Karolinska Institutet
Browse

Magnetic resonance imaging techniques for diagnostics in Parkinson's disease and atypical parkinsonism

Download (31.66 MB)
thesis
posted on 2024-09-02, 15:40 authored by Henrik SjöströmHenrik Sjöström

Background: Parkinson’s disease (PD) is a neurodegenerative disease characterized by rigidity, hypokinesia, tremor and postural instability. PD is a clinical diagnosis based on neurological examination, patient history and treatment response. Similar symptoms can be caused by other movement disorders such as progressive supranuclear palsy (PSP) and multiple system atrophy (MSA), making it difficult to clinically separate them in early stages. However, these diseases differ in underlying pathology, treatment and prognosis. PSP and MSA have more rapid deterioration and develop additional symptoms such as impaired eye movements or autonomic dysfunction. Magnetic resonance imaging (MRI) is commonly performed as part of the clinical work-up in patients presenting with parkinsonism. There are no overt changes on structural MRI in PD. In atypical parkinsonian syndromes there are typically no visible changes until later disease stages.

Purpose: The aim of this thesis is to evaluate novel MRI techniques for diagnostics and for investigation of disease processes in Parkinson’s disease, PSP and MSA.

Paper I: A retrospective cohort from Karolinska University Hospital (102 participants; 62 PD, 15 PSP, 11 MSA, 14 controls) was assessed using susceptibility mapping processed from susceptibility weighted imaging. We show that there is elevated susceptibility in the red nucleus and the globus pallidus in PSP compared to PD, MSA and controls. Higher susceptibility levels were also seen in MSA compared to PD in the putamen, and in PD compared to controls in the substantia nigra. Using the red nucleus susceptibility as a diagnostic biomarker, PSP could be separated from PD with an accuracy of 97% (based on the area under the receiver operating characteristic curve, AUC), from MSA with AUC 75% and from controls with AUC 98%. We concluded that susceptibility changes, particularly in the red nucleus in PSP, could be potential biomarkers for differential diagnostics in parkinsonism.

Paper II: A prospective cohort from Lund, the BioFINDER study (199 participants; 134 PD, 11 PSP, 10 MSA, 44 controls), was investigated using the susceptibility mapping pipeline developed for Paper I. The finding from Paper I with elevated susceptibility in the red nucleus was validated for PSP compared to PD, MSA and controls. The elevated putaminal susceptibility was also confirmed in MSA compared to PD. The potential role of red nucleus susceptibility as a biomarker for separating PSP from PD and MSA was also similar to the results in Paper I, with AUC 98% for separating PSP from PD and AUC 96% for separating PSP from MSA. We concluded that we could confirm our previous findings from Paper I, with the red nucleus susceptibility being a potential biomarker for separating PSP from PD and MSA.

Paper III: A retrospective cohort from Karolinska University Hospital (196 participants; 140 PD, 29 PSP, 27 MSA) was evaluated to employ automated volumetric brainstem segmentation using FreeSurfer. The volumetric approach was compared to manual planimetric measurements: midbrain-pons ratio, magnetic resonance parkinsonism index 1.0 and 2.0. Intra- and inter-scanner as well as intra- and inter-rater reliability were calculated. We found good repeatability in both automated volumetric and manual planimetric measurements. Normalized midbrain volume performed better than the planimetric measurements for separating PSP from PD. We concluded that, if further developed and incorporated in a radiology workflow, automated brainstem volumetry could increase availability of brainstem metrics and possibly save time for radiologists conducting manual measurements.

Paper IV: Two cohorts, a retrospective from Karolinska University Hospital (184 participants; 129 PD, 28 PSP, 27 MSA) and a prospective from Lund (185 participants; 125 PD, 11 PSP, 8 MSA, 41 controls), were studied to investigate a new method of creating T1-/T2-weighted ratio images and its diagnostic capabilities in differentiating parkinsonian disorders. In the explorative retrospective cohort, differences in white matter normalized T1-/T2- weighted ratios were seen in the caudate nucleus, putamen, thalamus, subthalamic nucleus and red nucleus in PSP compared to PD; in the caudate nucleus and putamen in MSA compared to PD and in the subthalamic nucleus and the red nucleus in PSP compared to MSA. These differences were validated externally in the prospective cohort, where the changes could be confirmed in the subthalamic nucleus and the red nucleus in PSP compared to PD and MSA. We concluded that there are different patterns of white matter normalized T1-/T2-weighted ratio between the disorders and that this reflects differences in underlying pathophysiology. The T1-/T2-weighted ratio should be further investigated for better understanding of pathological processes in parkinsonian disorders and could possibly be utilized for diagnostic purposes if further developed.

List of scientific papers

I. Quantitative susceptibility mapping differentiates between parkinsonian disorders. Sjöström H, Granberg T, Westman E, Svenningsson P. Parkinsonism & Related Disorders. 2017 Nov;44:51–57.
https://doi.org/10.1016/j.parkreldis.2017.08.029

II. Mapping of apparent susceptibility yields promising diagnostic separation of progressive supranuclear palsy from other causes of parkinsonism. Sjöström H, Surova Y, Nilsson M, Granberg T, Westman E, van Westen D, Svenningsson P, Hansson O. Scientific Reports. 2019 Apr 15;9(1):6079.
https://doi.org/10.1038/s41598-019-42565-4

III. Automated brainstem volumetry can aid in the diagnostics of parkinsonian disorders. Sjöström H, Granberg T, Hashim F, Westman E, Svenningsson P. [Submitted]

IV. Differences in neurodegeneration in Parkinsonian syndromes revealed by T1/T2-weighted ratio imaging in two large cohorts. Sjöström H, van Westen D, Hall S, Tjerkaski J, Westman E, Muehlboeck S, Hansson O, Svenningsson P, Granberg T. [Submitted]

History

Defence date

2020-06-12

Department

  • Department of Clinical Neuroscience

Publisher/Institution

Karolinska Institutet

Main supervisor

Svenningsson, Per

Co-supervisors

Granberg, Tobias; Westman, Eric

Publication year

2020

Thesis type

  • Doctoral thesis

ISBN

978-91-7831-789-9

Number of supporting papers

4

Language

  • eng

Original publication date

2020-05-20

Author name in thesis

Sjöström, Henrik

Original department name

Department of Clinical Neuroscience

Place of publication

Stockholm

Usage metrics

    Theses

    Categories

    No categories selected

    Keywords

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC