Multi-cohort and longitudinal Bayesian clustering study of stage and subtype in Alzheimer’s disease

Abstract

Understanding Alzheimer’s disease (AD) heterogeneity is important for understanding the underlying pathophysiological mechanisms of AD. However, AD atrophy subtypes may reflect different disease stages or biologically distinct subtypes. Here we use longitudinal magnetic resonance imaging data (891 participants with AD dementia, 305 healthy control participants) from four international cohorts, and longitudinal clustering to estimate differential atrophy trajectories from the age of clinical disease onset. Our findings (in amyloid-β positive AD patients) show five distinct longitudinal patterns of atrophy with different demographical and cognitive characteristics. Some previously reported atrophy subtypes may reflect disease stages rather than distinct subtypes. The heterogeneity in atrophy rates and cognitive decline within the five longitudinal atrophy patterns, potentially expresses a complex combination of protective/risk factors and concomitant non-AD pathologies. By alternating between the cross-sectional and longitudinal understanding of AD subtypes these analyses may allow better understanding of disease heterogeneity.

MRI data-driven clustering reveals different subtypes of Dementia with Lewy bodies

“…Thirdly, clustering in the current study is cross-sectional, like virtually all current MRI subtyping studies 5 , 8 . The advent of new longitudinal clustering methods 38 will open the door for future longitudinal subtyping studies in DLB, helping to better characterize disease progression of our current DLB subtypes.…”

Tau accumulates differently in four subtypes of Alzheimer's disease

“…Recently, subtype analyses of AD have been conducted using large multinational cohort studies. 6,7 Based on longitudinal MRI data, five subtypes have been categorized by Poulakis et al 7 These studies were based on a cohort study including ADNI and J-ADNI, which mainly targeted typical AD and may have excluded "atypical AD" such as PCA, LPPA, and FAD.…”

“…Based on pathological and positron emission tomography (PET) studies, deposition of Aβ is reported as the first step, followed by the accumulation and propagation of tau and neurodegeneration 4,5 . However, these pathological changes involve various brain regions and neuronal cell types and present heterogeneous clinical manifestations 6,7 . AD commonly presents as an amnestic syndrome, whereas non‐amnestic symptoms may dominate in less common subtypes of AD, including the posterior cortical atrophy (PCA), 8,9 frontal variant of AD (FAD), 10,11 and logopenic primary progressive aphasia (LPPA) 12 …”

“…4,5 However, these pathological changes involve various brain regions and neuronal cell types and present heterogeneous clinical manifestations. 6,7 AD commonly presents as an amnestic syndrome, whereas nonamnestic symptoms may dominate in less common subtypes of AD, including the posterior cortical atrophy (PCA), 8,9 frontal variant of AD (FAD), 10,11 and logopenic primary progressive aphasia (LPPA). 12 Utilizing PET tracers for Aβ and tau, we previously reported that expansion of tau distribution was observed from the parahippocampal gyrus to the cerebral cortex with advancing AD, whereas Aβ was already distributed widely even in the earliest clinical stage.…”

Prediction of misfolded proteins spreading in Alzheimer’s disease using machine learning

“…Furthermore, limbic-predominant and medial temporal lobe-sparing patterns, also highlight possible temporal patterns of atypical clinical variants of AD ( 36 ). These aspects appear to be yet open discussions, with studies hypothetizing that subtypes might be just different stages of the disease, urging for clearer longitudinal analysis ( 37,38 ), while other relatively confirming the sub-typing via longitudinal analysis ( 39 ). Given the complexity and the available data, investigation of different trajectories by using the proposed model is a future work.…”