Disease-specific patterns of cortical and subcortical degeneration in a longitudinal study of Alzheimer's disease and behavioural-variant frontotemporal dementia

- We report divergent patterns of cortical and subcortical involvement in AD and bvFTD with disease progression.
- AD show more atrophy in cortical regions (e.g., posterior cingulate) than in subcortical structures over time.
- The reverse pattern is seen in bvFTD, with greater involvement of subcortical striatal structures with disease progression.
- This paper demonstrates the utility of linear mixed effects models to examine disease-specific patterns of neurodegeneration.

BackgroundClinical differentiation between Alzheimer's disease (AD) and behavioural-variant frontotemporal dementia (bvFTD) is challenging due to overlapping clinical features at presentation. Whilst diagnostic criteria for both disorders incorporate evidence of frontal and temporal cortical atrophy, understanding of the progression of atrophy in these disorders is limited. This study aimed to elucidate common and disease-specific progressive changes in cortical and subcortical brain structures in AD and bvFTD.MethodsForty-one AD, 37 bvFTD and 33 healthy controls underwent baseline MRI and of these longitudinal follow-up was obtained for 20 AD and 20 bvFTD (1 to 4 years). A total of 87 AD and 70 bvFTD consecutive scans were included in the study. The trajectories of progression in cortical and subcortical structures were identified with FreeSurfer and linear mixed effect modelling.ResultsThe results uncovered cortical and subcortical disease-specific trajectories of neurodegeneration in AD and bvFTD. Specifically, direct comparisons between patient groups revealed that over time AD showed greater cortical atrophy in the inferior parietal and posterior cingulate cortex than bvFTD. Conversely, bvFTD patients showed greater atrophy in the striatum than AD over time.ConclusionsThese results indicate that atrophy in the posterior cingulate and the striatum diverges with disease progression in these dementia syndromes and may represent a potential diagnostic biomarker for tracking rates of progression of AD and bvFTD. These findings may help inform future drug trials by identifying appropriate outcome measures to quantify drug efficacy and their ability to modulate disease progression over time.