The Microstructural Basis of REM Sleep Behaviour Disorder

The Microstructural basis of REM Sleep Behaviour Disorder. Project proposal PhD Project Dr Hans Odd (Primary Supervisor Dr Christian Lambert, Wellcome Centre for Human Neuroimaging, UCL, Secondary Supervisor John Ashburner, UCL, ION). REM Sleep Behaviour Disorder (RBD) is a disorder of sleep where muscle atonia normally associated with the REM sleep phase fails, leading to dream enactment behaviour. This phenomenon is strongly prognostic for several neurodegenerative disorders, most commonly ⍺-synuclein mediated diseases including Parkinson’s disease (PD), dementia with Lewy bodies (DLB) and multiple system atrophy (MSA). Whilst >90% of individuals with RBD will ultimately convert to one of these conditions, this may range from a few years to many decades following symptom onset. At present it is not possible to predict who will convert, to what disease, or over what timeframe. Furthermore, the precise pathophysiological basis in humans remains unclear. This PhD project is based at the Wellcome Institute for Human Neuroimaging, Queen Sq, UCL, London. My primary supervisor is Dr Christian Lambert, the principal investigator for the “Quantitative MRI for Anatomical Phenotyping in Parkinson’s disease” study (qMAP-PD). This is a longitudinal observational study based at the Wellcome Centre for Human Neuroimaging (WCHN) at University College, London. It includes a group of participants with polysomnographically (PSG) diagnosed RBD who have undergone deep phenotyping (including detailed clinical assessments, cognitive battery, multi-domain symptom scales) and other investigations (quantitative MRI, serum biomarkers, genetics, 7-day actigraphy) to better understand the causes of variability between individuals with Parkinson’s disease, and other ⍺-synucleinopathies. The aim of my thesis is to dissect the microstructural basis of REM Sleep Behaviour Disorder in vivo. Specifically, I eventually aim to answer: A. What are the neural correlates underpinning idiopathic RBD in vivo? B. What are the di]erent patterns of tissue damage in iRBD, and do these correlate to di]erences in clinic phenotype? C. Can we track RBD disease progression in vivo using qMRI and is this linked to the baseline patterns of damage? There is a wide scientific consensus that the nuclei responsible for normal sleep architecture are found within the pons, along with involvement of the medulla. However, this is primarily due to research in animal lesion models, and their correlates in humans are poorly understood. To elucidate these three questions, I propose the following programme of research to be enable me to answer them. 1. By performing a literature search using the terms ‘RBD’, ‘REM Sleep Disorder’, ‘sleep disturbance’ ‘lesion’ ‘pontine’ and ‘pons’ to find cases of secondary RBD, that is to say, RBD caused by an identifiable lesions within the brain of any type, and from which the anatomical extent of the lesions are demonstrated or described su]iciently for comparison. 2. These lesions would then be manually segmented using the ITK-SNAP software package. These segmentations would be mapped to a common stereotactic space, namely an MNI-152 1mm T1w MRI template. This methodology’s robustness would be confirmed using a blinded test-retest process to constrain and minimise inter-operator variation. 3. These individual lesions would be averaged and thresholded then normalised to create a lesion probability map. From this normalised map I intend to find the largest ROIs in terms of highest number of contiguous voxels. 4. Using these ROIs, I would then perform tractography on one of the Human Connectome Project’s open-source data sets, namely HCP 100, which is made of up of 100 healthy human control volunteers. This tractographic study would enable me to determine whether a candidate network passes through these ROIs connecting plausible regions associated with a REM-sleep associated atonia network and the pontine lesions found in the literature search. 5. Should a plausible candidate network be found, I then plan to use it to explore the microstructural integrity of white matter tracts of subjects from the qMAP-PD cohort, using voxel-based morphometry (VBM), and voxel-based quantification (VBQ), to identify microstructural any changes in the 40+ participants with PSG confirmed RBD. To enable this, each member of the qMAP cohort underwent a one-hour MR protocol which included: • 0.8mm isotropic qMRI multiparameter mapping sequences (MPMS), providing quantitative magnetization transfer (MT) proton density(A), R1 and R2* maps with prospective motion correction. • Multishell high angular resolution di]usion weighted imaging (DWI) • Resting state fMRI (rsfMRI) 6. There are several di]erent potential patterns of ⍺-synuclein pathology that could plausibly underpin RBD. One mechanism is focal damage to small pontine nuclei known to be involved in sleep regulation and which are known to be disproportionately a]ected by ⍺-synuclein pathology (notably the locus coeruleus/peri-coeruleus region). It is also conceivable that the pattern of disease may be characterised by di]use levels of WM damage, which would explain the wide variety of clinical presentations of RBD patients, i.e. di]erent clinical phenotypes of RBD disease. 7. The possibility that these di]erent phenotypes might be associated with specific patterns of disease prodrome demands an approach that can measure the association of specific neuronal/WM patterns of damage with specific symptoms. I propose to analyse this putative association by using a multivariate analysis approach such as canonical correlation analysis (CCA) to establish whether these varying patterns of disease may explain di]erent phenotypes of disease. I believe that this proposed research pipeline applied to the data sets acquired by the qMAP-PD project will yield valuable insights into the microstructural basis of RBD and provide tools to reliably and quantifiably establish the presence of RBD, o]er insights into disease prognosis that will benefit patients and clinicians alike.