Abstract
Background
We used longitudinal Aβ imaging data from the AIBL and ADNI studies to calculate the rates of global and regional Aβ-amyloid deposition.
Methods
Four hundred and eighty-five participants from AIBL and ADNI (321 HC; 135 MCI; 29 AD) with 3 or more Aβ imaging assessments were included in the study. While AIBL participants underwent Aβ imaging with PiB and NAV4694 (n=220, median follow-up of 4.8 –range 2.5–10.6– years), or flutemetamol (n=94, median follow-up of 3.2 –range 3.0–3.8– years), ADNI participants underwent Aβ imaging with florbetapir (n=171, median follow-up of 4.1 –range 2.9–5.7 – years). Aβ deposition was derived from the slope of the linear regression plots, and rates of accumulation were calculated from the time it took to go from the median levels of Aβ- HC to the median levels of Aβ+ AD, and they were expressed as both SUVR and Centiloids (CL).
Results
Of 485 initial participants, 417 (86%) (270 HC; 118 MCI; 29 AD) showed positive rates of Aβ accumulation over 3 or more assessments. Irrespective of the Aβ tracer used, Aβ deposition spans more than two decades, averaging 31.6±5.7 (mean±SEM) years to go from the median levels observed in Aβ- HC (1.04 SUVR / 3.8 CL) to the median levels observed in mild Aβ+ AD (2.06 SUVR / 98.4 CL) confirming our previous estimates of 31.2 yrs. This translates in rates of Aβ accumulation of 0.032 SUVR/yr (2.99 CL/yr). The regional assessment showed faster Aβ accumulation in the posterior cingulate/precuneus and frontal areas, and slowest in the hippocampus.
Conclusions
Our new assessment with a longer follow-up confirmed our previous findings that Aβ-amyloid deposition is a slow and protracted process, extending for more than two decades. Centiloids can be used to combine longitudinal Aβ imaging data obtained with different Aβ tracers.