Predicting pattern and spread of brain atrophy and tau deposition in progressive supranuclear palsy.
Predicting pattern and spread of brain atrophy and tau deposition in progressive supranuclear palsy.
Where did the research take place?
The study site has not been established. Author addresses may differ from where the research occurred.
Milan, IT · Author affiliation
Department of Neurology, Mayo Clinic, Rochester, MN, USA; Neuroimaging Research Unit, Division of Neuroscience, IRCCS San Raffaele Scientific Institute, Via Olgettina, 60, 20132 Milan, Italy; Neurology Unit, IRCCS San Raffaele Scientific Institute, Milan, Italy; Vita-Salute San Raffaele University, Milan, Italy.Location evidence
US · Author affiliation · country only
Department of Neurology, Mayo Clinic, Rochester, MN, USA; Neuroimaging Research Unit, Division of Neuroscience, IRCCS San Raffaele Scientific Institute, Via Olgettina, 60, 20132 Milan, Italy; Neurology Unit, IRCCS San Raffaele Scientific Institute, Milan, Italy; Vita-Salute San Raffaele University, Milan, Italy.Location evidence
A plain-language reading has not been prepared for this paper yet.
Original abstract
BackgroundProgressive supranuclear palsy (PSP) is a 4-repeat tauopathy characterized by clinicopathological heterogeneity. The complex interplay between tau deposition, structural changes, and disease spread are unclear.ObjectivesTo investigate the temporospatial patterns of tau propagation and neurodegeneration using multimodal imaging with MRI and flortaucipir (FTP) PET, and determine relationship with clinical heterogeneity.Methods150 PSP patients (n = 66 Richardson's syndrome [PSP-RS], n = 26 parkinsonism [PSP-P], n = 25 speech-language [PSP-SL], n = 13 progressive gait freezing (PSP-PGF), n = 10 corticobasal syndrome [PSP-CBS], n = 3 frontal, n = 1 oculomotor and n = 6 postural instability) underwent 3 T-MRI and FTP-PET. Forty-three patients died and underwent autopsy. Subtype and Stage Inference (SuStaIn), an unsupervised machine learning algorithm that separates data-driven disease phenotypes distinguished by diverse temporal progression patterns, was applied to both MRI and FTP-PET W-scores (adjusted for age, sex and scanner, using 102 controls). Longitudinal MRI (n = 76) and FTP (n = 56) data, analysed with linear mixed models, were used to assess regional progression patterns and compared with the machine learning predictions.ResultsTwo subtypes emerged across modalities. Subtype 1 exhibited initial subcortical involvement, mainly included PSP-RS and PSP-P patients and mostly featured PSP pathology, while subtype 2 exhibited early cortical involvement, PSP-SL patients and CBD pathology. FTP-PET stages preceded MRI stages, suggesting tau deposition anticipates atrophy. MRI stages were better in capturing clinical progression and predicting longitudinal disease evolution.ConclusionsThese findings suggest the existence of subcortical and cortical subtypes of PSP, with distinct clinicopathological features. Tau PET and MRI provide complementary insights into disease progression, with MRI more closely reflecting clinical evolution.