Diagnostic reclassification under the AA-2024 biological framework: A real-world comparison with ATN 2018 using cerebrospinal fluid biomarkers
Caio Viana Botelho, Amandha Carneiro, Carolina Regadas, Cintia Cavalcante, Ian Ribeiro, Samia Rabelo, Vania Araujo, Glauber Ferreira, Bruno Diógenes Iepsen, Norberto Anizio Ferreira FrotaBackground
Alzheimer's disease (AD) has evolved from a syndromic diagnosis to a biologically defined entity supported by cerebrospinal fluid (CSF) biomarkers. The Alzheimer's Association 2024 (AA-2024) criteria are expected to expand biological AD classification compared with the 2018 NIA–AA AT(N) framework.
Objective
To compare biological AD classification using the AT(N) 2018 framework and the AA-2024 criteria in a real-world cohort.
Methods
We conducted a retrospective study including 146 consecutive patients evaluated for cognitive impairment in a tertiary neurology center who underwent CSF biomarker testing (Aβ 42 , Aβ 40 , Aβ 42 /Aβ 40 ratio, phosphorylated tau, and total tau). Each patient was classified according to AT(N) 2018 and AA-2024 criteria. Diagnostic distributions, reclassification patterns, and agreement (Cohen's κ) were analyzed. Group comparisons used Wilcoxon rank-sum and Fisher's exact tests.
Results
Under AT(N) 2018, 40/146 (27.4%) patients met biological AD criteria, whereas AA-2024 classified 72/146 (49.3%) as AD. Reclassification occurred exclusively from non-AD to AD (32 patients, 21.9%), with no reverse reclassification. Agreement was moderate (κ = 0.57). Concordant AD cases (AD+/AD+) showed lower Aβ 42 /Aβ 40 ratios and higher total and phosphorylated tau levels than reclassified cases (AD−/AD+) (all p < 0.001), as well as higher neurodegeneration marker positivity (70% versus 25%; p < 0.001). Final clinical AD diagnosis was assigned to 98% of concordant cases versus 38% of reclassified cases (p < 0.001).
Conclusions
AA-2024 increases biological AD classification by identifying amyloid-positive individuals with lower tau burden and clinical heterogeneity, underscoring the need to integrate biomarker-based definitions with clinical and longitudinal assessment.