Alzheimer’s Disease Diagnosis Based on the Amyloid,Tau,and Neurodegeneration Scheme (ATN) in a Real-Life Multicenter Cohort of General Neurological Centers

Abstract

Background:

The ATN scheme was proposed as an unbiased biological characterization of the Alzheimer’s disease (AD) spectrum, grouping biomarkers into three categories: brain Amyloidosis-A, Tauopathy-T, Neurodegeneration-N. Although this scheme was mainly recommended for research, it is relevant for diagnosis.

Objective:

To evaluate the ATN scheme performance in real-life cohorts reflecting the inflow of patients with cognitive complaints and different underlying disorders in general neurological centers.

Methods:

We included patients (n = 1,128) from six centers with their core cerebrospinal fluid-AD biomarkers analyzed centrally. A was assessed through Aβ₄₂/Aβ₄₀, T through pTau-181, and N through tTau. Association between demographic features, clinical diagnosis at baseline/follow-up and ATN profiles was assessed.

Results:

The prevalence of ATN categories was: A-T-N-: 28.3%; AD continuum (A + T-/+N-/+): 47.8%; non-AD (A- plus T or/and N+): 23.9%. ATN profiles prevalence was strongly influenced by age, showing differences according to gender, APOE genotype, and cognitive status. At baseline, 74.6% of patients classified as AD fell in the AD continuum, decreasing to 47.4% in mild cognitive impairment and 42.3% in other neurodegenerative conditions. At follow-up, 41% of patients changed diagnosis, and 92% of patients that changed to AD were classified within the AD continuum. A + was the best individual marker for predicting a final AD diagnosis, and the combinations A + T+ (irrespective of N) and A + T+N+ had the highest overall accuracy (83%).

Conclusion:

The ATN scheme is useful to guide AD diagnosis in real-life neurological centers settings. However, it shows a lack of accuracy for patients with other types of dementia. In such cases, the inclusion of other markers specific for non-AD proteinopathies could be an important aid to the differential diagnosis.