Differential association between tau pathology and longitudinal cortical atrophy in subtypes of Alzheimer’s disease

Authors: Rosaleena Mohanty, Daniel Ferreira, Eric Westman

Published: 2020-12-07

Abstract

AbstractBackgroundTau neurofibrillary tangle pathology and cortical atrophy progress at different rates when investigated cross‐sectionally or retrospectively (LaPoint 2017). However, convergence of the two over time remains unclear, especially in the subtypes of AD, which could inform about the underlying neural mechanisms. Here, we studied the association between tau pathology with positron emission tomography (PET) and cortical atrophy with retrospective and prospective structural magnetic resonance imaging (sMRI) in Alzheimer’s disease (AD) dementia. We identified subtypes using both modalities to compare their temporal evolution.MethodsWe identified 383 participants from the ADNI database, which comprised of 30 Aβ+ AD dementia patients, 54 Aβ+ prodromal AD patients, 99 Aβ+ preclinical AD and 200 Aβ‐ healthy individuals (HC). All subjects had tau PET (AV‐1451) and concurrent sMRI (T1‐weighted). Additionally, retrospective and prospective sMRI were available. Preliminarily, we analyzed data from the AD dementia group. Tau PET and longitudinal sMRI were preprocessed to generate regional measures (Desikan 2006) of partial volume corrected standardized uptake value ratio (SUVR) (Greve 2016) and cortical volume, thickness respectively. AD subtypes were characterized by regional volumes and tau SUVR as in a previous study (Byun 2015). Four subtypes were identified: typical AD (TAD), limbic predominant (LP), hippocampal sparing (HS) and minimal atrophy (MA). We investigated the association between tau SUVR and longitudinal cortical thickness in AD dementia and the four subtypes.ResultsIn the preliminary analysis, we observed a significant negative association between tau SUVR and longitudinal cortical thickness (baseline and prospectively) in the entorhinal cortex. Significant association between tau SUVR and cortical thickness was also observed in temporal and parietal areas prospectively. Additionally, subtypes identified based on tau PET and sMRI did not always coincide at baseline, but prospective tracking was indicative of potential convergence between the two (Figure 1).ConclusionThe association of increased tau PET binding with prospective cortical thinning suggests that including prospective imaging is critical to track disease progression. Preliminary evidence suggests that the convergence between tau pathology and cortical atrophy could occur at different rates in different subtypes. Learning subtype‐specific trajectory would be valuable for the design of disease‐slowing/modifying clinical trials.