Head-to-head comparison of neurodegeneration biomarkers across two analytical platforms in Alzheimer’s disease

BACKGROUND: Blood-based biomarkers are increasingly used to support the identification of Alzheimer's disease (AD) in clinical and research settings. However, variability between analytical platforms limits their interchangeability and may affect their clinical interpretation. Direct head-to-head comparisons across assays and their alignment with clinical phenotypes remain limited. OBJECTIVE: To compare the inter-test agreement of blood-based biomarkers measured using Roche and Fujirebio assays and to determine which biomarker shows the most robust analytical concordance and clinical relevance in AD. METHODS: Participants with AD (n = 86) and cognitively healthy controls (HC; n = 56) underwent blood biomarker assessment using Roche and Fujirebio platforms. Plasma phosphorylated tau at threonine 217 (pTau217), plasma phosphorylated tau at threonine 181 (pTau181), Amyloid-β42 (Aβ42), amyloid-β40 (Aβ40), and the amyloid-β42/amyloid-β40 ratio (Aβ42/Aβ40) and glial fibrillary acidic protein (GFAP) were analyzed. Inter-test agreement was assessed using intraclass correlation coefficients, concordance analyses, and regression-based methods. Clinical relevance was evaluated by association with measures of cognitive and functional impairment evaluated using the Mini-Mental State Examination (MMSE) and the Clinical Dementia Rating Scale - Sum of Boxes (CDR-SB). Analyses were adjusted for age and sex and complemented by sensitivity and stability analyses. RESULTS: Inter-test agreement varied across biomarkers, with pTau217 demonstrating the highest and most stable concordance between Roche and Fujirebio assays. Among the evaluated markers, pTau217 also showed the strongest and most consistent associations with global cognition measures across diagnostic groups and best discriminated between individuals with AD and HC. GFAP demonstrated robust inter-test agreement but weaker alignment with clinical measures, while amyloid-based markers showed lower concordance and greater variability across platforms. CONCLUSIONS: Head-to-head comparison of Roche and Fujirebio assays revealed biomarker-specific differences in inter-test agreement, limiting the direct interchangeability of measurements across platforms. Plasma pTau217 showed the most consistent analytical concordance and clinical associations in this setting, supporting its potential utility in cross-platform applications. However, these findings also highlight the need for assay-specific calibration and cautious interpretation of biomarker thresholds in clinical and research use and the need for further validation across diverse populations. TRIAL REGISTRATION: NCT05059158, NCT05317871.