 and 0.76 respectively) and A−T−N−/A+T+N+ profiles (0.83 and 0.85). Performances of the Simoa 3-PLEX Aβ(42/40) ratio were more modest. Pilot longitudinal analysis on the progression of plasma biomarkers indicates that IPMS-Shim can detect the decrease in plasma Aβ(42) that is specific to AD patients. CONCLUSIONS: Our study confirms the potential usefulness of amyloid plasma biomarkers, especially the IPMS-Shim technology, as a screening tool for early AD patients. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s13195-023-01188-8.