Quantification of baseline amyloid PET in individuals with subjective cognitive decline can identify risk of amyloid accumulation and cognitive worsening: the FACEHBI study

Abstract

Purpose: Amyloid PET imaging is capable of measuring brain amyloid load in vivo. The aim of this study is to assess the relationship of the baseline amyloid with its accumulation over time and with cognition in individuals with subjective cognitive decline (SCD), giving a focus on those below Aβ positivity thresholds.

Methods: 118 of 197 individuals with SCD from the Fundació ACE Healthy Brain Initiative underwent three [¹⁸F]florbetaben scans and the remaining 79 underwent two scans in a 5-year span. Individuals were categorised based on baseline Centiloid values (CL) into amyloid positive (Aβ+; CL > 35.7), Grey Zone (GZ; 20 < CL ≤ 35.7), and amyloid negative (Aβ-; CL ≤ 20). Relationship between conversion to mild cognitive decline (MCI) and baseline amyloid levels was assessed. Then, to focus on sub-threshold individuals with amyloid accumulation, the Aβ- group was split into two groups (N1 (CL ≤ 13.5) and N2 (13.5 < CL ≤ 20)), Aβ accumulation was determined, and a parametric image analysis of the Aβ accumulators in the N1 group was performed.

Results: At baseline, 20 individuals were Aβ+, 8 GZ, 160 N1, and 9 N2. Higher Aβ load, older and less educated individuals presented increased risk of MCI-conversion. Longitudinally, 19% of N1 individuals were accumulators despite very low Aβ burden at baseline. Meanwhile, 89% of the N2 group accumulated Aβ as well as all GZ individuals (which had the highest rate of amyloid accumulation, 5.1 CL/year). In the parametric image analysis of N1 accumulators, a region within the precuneus was linked to increased Aβ over time.

Conclusion: Baseline amyloid levels differentiate individuals who accumulate amyloid over time and that are at risk for cognitive decline, including those at sub-threshold levels of Aβ. This can be valuable to identify pre-clinical AD in a SCD population.

Keywords: Alzheimer’s disease; Amyloid PET; FACEHBI; Florbetaben; Longitudinal study; Subjective Cognitive Decline.

Fig. 1

FACEHBI study timeline

Fig. 2

Centiloid classification, split into groups, and the sequential steps of the analyses carried in this study. Number of individuals in each group at baseline is indicated with an asterisk

Fig. 3

Histogram of Centiloid values at baseline. Vertical lines display the cut-offs for Aβ- (CL ≤ 20, blue) and Aβ+ (CL > 35.7 red)

Fig. 4

A Cox Proportional Hazard analysis of the risk of SCD to MCI conversion based on the CL level at baseline, age, and education. B Kaplan-Meyer-like adjusted curves indicating the likelihood of remaining SCD over time

Fig. 5

Longitudinal change in CL. Individuals that changed classification between first and last scan are highlighted in yellow, orange or red depending on their progression. Horizontal dashed lines show the cut-off for Aβ- (CL ≤ 20) and Aβ+ (CL > 35.7) scans. Individuals that were Aβ+ at baseline are displayed in slightly darker grey

Fig. 6

Visualisation of the Linear Mixed Model results (lines) with the measured CL values (dots). The dashed lines correspond to individual-level results and the full lines are group-level results

Fig. 7

Histogram of the CL ARC for the whole cohort fitted with a 2-curve GMM

Fig. 8

Centiloid annualised rate of accumulation (CL ARC) as function of measured baseline CL. Colours indicate amyloid level groups and dashed vertical lines indicate thresholds for grouping at CL below 13.5 (purple), below 20 (blue), below 35.7 (green) and above 35.7 (red). Dashed horizontal line shows the cut-off for CL ARC classification of individuals with Aβ accumulation. Those considered accumulators are indicated by the triangular symbol

Fig. 9

Volume with higher baseline signal for N1 accumulators and N1 non-accumulators of amyloid (red line) overlayed on the scan of an N1 accumulator (MNI space). Grey scale shows SUVR range. Blue line shows the Centiloid target region and the green line shows the reference region (whole cerebellum)