Preclinical detection of Alzheimer's disease pathology using conceptual discrimination abilities

Abstract

Background: Performance on the Conceptual Matching Task (CMT), a measure of discrimination between conceptually confusable items, has been suggested as a cognitive marker of rhinal cortex atrophy, one of the first brain regions affected by Alzheimer's disease (AD) pathology.

Objectives: We aimed to determine whether CMT can detect preclinical AD, and whether CMT performance is related to regional deposition of tau protein or other AD-associated lesions including amyloid (Aβ) accumulation and white matter hyperintensities (WMH).

Design, setting and participants: This cross-sectional study include 101 participants from the UCL2016-121 cohorts in Brussels, Belgium, classified as 56 Aβ-negative cognitively unimpaired (Aβ-CU), 25 Aβ-positive CU (Aβ+CU, preclinical AD), and 20 Aβ-positive mildly cognitively impaired (Aβ+MCI, prodromal AD) individuals.

Measurements: Participants underwent CMT and a standard neuropsychological assessment that included the Preclinical Alzheimer Cognitive Composite (PACC5), an Aβ status examination, a 3D-T1 MRI and a [¹⁸F]MK-6240 tau-PET scan.

Results: CMT performance was lower among Aβ+MCI and Aβ+CU than Aβ-CU individuals. The effect of Aβ on CMT performance was stronger in the presence of WMH, but rhinal tau burden did not explain CMT performance beyond the effects of Aβ and WMH. CMT performance correlated with executive, memory, and language performance. Finally, CMT was more sensitive than PACC5 to detect CU individuals with Aβ or tau pathology.

Conclusion: Given that impaired performance is observed earlier in the CMT than in standard neuropsychological tests, this test shows promise as an early diagnostic tool for AD and may offer significant utility in the context of clinical trials.

Keywords: Amyloid; Conceptual discrimination; Preclinical Alzheimer's disease; Rhinal cortex; Tau.

Fig. 1

Example of trials of the Conceptual Matching Task (CMT) [21]. Legend: The participant was instructed to determine whether the word and the picture represented the same concept. The first example illustrates an item from matching trials, the second from the non-match distant trials, and the last one an item from the non-match close trials.

Fig. 2

Boxplots of CMT performance in each group according to condition. Legend: A. Boxplots of CMT performance according to amyloid and cognitive status. B. Boxplots of CMT performance according to tau and cognitive status. In this analysis, we grouped Tau-MCI and Tau+MCI because of the small number of participants in the first subgroup (n = 2). All comparisons were corrected for the effect of age, sex and education. TO= Trade-Off (higher score reflects poorer performance); Aβ = amyloid; CU = cognitively unimpaired participants; MCI = mild cognitive impaired; *: p<.05, **: p<.01.

Fig. 3

Illustration of the interaction effect between amyloid load and white matter hyperintensities in the entire sample. WMH = white matter hyperintensities volume (mm³); Aβ = amyloid-beta; CU= cognitively unimpaired participants; MCI= mild cognitively impaired.

Fig. 4

ROC Curves comparing performance to distinguish Aβ± or Tau± cognitively unimpaired individuals. ROC curves data table evaluating the cognitive metrics to distinguish Aβ-CU vs Aβ+CU (left side) and Tau-CU vs Tau+CU (right side). Amyloid positivity was defined as a CL value>20 and Tau positivity was established on the basis of a Braak stage >0. The best AUC curves for each comparison are in bold. Legend: CU = cognitively unimpaired; Aβ = amyloid-β; CMT = Conceptual Matching Task; TO= Trade-Off; PACC5 = preclinical Alzheimer cognitive composite; AUC = area under the curve; Sn = sensitivity; Sp = specificity.