Cognitive impairment and conversion to dementia in Parkinson's disease: an imaging study of amyloid PET and diffusion MRI

Objective To investigate the association between amyloid deposition or white matter degeneration with cognitive impairment and conversion to dementia in Parkinson’s disease (PD). Background Cognitive decline and dementia are common in Parkinson's disease, however the pathophysiological basis of cognitive impairment in PD is unresolved. The time-course from diagnosis to development of dementia is highly variable, and imaging biomarkers are urgently needed to assist estimation of long-term cognitive outcomes, and enable targeted therapeutic interventions in early disease. Misfolded beta-amyloid protein aggregates, or amyloid plaques, are a significant pathology in Alzheimer's disease, and may play a part in future cognitive decline in PD. Measures of cerebral blood flow and white matter micro- and macro-structural degeneration may correlate more directly with current cognitive impairment, and may interact with amyloid to affect cognitive decline in PD. This thesis comprehensively investigates these measures in PD and aims to differentiate pathology and age-related effects, using both cross-sectional and longitudinal (three-year) study designs. Methods We acquired [18F]-Florbetaben (FBB) amyloid PET, arterial spin labelling perfusion MRI, and structural MRI in 115 patients with Parkinson’s disease, recruited from the Movement Disorders clinic at the New Zealand Brain Research Institute. Movement Disorders Society level II criteria were used to classify PD patients as having normal cognition (PDN, n=23), mild-cognitive impairment (PD-MCI, n=76), or dementia (PDD, n=16), at study baseline and over the course of three-year neuropsychological follow-up. The relationship between amyloid deposition and cognitive classification, global cognitive ability, cerebral blood flow, and conversion to dementia during the three-year follow-up was assessed using both Bayesian regression and whole-brain voxel-wise analysis. High angular resolution diffusion imaging (HARDI) MRI was acquired in a wider cohort of 123 PD participants and 37 controls, for the investigation of white matter integrity in PD across the cognitive spectrum (PDN, n=46; PD-MCI, n=66; PDD, n=11). Cross-sectional fixel-based analysis investigated measures of fibre density, fibre cross-section, and combined fibre-density-cross-section across PD, controls, and PD cognitive subgroups. Clinical measures of global cognitive ability and motor impairment were also assessed for association with these fixel-based metrics. Results We observed significantly higher cortical amyloid accumulation in our PDD group relative to other cognitive subgroups, but model comparison indicated this was due to an effect of age. Cortical amyloid was seen to be present in PD at levels comparable to healthy ageing. Longitudinal assessment identified that, while increased cortical and subcortical amyloid was associated with conversion to dementia within three years, this did not represent a clinically relevant effect. There was no evidence of an interactive effect of amyloid with cerebral blood flow to affect cognition. Reduced fibre density in the substantia nigra correlated with disease, however age exhibited the most widespread association with white matter metrics in this cohort. Conclusions This thesis investigated the effect of amyloid on cognitive impairment within a large, well-characterised, longitudinal PD cohort, and subsequently challenged existing characterisations of regional amyloid deposition relating to cognitive decline. This work also represents the largest application to-date of fixel-based analysis for the investigation of white matter degeneration across the cognitive spectrum in PD