Molecular imaging of tau in the pathological cascade of Alzheimer’s disease

Abstract

The pathology of Alzheimer’s disease (AD) is characterised by the misfolding and aggregation of amyloid-β (Aβ) into extracellular plaques and aggregation of tau into intracellular neurofibrillary tangles. Recent advances in molecular imaging have allowed the development of positron emission tomography (PET) tracers for the in vivo detection of Aβ plaques while current efforts focus on the evaluation of recently proposed tracers targeting tau pathology. This thesis is composed of three main parts. Part one compares two Aβ PET tracers ([11C]PIB and [18F]florbetapir) when administered to different but matched patient cohorts, and explores the effect of age on the distribution of Aβ-positive PET scans. Part two focuses on the first in vivo evaluation of the tau-specific tracer [18F]THK5317, using a longitudinal multi-modal design, in a sample of cognitively normal volunteers, patients at different clinical stages of AD and individual patients with atypical parkinsonism. The third part describes the direct in vivo comparison of the binding properties of two tau-specific tracers ([11C]THK5351 and [11C]PBB3) when injected into the same patients with AD on the same day.

The results indicated that, firstly, the binding of the Aβ-specific PET tracers, [11C]PIB and [18F]florbetapir, was highly comparable in individuals from different cohorts. Furthermore, age plays an important role in the distribution of Aβ-positive PET scans, with the oldest old patients with cognitive complaints appearing to benefit substantially from clinical assessment with Aβ PET. Secondly, the tracer [18F]THK5317 detected the expected load and regional distribution of tau pathology in vivo in a sample of patients with AD and patients with atypical parkinsonism. The distribution of [18F]THK5317 binding differed from that of Aβ deposition in patients with AD, although regional correlations existed, indicating areas where Aβ and tau pathologies were co-located. The regional load of tau pathology ([18F]THK5317) was associated with measures of global cognition and episodic memory, with local hypometabolism playing a mediating role in this relationship. Longitudinally, a heterogeneous pattern of changes was observed in the binding of the tau tracer [18F]THK5317 in patients with AD, in contrast to the homogeneous changes in glucose metabolism that better tracked cognitive deterioration.

The build-up of tau pathology ([18F]THK5317) and the development of local hypometabolism appeared temporally dissociated, with a stronger relationship detected between the two when hypometabolism changes became more prevalent, in the later stages of the disease. Finally, different tau-specific tracers ([11C]THK5351 and [11C]PBB3) seemed to bind in vivo to different molecular targets; [11C]PBB3 binding appeared to correlate closer to Aβ deposition, while [11C]THK5351 binding followed the expected regional pattern of tau pathology in AD and related closer to downstream markers of the disease. Further investigation of the existing PET tracers and development of new tracers is required for shedding light on the pathological processes that contribute to neurodegeneration in AD and for developing clinical markers that allow early and highly accurate discrimination between different proteinopathies.