Upload
others
View
8
Download
0
Embed Size (px)
Citation preview
Neural Substrates of Parkinson’s Disease
by
Yuko Koshimori
A thesis submitted in conformity with the requirements
for the degree of Doctor of Philosophy
Institute of Medical Science
University of Toronto
© Copyright by Yuko Koshimori 2016
ii
Neural Substrates of Parkinson’s Disease
Yuko Koshimori
Doctor of Philosophy
Institute of Medical Science
University of Toronto
2016
Abstract
Parkinson’s disease is characterized by selective degeneration of dopaminergic neurons in the
substantia nigra, the presence of Lewy bodies and Lewy neurites containing alpha-synuclein
aggregates in such neurons, and neuroinflammation mediated by the activation of microglia and
astrocytes. These pathological changes collectively contribute to the cardinal motor symptoms of
Parkinson’s disease such as resting tremor, bradykinesia, rigidity, and postural instability. In
addition, dysfunction of non-dopaminergic neurons such as serotonergic, noradrenergic, and
cholinergic neurons, the Lewy pathology, and neuroinflammation outside of the nigro-striatal
regions are thought to play a major role in non-motor symptoms of Parkinson’s disease such as
psychiatric problems, cognitive dysfunction including dementia, autonomic symptoms, and
rapid-eye movement sleep disorder. Therefore, Parkinson’s disease is a multisystem disease,
affecting diffuse area of brain and producing various symptoms. Current trends toward use of
data driven analytic methods can facilitate to expand our knowledge about Parkinson’s disease
beyond the classic basal ganglia motor model and DA dysfunction. The general aim of this thesis
is to investigate a whole brain in vivo to further understand the complex neural substrates of the
iii
disease using multimodal neuroimaging tools and techniques. This thesis has shed light on the
cortical and subcortical structural and functional brain changes underpinning for the clinical
manifestations in Parkinson’s disease. Changes in gray and white matter as well as functional
changes were evident using magnetic resonance imaging (MRI). Gray matter and functional
changes occurred in the frontal cortex while white matter changes occurred more diffuse area of
the brain including bilateral frontal and temporal regions and left parietal and occipital regions as
well as subcortical white matter tracts. These changes were also associated with motor and
cognitive dysfunction, suggesting that structural MRI and diffusion tensor imaging as well as
resting functional MRI may be promising as in vivo biomarkers for Parkinson’s disease. Our
investigation of neuroinflammation using Positron Emission Tomography and an improved
second generation radioligand did not find any changes in the striatum of patients with
Parkinson’s disease. However, demonstrated gray matter, white matter, and functional changes
warrant further investigations of neuroinflammation in extra-striatal and white matter regions.
iv
Acknowledgements
My PhD experiences supported by amazing individuals have inspired me to further pursuit a
career in brain research. Foremost, I would like to thank my supervisor, Dr. Antonio Strafella
who provided me with this invaluable opportunity to completely my PhD degree in his lab.
Antonio is a role mode for a research scientist, supervisor, and a personal figure. He has been a
reliable and responsible supervisor, providing me with continued solid guidance. I am inspired
by his creative and passionate mind, diligence, positive attitude, warmth, and kindness. I greatly
appreciate his mentorship throughout my PhD years.
I would like to acknowledge my other mentors, Dr. Romina Mizrahi and Dr. Clement Hamani
who have provided me with indispensable feedback, guidance, and research ideas. Thank you to
Dr. Pablo Rusjan for teaching me the fundamentals of PET analyses and assisting me with the
analyses and to Dr. Nancy Lobaugh for her guidance for image quality control and imaging
analyses of MRI. Thanks to Dr. Ariel Graff for his time and effort as an internal examiner.
I would like to thank all of my current and past lab members who have created such a supportive
and enjoyable learning environment over the past years. Because of their knowledge, skills,
kindness, and friendship, my PhD experiences have been incredibly positive and inspiring.
Thanks to Leigh Christopher for providing continued support, laughs, and friendship throughout
my PhD years. Thanks to Sang Soo Cho for her invaluable support and advice, teaching me
every aspect of research. I would also like to acknowledge Marion Criaud, Mark Jacobs, Sarah
Coakley, and Crystal Li for their help and friendship. Thank you to Christine Ghadery, Alex
Mihaescu, and Jin-hee Kim for their warm support and friendship. Thank you to Ji Hyun Ko who
first taught me the PET analysis and provided me continued support throughout the project,
Giovanna Pellachia and Barbara Segura for their help with imaging data processing and analyses,
Kelly Aminian, and Rostom Mabrouk for their support and kindness. I would also like to thank
my friends I made through my graduate studies, Danielle DeSouza, Ivonne Suridjan, Aaron
Kucyi, and Shinichiro Nakajima for all of our shared experiences, friendship and laughter that I
will treasure.
v
Thank you to Alvina Ng, Laura Nguyen, Anusha Ravichandran, Natalie Freeman, Sajid Shaikh,
Jun Parkes, and Irina Vitcu for all of the research and technical support. Thank you to Dr.
Anthony Lang and Dr. Sylvain Houle for the incredible opportunities and variable input.
I would like to acknowledge the Canadian Institute of Health Research and Parkinson Society
Canada, and Society for Neuroscience for providing financial support for this thesis work and for
providing opportunities to present the data, learn most updated research findings, and interact
with other fellow researchers. I also deeply appreciate time and effort of all the research
participants and their family members.
Lastly, I would like to extend my gratitude to my family in Japan and friends in Canada and
Japan. Thanks to my loving, supportive, and understanding mother and father. They encouraged
me to work hard, persevere, and achieve my goal. I also thank all of my friends for laughs,
kindness, and support over the years.
Without all of these people, my PhD would not have been possible. Thank you very much.
vi
Contributions
Yuko Koshimori (author) was the sole author of this thesis. All work required for this thesis,
including the planning, writing, analysis and research was performed either in whole or in part by
the author. Contributions from other individuals are acknowledged:
Dr. Antonio Strafella (Supervisor) – provided guidance in planning and execution of projects;
laboratory resources; guidance in interpretation of results and preparation of manuscripts
Dr. Romina Mizrahi – provided mentorship and guidance in interpretation of results
Dr. Clement Hamani – provided mentorship and guidance in interpretation of results
Dr. Leigh Christopher – provided assistance with data collection and experiments and
preparation of manuscripts for Chapters 3, 4 & 5
Dr. Anthony Lang – provided guidance in the write-up of the manuscripts for Chapters 3, 4 & 5
Dr. Sylvain Houle – provided guidance in the write-up of the manuscripts for Chapters 3, 4 & 5
Dr. Sarah Duff-Canning – provided guidance for all of the neuropsychological testing for
Chapters 3
Dr. Barbara Segura – provided assistance with data analysis for Chapters 3
Dr. Nancy Lobaugh – provided guidance with the MRI imaging methods for Chapter 3
Kelly Aminian – provided assistance with preparation of manuscripts for Chapter 3
vii
Mark Jacobs - provided assistance with preparation of manuscripts for Chapter 4 and 5
Dr. Sang-Soo Cho – provided assistance with data collection and experiments; preparation of
manuscripts for Chapters 4
Dr. Marion Criaud – provided assistance with image quality control for Chapters 4
Dr. Christine Ghadery – provided assistance with preparation of manuscripts for Chapters 4
Sarah Coakeley – provided assistance with preparation of manuscripts for Chapters 4
Madeleine Harris - provided assistance with preparation of manuscripts for Chapters 4
Dr. Ji-Hyun Ko – provided assistance with data collection and experiments and preparation of
manuscripts for Chapters 5
Dr. Pablo Rusjan – provided guidance with the PET imaging analysis for Chapters 5
Dr. Rostom Mabrouk - assistance with the PET imaging analysis for Chapters 5
Dr. Alan Wilson – provided assistance with experiments for Chapter 5
viii
Table of Contents
Acknowledgements ............................................................................................................ iv
Contribution ....................................................................................................................... vi
Table of Contents ............................................................................................................. viii
List of Tables ................................................................................................................... xvi
List of Figures ................................................................................................................. xvii
List of Abbreviations ....................................................................................................... xix
1.0 Literature Review .......................................................................................................... 1
1.1 Parkinson’s disease ................................................................................................ 1
1.2 Clinical symptoms of Parkinson’s disease ............................................................. 2
1.3 Neuropathology of Parkinson’s disease ................................................................. 3
1.3.1 Dysfunction in dopamine pathways and basal ganglia-thalamocortical
circuits ................................................................................................................... 3
1.3.2 Dysfunction of non-dopaminergic systems ................................................. 5
1.3.3 Lewy pathologies and alpha-synuclein ........................................................ 7
1.3.4 Neuroinflammation .................................................................................... 10
1.3.4.1 Microglia.......................................................................................... 10
1.3.4.2 Astrocytes ........................................................................................ 11
1.3.4.3 Neuroinflammation in aging condition ............................................ 12
1.3.4.4 Neuroinflammation in Parkinson’s disease ..................................... 13
ix
1.3.4.5 Potential mechanisms underlying for neuroinflammation in
Parkinson’s disease ...................................................................................... 14
1.3.5 Summary of Parkinson’s disease ............................................................... 15
1.3.6 Potential mechanisms underlying neuroimaging changes in Parkinson’s disease
............................................................................................................................ 16
1.4 Magnetic Resonance Imaging .............................................................................. 18
1.4.1 Principles of Magnetic Resonance Imaging .............................................. 18
1.4.2 Gray matter imaging .................................................................................. 19
1.4.2.1 Voxel based morphometry ............................................................... 19
1.4.2.2 Cortical thickness analysis ............................................................... 20
1.4.3 White matter imaging ................................................................................ 22
1.4.3.1 Diffusion-weighted imaging/Diffusion tensor imaging................... 23
1.4.3.2 Diffusion tensor tractography .......................................................... 27
1.4.3.3 Tract-based Spatial Statistics ........................................................... 28
1.4.4 Functional MRI ............................................................................................. 29
1.4.4.1 Blood oxygen level-dependent signal .............................................. 29
1.4.4.2 Resting state functional MRI ........................................................... 30
1.4.4.3 Seed-based functional connectivity analysis ................................... 30
1.4.4.4 Independent Components Analysis ................................................. 31
1.4.4.5 Graph theoretical analysis................................................................ 31
1.4.4.5.1 Processing and analysis steps ................................................ 32
x
1.4.4.5.2 Graph measures ..................................................................... 33
1.5 Positron Emission Tomography ........................................................................... 36
1.5.1 Principles of Positron Emission Tomography imaging ............................. 36
1.5.2 PET radioligands ....................................................................................... 38
1.5.3 Quantification of PET radioligand binding in vivo ................................... 38
1.5.3.1 Kinetic modeling of time-activity curve .......................................... 38
1.5.3.2 Two-tissue compartment model and its outcome measures ............ 39
1.5.3.3 One-tissue compartment model ....................................................... 40
1.5.3.4 A simplified reference tissue model ................................................ 41
1.6 Neuroimaging studies in Parkinson’s disease ...................................................... 42
1.6.1 Structural changes in Parkinson’s disease ................................................. 42
1.6.2 Brain network changes in Parkinson’s disease using rsfMRI and graph
theoretical approach ............................................................................................ 46
1.6.3 Neuroinflammation in PD using Translocator protein 18 kDa imaging .... 48
1.6.3.1 Translocator protein 18 kDa ............................................................ 48
1.6.3.2 Prototypical TSPO radioligand: [11C]-PK11195 ............................. 50
1.6.3.3 Second-generation TSPO radioligands ............................................ 50
1.6.3.4 TSPO polymorphism and binding affinity of second generation TSPO
radioligands.................................................................................................. 52
1.6.3.5 [18F]-FEPPA .................................................................................... 54
1.6.3.5.1 Evaluations in vitro and ex vivo of [18F]-FEPPA .................. 54
xi
1.6.3.5.2 Evaluation of quantification in vivo of [18F]-FEPPA in human
gray matter ............................................................................................ 55
1.6.3.5.3 Evaluation of quantification in vivo of [18F]-FEPPA in human
white matter .......................................................................................... 56
1.6.3.5.4 TSPO polymorphism and [18F]-FEPPA VT ........................... 56
1.6.3.5.5 [18F]-FEPPA study on neuroinflammation and aging ........... 57
1.6.3.5.6 [18F]-FEPPA studies in clinical populations .......................... 57
1.6.3.5.7 TSPO imaging in Parkinson’s disease ................................... 58
2.0 Aims and Hypotheses ................................................................................................. 60
2.1 Study 1: Imaging changes associated with cognitive abnormalities in Parkinson’s
disease ........................................................................................................................ 60
2.2 Study 2: Disrupted nodal and hub organization account for abnormalities across
brain networks and clinical manifestations of Parkinson’s disease ........................... 61
2.3 Study 3: Imaging striatal microglial activation in patients with Parkinson’s
disease ........................................................................................................................ 61
3.0 Study 1: Imaging changes associated with cognitive abnormalities in Parkinson’s
disease ............................................................................................................................... 62
3.1 Introduction .......................................................................................................... 62
3.2 Methods ............................................................................................................... 65
3.2.1 Participants ................................................................................................ 65
xii
3.2.2 Neuropsychological Assessment ............................................................... 65
3.2.3 MRI Acquisition ........................................................................................ 66
3.2.3.1 Cortical Thickness Analysis ............................................................ 67
3.2.3.2 Subcortical Volume Analysis .......................................................... 67
3.2.3.3 TBSS ................................................................................................ 68
3.2.3.4 Statistical Analysis .......................................................................... 68
3.3 Results .................................................................................................................. 69
3.3.1 Demographic, Clinical, and Cognitive Characteristics .............................. 69
3.3.2 Cortical Thickness ..................................................................................... 71
3.3.3 Subcortical Volume ................................................................................... 74
3.3.4 White matter .............................................................................................. 74
3.4 Discussion ............................................................................................................ 77
4.0 Study 2: Disrupted nodal and hub organization account for brain network abnormalities in
Parkinson’s disease ........................................................................................................... 82
4.1 Introduction .......................................................................................................... 82
4.2 Methods ............................................................................................................... 85
4.2.1 Participants ................................................................................................ 85
4.2.2 MRI image acquisition .............................................................................. 85
4.2.3 rsfMRI preprocessing ................................................................................ 86
4.2.4 Network nodes ........................................................................................... 86
4.2.5 Computation of network measures ............................................................ 90
xiii
4.2.6 Seed-based functional connectivity analysis ............................................. 91
4.2.7 Statistical analysis ...................................................................................... 91
4.3 Results .................................................................................................................. 92
4.3.1 Demographic and clinical characteristics .................................................. 92
4.3.2 Head motion parameters and outliers ........................................................ 94
4.3.3 Group comparisons of graph measures ...................................................... 95
4.3.4 Hub analysis ............................................................................................... 99
4.4 Discussion .......................................................................................................... 102
4.4.1 Changes in the sensorimotor network ...................................................... 102
4.4.2 Changes in the cognitive subnetworks .................................................... 103
4.4.3 Hub reorganization .................................................................................. 104
4.4.4 Conclusions .............................................................................................. 106
5.0 Study 3: Imaging striatal microglial activation in patients with Parkinson’s
disease ............................................................................................................................. 107
5.1 Introduction ........................................................................................................ 107
5.2 Methods ............................................................................................................. 109
5.2.1 Subjects .................................................................................................... 109
5.2.2 PET data acquisition ................................................................................ 109
5.2.3 MRI acquisition ....................................................................................... 110
5.2.4 Input function measurement .................................................................... 111
xiv
5.2.5 Generation of ROI-based time activity curve ......................................... 111
5.2.6 Kinetic analysis ........................................................................................ 112
5.2.7 DNA extraction and polymorphism genotyping ...................................... 112
5.2.8 Statistical analysis .................................................................................... 113
5.3 Results ................................................................................................................ 113
5.3.1 Demographic and clinical characteristics ................................................ 113
5.3.2 Genotype and disease effects on TSPO binding ...................................... 114
5.4 Discussion .......................................................................................................... 117
6.0 General Discussion ................................................................................................... 120
6.1 Summary of Findings ......................................................................................... 120
6.2 Structural and functional changes in sensorimotor nodes in Parkinson’s
disease ...................................................................................................................... 121
6.3 Structural and functional changes in cognitive nodes in Parkinson’s disease ... 123
6.4 Dopaminergic modulation on resting-state functional connectivity .................. 126
6.5 White matter changes in Parkinson’s disease .................................................... 128
6.6 Neuroinflammation in Parkinson’s disease ....................................................... 130
6.6.1 Interindividual variability in total distribution volume .............................. 130
6.6.2 [18F]-FEPPA binding .................................................................................. 132
6.6.3 Neuroinflammation in Parkinson’s disease ................................................ 132
6.7 Neuroimaging biomarkers for Parkinson’s disease ........................................... 133
xv
6.8 Limitations ................................................................................................................ 135
7.0 Conclusion ................................................................................................................ 137
8.0 Future Directions ...................................................................................................... 138
References ....................................................................................................................... 144
xvi
List of Tables
Table 3-1. Demographic, clinical, and cognitive characteristics of PD patients and HCs
........................................................................................................................................... 70
Table 4-1. Node labels of four subnetworks and their MNI coordinates .......................... 88
Table 4-2. Demographic and clinical characteristics of patients with PD and HCs ......... 93
Table 4-3. Head motion parameters and outliers of patients with PD and HCs ............... 94
Table 4-4. Graph measure changes in patients with PD compared with
HCs ................................................................................................................................... 97
Table 4-5. FC of three nodes that showed group differences ........................................... 98
Table 4-6. Hub regions in HCs and patients with PD ..................................................... 101
Table 5-1. Demographic and clinical characteristics and PET measures of HC subjects and PD
patients ............................................................................................................................ 114
Table 5-2. Demographic and clinical characteristics of PD patients .............................. 114
xvii
List of Figures
Figure 1-1. BG thalamocortical model in normal state and parkinsonian state .................. 4
Figure 1-2. Three main cortico-BG thalamocortical circuits: motor, associative, and limbic
circuits. ................................................................................................................................ 5
Figure 1-3. Affected various neurotransmitter systems in PD ............................................ 6
Figure 1-4. Six stages of spread of α-synuclein inclusions across brain regions in idiopathic
PD ....................................................................................................................................... 9
Figure 1-5. Neuronal transmission of α-synuclein aggregates ........................................... 9
Figure 1-6. PD pathology. ................................................................................................. 16
Figure 1-7. Estimations of CTh using slice data. .............................................................. 21
Figure 1-8. Measurement of CTh. ..................................................................................... 22
Figure 1-9. Anisotropic diffusion in white matter ............................................................ 26
Figure 1-10. Spherical tensor shape with three orthogonal eigenvectors and their associated
eigenvalues (λ1, λ2, λ3) .................................................................................................... 26
Figure 1-11. Spherical tensor shapes illustrating various anisotropic degrees ................. 27
Figure 1-12. Mean FA skeleton projected on a skull-stripped DWI image. ..................... 29
Figure 1-13. Graph measures. ........................................................................................... 35
Figure 1-14. Principles of PET ......................................................................................... 37
Figure 1-15. 2TCM. .......................................................................................................... 40
Figure 1-16. 1TCM ........................................................................................................... 40
xviii
Figure 1-17. SRTM. .......................................................................................................... 41
Figure 3-1. a. Cortical areas showing significant cortical thinning in PD patients compared to
HCs. b. Bar graphs on extracted CTh values (mm) from the significant clusters in the left SFG
and left precentral gyrus between HCs and PD patients ................................................... 73
Figure 3-2. Partial correlations with age as a covariate between SFG thickness (extracted from
the significant cluster) and global composite z (left) and executive composite z (right) in 26 PD
patients showing the significant positive correlations.. .................................................... 74
Figure 3-3. a. Clusters of significantly increased MD in 16 patients with PD compared with 15
HCs in TBSS. b. Bar graphs on mean MD derived from the significant clusters in TBSS between
HC and PD groups ... ........................................................................................................ 76
Figure 4-1. Medial and lateral views of brain images with 120 regions of interest of four
subnetworks ...................................................................................................................... 87
Figure 4-2. Dorsal view of brain image presenting four nodes showing group differences
........................................................................................................................................... 96
Figure 4-3. Caudal view of brain image presenting 10 hub nodes.... ............................. 100
Figure 5-1. Graphs of PVEC VT in the CN and in the putamen.... ................................. 116
Figure 5-2. Graphs of VT in the CN and in the putamen.... ............................................ 116
Figure 6-1. Neuroimaging biomarkers for Parkinson’s disease.... .................................. 135
xix
List of Abbreviations
%COV Percent coefficient of variation
1H Hydrogen
1HMRS Proton magnetic resonance spectroscopy
1TCM One-tissue compartment model
2D Two-dimensional
2TCM Two-tissue compartment model
3D Three-dimensional
α-synuclein alpha-synuclein
ACC Anterior cingulate cortex
ACR Anterior corona radiata
AD Alzheimer’s disease
ADC Apparent diffusion coefficient
AIC Akaike information criterion
AIns Anterior insula
ALIC Anterior limb of internal capsule
ANOVA Analysis of variance
ASSET Array spatial sensitivity encoding technique
ATR Anterior thalamic radiation
AUC Area under the curve
BA Brodmann area
BBB Blood-brain barrier
BC Betweenness centrality
BDI Beck depression inventory
BET Brain extraction tool
BG Basal ganglia
BIC Bayesian information criterion
BOLD Blood oxygen level dependent
BPND Binding potential non-displaceable
CAMH Centre for addiction and mental health
xx
CBR Central benzodiazepine receptors
CC Corpus callosum
CF Concentration of free radioligand in tissue
CN Caudate nucleus
CND Concentration of non-displaceable radioligand in tissue
CNS Central nervous system
CNS Concentration of non-specifically bound radioligand in tissue
CP Concentration of radioactivity in blood plasma
CS Concentration of specifically bound radioligand in tissue
CSF Cerebrospinal fluid
CTA Cortical thickness analysis
CTh Cortical thickness
CT Concentration of radioactivity in tissue
CVLT California verbal learning and memory test
D Effective diffusion tensor
D2 receptor Dopamine D2 receptor
DA Dopamine
DAD Disability assessment for dementia
DAN Dorsal attention network
DBI Diazepam binding inhibitor
D-KEFS Delis-kaplan executive function system
DLB Dementia with Lewy bodies
DLPFC Dorsolateral prefrontal cortex
DMN Default mode network
DRT Dopamine replacement therapy
DTI Diffusion tensor imaging
DWI Diffusion-weighted imaging
EC External capsule
EEG Electroencephalogram
FA Fractional anisotropy
FC Functional connectivity
FLIRT FSL linear registration tool
xxi
fMRI Functional magnetic resonance imaging
FNIRT FSL non-linear registration tool
FOV Field of view
fp Plasma free fraction
FPN Fronto-parietal network
FSL FMRIB software library
FWHM Full width half maximum
GABA γ-aminobutyric acid
GAT Graph-theoretical analysis toolbox
GE General electric
GFAP Glial fibrillary acid protein
GLM General linear model
Gln Glutamine
GM Gray matter
Glu Glutamate
GP Globus pallidus
GPi Globus pallidus pars interna
GTM Geometric transfer matrix
HAB High affinity binder
Hb Hemoglobin
HC Healthy control
HC-HAB Healthy control with high affinity binder
HC-MAB Healthy control with mixed-affinity binder
HIV Human immunodeficiency virus
HRRT High resolution research tomograph
Iba1 Ionized calcium binding adaptor molecule 1
ICA Independent components analysis
ICBM International consortium for brain mapping
ICC Interrater correlation coefficient
ICD Impulse control disorders
ICV Intracranial volume
IFO Inferior fronto-occipital fasciculus
xxii
IL Interleukin
ILF Inferior longitudinal fasciculus
Ins Insula
JHU Johns Hoskins University
JLO Judgment of Line Orientation
keV Kilo-electron volts
LAB Low affinity binder
LDL Low-density lipoprotein
LFP Local field potential
LEDD Levodopa equivalent daily dose
M1 Microglia M1
M1 Primary motor cortex
M2 Microglia M2
MAB Mixed affinity binder
MCI Mild cognitive impairment
MD Mean diffusivity
MDE Major depressive episodes
MDS Movement disorders society
MEG Magnetoencephalography
MFG Middle frontal gyrus
MHC Major histocompatibility complex
MIns Mid-insula
MMSE Mini-mental status examination
MNI Montreal neurological institution
MoCA Montreal cognitive assessment
mPEC Medial prefrontal cortex
MRI Magnetic resonance imaging
MS Multiple sclerosis
NA Nucleus accumbens
NEX Number of excitations
NMDA N-methyl-D-aspartate
NMS Non-motor symptoms
xxiii
NO Nitric oxide
NSAID Nonsteroidal anti-inflammatory drugs
PBR Peripheral benzodiazepine receptors
PCR Polymerase chain reaction
PD Parkinson’s disease
PDD Parkinson’s disease with dementia
PD-HAB Parkinson’s disease with high affinity binder
PD-MAB Parkinson’s disease with mixed affinity binder
PD-MCI Parkinson’s disease Mild Cognitive Impairment
PET Positron emission tomography
PFC Prefrontal cortex
PiB Pittsburgh compound B
PLIC Posterior limb of internal capsule
PostIns Posterior insula
PPN Pedunculopontine nucleus
PVE Partial volume error
PVEC Partial volume error correction
RBD Rapid-eye-movement sleep behavior disorder
RF Radiofrequency
ROI Region of Interest
ROS Reactive oxygen species
rsfMRI Resting state functional magnetic resonance imaging
RSN Resting state network
SAL Salience network
SCR Superior corona radiata
SCZ schizophrenia
SFG Superior frontal gyrus
SLF Superior longitudinal fasciculus
SMA Supplementary motor area
SMN Sensorimotor network
SN Substantial nigra
SNP Single-nucleotide polymorphism
xxiv
SNpc Substantia nigra pars compacta
SNpr Substantia nigra pars reticulata
SPM Spatial parametric mapping
SPSS Statistical package for the social sciences
SRTM Simplified reference tissue model
STN Subthalamic nucleus
T1 Relaxation time 1
T2 Relaxation time 2
TAC Time-activity curve
TBSS Tract-based spatial statistics
TE Echo Time
TFCE Threshold free cluster enhancement
TI Inversion time
TLR2 Toll-like receptor2
TNF-α Tumor necrosis factor- α
TR Repetition Time
TSPO Translocator protein 18 kDa
UF Uncinate fasciculus
UPDRS Unified Parkinson’s disease rating scale
VBM Voxel based morphometry
Vim Vimentin
VND Non-displaceable distribution volume
VS Specific distribution volume
VT Total distribution volume
WM White matter
WMS Wechsler memory scale
1
1.0 Literature Review
1.1 Parkinson’s disease
Parkinson’s disease (PD) was originally described as a shaking palsy by James Parkinson in
1817 and later named after him. PD is the second most common age-related
neurodegenerative disease after Alzheimer’s disease (AD) and the most common movement
disorder (Goedert, Spillantini, Del Tredici, & Braak, 2013). It affects about 1% of population
aged over 60 years and 4% of those over 80 in industrial countries. The onset of disease
primarily occurs during the sixth decade, but there is also a young onset between 20 and 50
years of age (Dexter & Jenner, 2013).
PD diagnosis is made based on the unilateral onset of the cardinal motor symptoms such as
tremor, rigidity, and bradykinesia, which later spread to become bilateral, levodopa
responsiveness, and absence of markers suggestive of other disease. Pathologic confirmation
includes alpha-synuclein (α-synuclein) deposition and dopamine (DA) neuronal loss in the
substantia nigra pars compacta (SNpc) (Berg et al., 2014).
Etiology of PD remains unknown. Familial PD accounts for approximately 10% (Dexter &
Jenner, 2013). Gene mutations causing familial PD include autosomal dominant and
autosomal recessive gene mutations in SNCA, LRRK2, and PAKIN (Houlden & Singleton,
2012). The rest comprises of idiopathic PD where both genetic risk and environmental
factors play roles in its etiology.
PD is classically characterized by the cardinal motor symptoms such as tremor, rigidity,
bradykinesia, and postural instability (Dexter & Jenner, 2013). However, it is recognized to
be a more complex disease producing a number of non-motor symptoms (NMS) such as
rapid-eye-movement sleep behavior disorder (RBD), hypomia, constipation, autonomic
dysfunction, visual disturbances, depression, and cognitive impairment. Some of these NMS
often occur well before PD diagnosis (Berg et al., 2014; Langston, 2006; Muzerengi,
Contrafatto, & Chaudhuri, 2007) and are generally associated with increased disability, poor
quality of life and entry into long-term care (Chaudhuri & Schapira, 2009).
2
The treatment for motor symptoms of PD include DA replacement therapy employing
Levodopa and DA agonists, supported by the use of enzyme inhibitors such as peripheral
dopa decarboxylase inhibitors, catechol-O-methyl transferase inhibitors, and selective
monoamine oxidase type B inhibitors (selegiline and rasagiline) (Dexter & Jenner, 2013;
Tufekci, Meuwissen, Genc, & Genc, 2012). Besides these dopaminergic medications, the
weak N-methyl-D-aspartate receptor antagonist, amantadine and anticholinergics are also in
use. Alternative treatment strategies include deep brain stimulation in more advanced stages
of the disease. Unlike the motor symptoms of PD, the NMS such as cognitive impairment is
not effectively treated because of the limited knowledge of its pathology.
1.2 Clinical symptoms of Parkinson’s disease
In addition to the cardinal motor symptoms such as tremor, rigidity, bradykinesia, and
postural instability (Dexter & Jenner, 2013), other motor features include gait and posture
changes that manifest as festination (rapid shuffling steps with a forward-flexed posture
when walking), speech and swallowing difficulties, and a masklike facial expression and
micrographia (Jankovic, 2008). Among NMS, cognitive impairment frequently occurs in PD.
Twenty to 60% of patients may show impairment in one or more cognitive domains
(Aarsland, Bronnick, & Fladby, 2011; Williams-Gray, Foltynie, Brayne, Robbins, & Barker,
2007).
The most common cognitive abnormality in PD is executive dysfunction (Litvan et al., 2012)
such as problem solving, planning, set-shifting (switching goals or tasks) and inhibition.
Additionally, impairment in attention and working memory, visuospatial function, memory
and language are also reported in PD. PD patients showing these cognitive deficits, but no
significant impairment in daily function are termed Mild Cognitive Impairment in PD (PD-
MCI) (Caviness et al., 2007). The PD-MCI is classified in two ways based on a standard
cognitive assessment scale and/or the neuropsychological testing (Litvan et al., 2012). It is
now evident that PD patients with MCI have a six-fold increase in the risk of developing
dementia (Aarsland et al., 2011; Janvin, Larsen, Aarsland, & Hugdahl, 2006).
3
1.3 Neuropathology of Parkinson’s disease
Neuropathology of PD is primarily characterized by loss of dopaminergic neurons in the
SNpc, resulting in depletion of dopaminergic input to the striatum as well as widespread
aggregation of α-synuclein, the main component of the Lewy pathologies. However, a
number of other neuropathological processes have been indicated such as dysfunction of
non-dopaminergic systems, mitochondrial dysfunction, impaired autophagy, glutamate (Glu)
excitotoxicity, and neuroinflammation (Dexter & Jenner, 2013). The following sub-sections
will discuss some of the neuropathology of PD including dopaminergic and non-
dopaminergic dysfunction and α-synuclein aggregation, and neuroinflammation that has been
investigated in the study 3 of this thesis.
1.3.1 Dysfunction in dopaminergic pathways and basal ganglia-
thalamocortical circuits
There are four DA pathways of the brain: nigrostriatal, mesocortical, mesolimbic, and
tuberoinfundibular pathways. The first three pathways are most relevant to PD. The
nigrostriatal pathway originates from the SNpc of the midbrain and innervates the striatal
structures including the caudate nucleus (CN) and putamen, which is involved in motor,
cognitive, and behavioural/emotional functions. At a time of the diagnosis of PD, the
majority of dopaminergic neurons in the SNpc have been lost. This causes severe depletion
of dopaminergic input to the striatum, leading to pathophysiological changes within the basal
ganglia (BG) thalamocortical circuits including motor, limbic, and cognitive circuits (Figure
1-1; Figure 1-2). The mesocortical and mesolimbic pathways originate in the ventral
tegmental area of the brainstem. The former pathway innervates most regions of the cortex
with most dense projections to the entorhinal, anterior cingulate, insular and prefrontal
cortices and is associated with learning, memory and cognition (Oades & Halliday, 1987).
The latter pathway innervates mostly the nucleus accumbens (NA) or the ventral striatum
with some projections to the hippocampus, amygdala and thalamus (Oades & Halliday,
1987), and is involved in reward and emotional processing (Oades & Halliday, 1987).
Dysfunctions of these DA pathways and the BG-thalamocortical circuits contribute to motor
4
and cognitive/behavioural complications of PD such as cardinal motor symptoms (Stoessl,
Lehericy, & Strafella, 2014) as well as impairment in executive function (Floresco &
Magyar, 2006), reward-learning behaviour (Lee & Jeon, 2014), and other neuropsychiatric
complications (i.e., depression, visual hallucinations, and so forth) (Pagonabarraga,
Kulisevsky, Strafella, & Krack, 2015) (Figure 1-1; Figure 1-2).
Figure 1-1. BG thalamocortical model in normal state and parkinsonian state. In the
parkinsonian state, DA depletion in SNpc decreases input to the striatum, subthalamic
nucleus (STN), and globus pallidus pars interna-substantia nigra pars reticulata (GPi-SNpr)
illustrated in dotted lines, and the nuclei in the BG such as STN, GPi-SNpr, and
pedunculopontine nucleus (PPN) in darker color boxes become hyperactive (left).
Reproduced with permission from (Obeso & Lanciego, 2011).
5
Figure 1-2. Three main cortico-BG thalamocortical circuits: motor, associative, and limbic
circuits. The diagram demonstrates the anatomical and functional organization of circuits.
DA depletion in the posterior putamen affects the motor circuit, producing motor
dysfunction, and the dysfunction in the more anterior and ventral regions subserving the
associative and limbic circuits can affect cognitive and limbic functions. Reproduced with
permission from (Galvan, Devergnas, & Wichmann, 2015).
1.3.2 Dysfunction of non-dopaminergic systems
In addition to the dysfunction of the dopaminergic system caused by the degeneration of
dopaminergic neurons in the SNpc, non-dopaminergic nuclei may also be affected such as
the locus coeruleus (producing nor-epinephirine), raphe nucleus (producing serotonin), basal
nucleus of the Meyert (producing acetylcholine), (Jellinger, 2012) (Figure 1-3). Thus, other
neurotransmitter systems are also affected in PD where the cell bodies can be depleted
anywhere between 30 and 90% (Jellinger, 1991). These neuropathological changes likely
contribute to producing a number of NMS (Dexter & Jenner, 2013).
6
Figure 1-3. Affected various neurotransmitter systems in PD. Neurodegeneration in the
brainstem nuclei affecting widespread cortical regions as well as the cerebellum in PD.
Reproduced with permission from (Lang & Lozano, 1998), Copyright Massachusetts
Medical Society.
7
1.3.3 Lewy pathologies and alpha-synuclein
Cellular inclusions in nerve cell bodies and in nerve cell processes were first described in
brain regions outside of the substantia nigra (SN) of post-mortem PD brains by Fritz
Heinrich Lewy in 1912 and later called Lewy body and Lewy neurites, respectively after his
name (Goedert et al., 2013). There are two types of Lewy bodies (i.e. brain stem type and
cortical type) and lewy neurites occurring in the axons and dendrites of affected neurons.
The significance of the inclusions was poorly understood until 1997 when a missense
mutation in alpha-synuclein (α-synuclein) gene was found to cause familial PD
(Polymeropoulos et al., 1997). The α-synuclein is the main component of the Lewy
pathologies in PD (Spillantini et al., 1997). The α-synuclein is composed of 140 amino-acid
proteins (Benskey, Perez, & Manfredsson, 2016) and widely expressed throughout central
and peripheral nervous system. It is primarily considered a presynaptic protein (Iwai et al.,
1995), and thought to play a role in synaptic transmission and vesicular transport (Uversky,
2007) although the exact function remains unclear. The aggregation of α-synuclein from
monomers into fibrils via oligomeric intermediates occurs in the soma and neurites of
neurons in PD (Spillantini et al., 1997; Spillantini, Crowther, Jakes, Hasegawa, & Goedert,
1998) and is thought to be part of the toxic mechanism causing the disease (Glass, Saijo,
Winner, Marchetto, & Gage, 2010).
A staging scheme on the temporal and regional distribution of α-synuclein inclusions
associated with idiopathic PD was proposed by Braak, Tredici, Rub, et al (2003) (Figure 1-
4). It includes six stages (Braak et al., 2003; Braak & Del Tredici, 2009): In stage 1, the
pathology is present in the olfactory bulb, the anterior olfactory nucleus and/or dorsal motor
nucleus of the glossopharyngeal and vagal nerves in the brainstem. In stage 2, it develops in
the medulla oblongata and the pontine tegmentum (locus coeruleus, magnocellular nucleus of
the reticular formation, and lower raphe nuclei). In stage 3, it reaches the pedunculopontine
nucleus, the cholinergic magnocellular nuclei of the basal forebrain, amygdala and the SNpc.
Generally, during this stage the cardinal motor symptoms of PD begin to appear. In stage 4,
it progresses to the hypothalamus, parts of the thalamus, and the anteromedical temporal
mesocortex. During stage 5 and 6, it appears in the cortical regions, from high-order
8
association areas including the insular and cingulate cortices (stage 5) to secondary and
primary regions.
The α-synuclein deposits are also detected in the spinal cord, the enteric nervous system of
the peripheral nervous system connecting to the brain via the vagal nerve, plexuses in the
gastro-intestinal tract, sympathetic ganglia and trunk, the adrenal medulla, the sub-madibular
gland, and the heart. These α-synuclein deposits outside of the brain and those in the brain
from the stage 1-3 may provide pathological substrates for NMS (e.g., autonomic
dysfunction, hyposmia, depression and RBD) during the premotor stages of the disease
(Benskey et al., 2016; Langston, 2006; Schapira & Tolosa, 2010).
The transmission of α-synuclein between cells was evidenced in post-mortem PD brain
(Kordower, Chu, Hauser, Olanow, & Freeman, 2008; Li et al., 2008) and animals (Desplats
et al., 2009; Hansen et al., 2011). PD patients who received mesencephalic tissue transplants
showed that two to five percent of the grafted neurons displayed α-synuclein aggregates over
five years (Brundin, Li, Holton, Lindvall, & Revesz, 2008; Kordower et al., 2008; Kordower,
Chu, Hauser, Freeman, & Olanow, 2008; Li et al., 2008; Li et al., 2010). In transgenic mice
overexpressing human α-synuclein, the transgenic protein was transferred from the host
neurons to neural stem cells grafted into the hippocampus (Desplats et al., 2009), and to
dopaminergic neurons grafted into the striatum (Hansen et al., 2011).
The mechanisms underlying this cell-to-cell transmission of α-synuclein aggregates can be
exocytosis (Brundin, Melki, & Kopito, 2010; Lee, Bae, & Lee, 2014) and passive release
from cell death (Brundin et al., 2010), and endocytosis of neighboring neurons (Lee et al.,
2014). The spreading of α-synuclein aggregates across brain regions described in Braak’s
staging scheme may occur through trans-synaptic transmission (Brundin et al., 2010;
Hawkes, Del Tredici, & Braak, 2009; Volpicelli-Daley et al., 2011). However, this has not
been confirmed in in vivo models (Lee et al., 2014) (Figure 1-5).
9
Figure 1-4. Six stages of spread of α-synuclein inclusions across brain regions in idiopathic
PD. Disease progresses with additional brain regions from lower to higher brain regions and
worsening of pathology compared to the previously affected brain regions illustrated from
lighter to darker colors. Reproduced with permission from (Goedert et al., 2013).
Figure 1-5. Neuronal transmission of α-synuclein aggregates. a. Intracellular α-synuclein
aggregates can be released from neurons by exocytosis or cell death. Then, the aggregates are
taken up by neighboring neuronal cell bodies and are either retained in the cell soma (local
spread of pathology) or transported anterogradely by axons. Alternatively, they are taken up
10
by axon terminals and transported retrogradely to the cell soma. The protein aggregates can
spread between brain regions by axonal transport. b. Appearance of α-synuclein aggregates
from the initial stage in the dorsal motor nucleus of the vagal nerve in the brainstem and
anterior olfactory structures in darkest green to the final stage in the primary and secondary
cortical regions. Reproduced with permission from (Brundin et al., 2010).
1.3.4 Neuroinflammation
Inflammation is generated by an activated immune system, and is a strictly regulated self-
defensive mechanism against pathogenic stimuli or injury to protect the host organism by
clearing pathogenic stimuli or debris and promote the healing process (Khandelwal, Herman,
& Moussa, 2011; Stone, Reynolds, Mosley, & Gendelman, 2009). In the central nervous
system (CNS), glial cells including microglia and astrocytes are primary immune cells
protecting the brain against pathogenic stimuli (Glass et al., 2010).
1.3.4.1 Microglia
Microglia are the key immune cells of the CNS originated from primitive macrophage
progenitors in the yolk sac and migrate into the CNS during early embryogenesis (Saijo &
Glass, 2011). They constitute up to 5-15 % of the brain and 20 % of the glial cell population
(Kofler & Wiley, 2011; J. A. Smith, Das, Ray, & Banik, 2012). In humans, microglia are
present at a higher density in white matter (WM) than gray matter (GM) and localized mostly
in the medulla oblongata, pons, BG, and SN (Mittelbronn, Dietz, Schluesener, &
Meyermann, 2001). Under the normal physiological conditions, microglia have a unique
ramified morphology consisting of a small and round soma with numerous branching
processes (Smith et al., 2012). These “resting” microglia are constantly sampling their
surrounding environment by their processes undergoing repeated cycles of extension and
withdrawal while the cell bodies remaining relatively stable (Venneti, Wiley, & Kofler,
2009). Microglia mediate immune responses (Saijo & Glass, 2011), release neurotropic
factors, remove toxic substances, repair neurons, remodel and prune synapses, and guide
neural stem cells during synaptogenesis (Paolicelli et al., 2011; Ransohoff & Stevens, 2011;
Saijo & Glass, 2011).
11
Microglia become activated in response to various environmental challenges such as
bacterial and viral molecules, disease proteins (beta-amyloid and α-synuclein) and soluble
mediators released by damaged neurons such as α-synuclein and neuromelanin (Lull &
Block, 2010). The resting ramified microglia shift to the intermediate hyper-ramified
morphology, characterized by a larger soma and an amoeboid morphology to initiate
phagocytosis (Graeber & Streit, 2010). Microglia upregulate cell surface markers of
inflammation, proliferate, migrate to the injured site and secrete pro- and anti-inflammatory
cytokines such as tumor necrosis factor-α (TNF-α), interleukin-1B (IL-1B) and IL-10,
chemokines, oxidative and nitrodative stress-inducing factors such as nitric oxide (NO) and
superoxide (Lull & Block, 2010).
Activated microglia are often divided into two subsets: microglia M1 (M1) and microglia M2
(M2) based on the distinct molecular phenotype and effector function. The M1 phenotype
reflects classical activation where microglia produce oxidative metabolites, proteases, and
proinflammatory cytokines to defend the host organism against pathogens and tumor cells,
whereas the M2 phenotype reflects alternative activation where microglia secrete anti-
inflammatory cytokine IL-10 and transforming growth factor-β to repair and remodel tissue
(Czeh, Gressens, & Kaindl, 2011). However, this classification is overly simplistic in
diseased brains.
1.3.4. 2 Astrocytes
Astrocytes are abundant glial cells in the CNS and are five times greater in number than
neurons (Freeman, 2010). Astrocytes are dispersed throughout the CNS where each astrocyte
has own territorial region without overlapping with others. They exhibit numerous branches
extending through neurons and blood vessels to form functional networks via gap junctions,
called neurovascular units (Aarsland et al., 2011). Two main types of astrocytes in the CNS
include (1) protoplasmic astrocytes localized in the GM and envelope neuronal bodies and
synapses, and (2) fibrous astrocytes in the WM that interact with the nodes of Ranvier and
oligodendroglia (Halliday & Stevens, 2011; Oades & Halliday, 1987).
Astrocytes serve many essential functions: they perform regulatory and supportive roles in
the CNS such as biochemical and nutritional support for neurons, extracellular ion balance,
12
and repair of scarring of brain and spinal cord tissue (Sofroniew & Vinters, 2010), and form
a bridge between the vascular system and neurons, which makes them a key regulatory
element of neuronal activity and cerebral blood flow. Additionally, astrocytes have immune
functions that limit the spreading of inflammatory cells and infectious agents from the
damaged area to healthy parenchyma (Hamby & Sofroniew, 2010).
When astrocytes become activated, they undergo reactive astrogliosis including scar
formation. During the astrogliosis, astrocytes undergo morphological changes, resulting in
cell hypertrophy (Hamby & Sofroniew, 2010; Sofroniew & Vinters, 2010); upregulate
proteins such as the intermediate filaments vimentin (Vim) and glial fibrillary acid protein
(GFAP) (Alvarez, Katayama, & Prat, 2013); act as immune regulatory cells by presenting
antigens and pro- and anti-inflammatory molecules such as cytokines, chemokines, and
neurotropic factors (Sofroniew & Vinters, 2010); and activate distant microglia (Liu, Tang,
& Feng, 2011).
1.3.4.3 Neuroinflammation in aging condition
Age-associated changes in microglial functions have been characterized in two different
ways: (1) primed, sensitized or reactive and (2) senescent or dystrophic (Norden & Godbout,
2013; Streit & Xue, 2010; Venneti et al., 2009). The primed microglia are an intermediate
activation state, evidenced by morphological activation (deramification) and upregulation of
the major histocompatibility complex (MHC) class II cell surface markers (von Bernhardi,
Tichauer, & Eugenin, 2010), as well as elevated levels of proinflammatory cytokines and
downregulated anti-inflammatory cytokines (Dilger & Johnson, 2008). This creates a chronic
low-level inflammation and increased reactivity to stimuli in aged brain (Streit & Xue,
2010). On the other hand, the senescent microglia was referred to as the functional
impairments of microglia in phagocytosis (Norden & Godbout, 2013), characterized by
reduced phagocytosis of protein by microglia (Hickman, Allison, & El Khoury, 2008; Lee et
al., 2010), delayed recruitment of phagocytic cells and less clearance of debris (Zhao, Li, &
Franklin, 2006), and slower velocity towards the injury site and aggregation at the injury site
for a longer duration (Damani et al., 2011).
13
Astrocytes also become more inflammatory with age (Norden & Godbout, 2013). For
example, activated astrocytic profiles associated with inflammation (Eng, Ghirnikar, & Lee,
2000) such as increased expression of GFAP and Vim were found in brains of aged humans
(Finch, 2003; Nichols, Day, Laping, Johnson, & Finch, 1993; Porchet et al., 2003). These
activated astrocytes in aged brain may disrupt the communication with neurons and
microglia (Norden & Godbout, 2013).
The increased inflammatory markers on the glial cells in the aged brain set the stage for the
rapid and prolonged expression of pro-inflammatory molecules following a stimulus, which
can cause disease behaviour including cognitive impairment and depression (Norden &
Godbout, 2013; Streit, 2006). In addition, dysfunctional phagocytosis can lead to
accumulation of toxic materials and disease related proteins associated with
neurodegenerative disorders (Streit, 2006), which can induce more glial activation, creating
prolong inflammation that drives the chronic progression of neurodegenerative disorders
(Gao & Hong, 2008).
1.3.4.4 Neuroinflammation in Parkinson’s disease
In post-mortem PD brains, neuroinflammatory processes were confirmed by the presence of
microglia positive to MHC class II (Croisier, Moran, Dexter, Pearce, & Graeber, 2005;
Imamura et al., 2003; McGeer, Itagaki, Boyes, & McGeer, 1988), such as CR3/43 (Banati,
Daniel, & Blunt, 1998; Mirza, Hadberg, Thomsen, & Moos, 2000), ionized calcium binding
adaptor molecule 1 (Iba1) (Doorn, Moors et al., 2014b), as well as CD68, glycoprotein that
binds to low density lipoprotein and expressed on monocytes/macrophages (Doorn et al.,
2014b), and/or by the presence of astrocytes positive to GFAP (Banati et al., 1998; Mirza et
al., 2000; Mythri et al., 2011).
Activated microglial cells were first evidenced in the SN of PD patients (McGeer et al.,
1988). Subsequent studies corroborated their findings (Banati et al., 1998; Croisier et al.,
2005; Desai Bradaric, Patel, Schneider, Carvey, & Hendey, 2012; Doorn et al., 2014b;
Imamura et al., 2003; Mirza et al., 2000). Additionally, microglial activation occurred in
other brain regions such as putamen (Imamura et al., 2003), hippocampus (Doorn et al.,
2014b; Imamura et al., 2003), transentorhinal, cingulate, and temporal cortex (Imamura et
14
al., 2003), frontal cortex (Vroon et al., 2007), olfactory bulb (Vroon et al., 2007), and
anterior olfactory nucleus (Doorn, Goudriaan et al., 2014) . Reactive astrocytes were also
confirmed in the CN and frontal cortex (Mythri et al., 2011), and temporal and mesocortex
(Braak, Sastre, & Del Tredici, 2007). However, these neuroiflammatory processes were not
associated with clinical measures such as duration of disease (Banati et al., 1998; Croisier et
al., 2005; McGeer et al., 1988) or severity of disease (Mythri et al., 2011).
Inflammatory markers were also elevated including the levels of TNF-α (Mogi et al., 1994;
Nagatsu & Sawada, 2005) as well as IL-1 (Blum-Degen et al., 1995; Nagatsu & Sawada,
2005) and IL-6 in the cerebrospinal fluid (CSF) of PD patients (Blum-Degen et al., 1995;
Mogi et al., 1994; Muller, Blum-Degen, Przuntek, & Kuhn, 1998; Nagatsu & Sawada, 2005).
In addition, there were significant inverse correlations between IL-6 CSF levels and disease
severity measured by Unified Parkinson’s Disease Rating Scale (UPDRS) I, II, III,
bradykinesia, resting tremor, and rigidity (Muller et al., 1998).
1.3.4.5 Potential mechanisms underlying for neuroinflammation in
Parkinson’s disease
Neuroinflammatory processes can be activated by injured neurons releasing noxious self-
compounds such as membrane breakdown products, aggregated α-synuclein, and
neuromelanin in the extracellular milieu in PD (Gao & Hong, 2008). They were evidenced in
post-mortem brains where activated microglia was in close proximity to melanin-containing
dopaminergic cells and to free melanin in the SN (McGeer et al., 1988), as well as α-
synuclein depositions in the SN (Croisier et al., 2005; Doorn et al., 2014b), hippocampus
(Doorn et al., 2014b), cingulate cortex (Croisier et al., 2005), and in damaged dopaminergic
and serotonergic neuritis (Croisier et al., 2005). In addition, α- synuclein-positive astrocytes
were detected in prosencephalic regions such as amygdala, thalamus, septum, striatum,
claustrum and cerebral cortex (Braak et al., 2007).
Recent data also suggest that extracellular α-synuclein oligomers released by exocytosis from
neuronal cells play an important role in initiating neuroinflammatory processes. For example,
they can directly activate microglia by binding Toll-like receptor2 (TLR2) on microglia (Kim
15
et al., 2013). This response was not observed if microglia lacks TLR2. In addition,
microarray analysis of astrocytes exposed to neuron-released α synuclein induced specific
genes that are involved in proinflammatory responses, such as production of cytokines and
chemokines (Lee, Kim, & Lee, 2010). Proinflammatory reactions of astrocytes could also act
as a precursor for more extensive inflammation, possibly through recruitment and activation
of microglia (Lee et al., 2010). Both microglia and astrocytes were also activated to clear
extracellular α-synuclein aggregates although the former was shown to be principle
scavenger of extracellular α-synuclein aggregates (Lee, Suk, Bae, & Lee, 2008).
PD brain releases a greater amount of α-synuclein aggregates than normal brain (Lee et al.,
2014). Glial overactivation results in increased release of pro-inflammatory cytokines, NO,
and reactive oxygen species (ROS) (Deshpande et al., 2005; Lee et al., 2010; Qian & Flood,
2008; Zhang et al., 2005), which can create a chronic inflammatory microenvironment in PD
patients (Lee et al., 2014).
1.3.5 Summary of Parkinson’s disease
To summarize, PD is characterized by selective degeneration of dopaminergic neurons in the
SN, the presence of Lewy bodies and Lewy neurites containing α-synuclein aggregates in
such neurons, and neuroinflammation mediated by the activation of microglia and astrocytes.
These pathological changes collectively contribute to the cardinal motor symptoms of PD
such as resting tremor, bradykinesia, rigidity, and postural instability. In addition,
dysfunction of non-dopaminergic neurons such as serotonergic, noradrenergic, and
cholinergic neurons, the Lewy pathology, and neuroinflammation outside of the nigro-striatal
regions are thought to play a major role in NMS of PD such as psychiatric problems,
cognitive dysfunction including dementia, autonomic symptoms, and RBD. Therefore, PD is
a multisystem disease, affecting diffuse area of brain and producing various symptoms
(Figure 1-6). The complex neural substrates of PD reflecting these pathological changes can
be studied in vivo using multimodal neuroimaging tools in the attempt of identifying
neuroimaging biomarkers associated with motor and cognitive symptoms.
16
Figure 1-6. PD pathology. PD is characterized by degeneration of dopaminergic neurons and
non-dopaminergic neurons such as serotonergic, noradrenergic, and cholinergic neurons, as
well as aggregated α-synuclein, which can induce neuroinflammation mediated by the
activation of microglia and astrocytes. These pathological changes collectively contribute to
the cardinal motor symptoms of PD such as resting tremor, bradykinesia, rigidity, and
postural instability and NMS such as cognitive dysfunction, depression, impairment in
reward processing, hyposmia and RBD. Using in vivo neuroimaging tools, brain changes
such as structural as well as functional changes reflecting these pathological changes can be
quantified. Adapted from (Brudin, 2010).
1.3.6 Potential mechanisms underlying neuroimaging changes in
Parkinson’s disease
GM changes in PD reflect different neuropathologies such as dysfunction/degeneration of
different neurotransmitter systems, Lewy pathology and/or secondary effect of Lewy
pathology in subcortical regions, as well as neuroinflammation. These contribute to GM
changes varying from subtle changes such as reduction in size of cell bodies, dendritic
17
arborisation and presynaptic terminals (Morrison & Hof, 1997; Pellicano et al., 2012) to
synaptic and cell loss. Similarly, WM changes are possibly due to gliosis, Lewy pathologies,
and neuronal loss (Braak et al., 2004). For example, neuronal losses in subcortical projection
nuclei such as the noradrenergic locus ceruleus, the serotonergic raphe nuclei, and the
cholinergic nucleus basalis of Meynert (Jellinger, 1991) likely lead to axonal degeneration.
Lewy pathology can lead to axonal dysfunction and/or altered axonal microstructure
(Bellucci et al., 2016). Postmortem brains with dementia with Lewy bodies (DLB) showed
axonal transport dysfunction leading to accumulation of axonal transported substances
(Katsuse, Iseki, Marui, & Kosaka, 2003). This can change axonal microstructure by swelling
as well as retrograde and anterograde degeneration of the axonal projections (Hattori et al.,
2012), which may eventually leads cell body damage. These GM and WM changes can be
detected using magnetic resonance imaging (MRI).
These structural deteriorations mentioned above would affect functional connectivity (FC) of
the brain regions. In fact, direct comparisons between functional and structural connectivity
in the same individuals suggest that the FC can be predicted from the anatomical structure
(Honey et al., 2009), and the microstructure of cingulum that interconnects key regions of the
default mode network (DMN) was associated with the level of FC between these regions
(van den Heuvel, Mandl, Luigjes, & Hulshoff Pol, 2008). In addition, PD neuropathologies
can directly affect functionality. For example, the level of CSF α-synuclein was associated
with reduced sensorimotor network (SMN) connectivity in PD patients (Campbell et al.,
2015). Furthermore, the associations between neuroinflammation and FC were demonstrated
in patients with multiple sclerosis (Colasanti et al., 2016).
One of the potential mechanisms underlying structural and functional changes detected by
MRI is neuroinflammation, which can be specifically investigated utilizing Positron
Emission Tomography (PET) with a radioligand targeting for translocator protein (TSPO), a
biomarker for microglia activation.
18
1.4 Magnetic Resonance Imaging
1.4.1 Principles of Magnetic Resonance Imaging
MRI is primarily based on detection of the hydrogen (1H) signal from water because a 1H
nucleus or proton has a significant magnetic moment (meaning a small magnetic field like a
bar magnet generated from its positively charged and spin characteristics) and is abundant in
the body. When the protons are surrounded by a strong external magnetic field, they align in
either parallel or anti-parallel to the magnetic field. More nuclei take the parallel alignment,
as this requires lower energy than the other orientation. The protons also precess at some
frequency. This longitudinal magnetization then reaches thermal equilibrium. In this
condition, however, the individual protons creating the longitudinal magnetization do not
precess in phase. As the result, the longitudinal magnetization does not produce a measurable
signal.
To produce signal, a radiofrequency (RF) pulse in a 90° flip angle (that produces the
maximum signal) is introduced at the frequency at which the protons precess. As the RF
pulse continues, some of the protons in the low energy state absorb energy from the RF field
and make a transition into the high energy state, resulting in “tipping” the longitudinal
magnetization toward the transverse plane. The RF pulse results in the synchronized
precessing of proton spins in the transverse plane, producing an oscillating MR signal in a
receiver coil. After the RF transmitter is turned off, the protons start to lose precessing and
the transverse magnetization decays (T2 relaxation). Simultaneously, the longitudinal
magnetization increases toward its pre-excitation value over time (T1 relaxation). T2*
relaxation describes the decay in the transverse plane following a RF pulse due to
inhomogeneities in the magnetic field that are introduced by objects with magnetic
susceptibility.
The times required for the transverse and longitudinal magnetizations to return to their
equilibrium states are called relaxation times. T1 relaxation time (T1) indicates the time
required for the system to recover 63% of its equilibrium value (generally in the order of
19
seconds) and the T2 relaxation time (T2) is the time it takes for dephasing to decay to 37% of
its original value (generally tens to hundreds of milliseconds).
Image contrast can be modified by varying how often RF pulses are applied. The interval
between 90° pulses is called the repetition time (TR). The magnitude of the signal detected
also depends on how slowly the transverse magnetization decays from its initial maximum
value. The amount of time allowed for decay to occur following the initial 90° flip is called
the echo time (TE). The TR is shorter than the time necessary for total longitudinal relaxation
produces T1-weighted images. Long TR and long TE create T2-weighted images. When TR
is long so that protons fully recover longitudinal magnetization between repetitions, but TE is
short when the effect of T2 decay is minimized, then the image will not be dependent either
on T1 or T2, but only on the tissue’s proton density and it is called a proton density weighted
image.
1.4.2 Gray matter imaging
Given the sensitivity of T1-weighted imaging in detecting differences between tissue types,
structural analyses of brain GM are based on T1-weighted MR images. The two most
common quantitative analysis methods, voxel based morphometry (VBM) and cortical
thickness analysis (CTA) are discussed.
1.4.2.1 Voxel based morphometry
VBM allows for the whole brain voxel-wise comparison of GM volume between groups. It is
a fully automated and operator-independent analysis method. The data processing is
relatively straightforward and involves spatial normalization of individual T1-weighted
images into the common stereotactic space (e.g., the Montreal Neurological Institute (MNI)
152 standard space), followed by segmentation of GM, and smoothing for noise reduction.
Voxel-wise parametric statistical tests are performed on the smoothed images from the
groups (Ashburner & Friston, 2000). It can be carried out using the Spatial Parametric
Mapping (SPM) or FMRIB Software Library (FSL) software. Because of these simple,
automated pipelines to investigate the entire GM in the brain, it has been widely used to
study diseased brains including PD (Pan, Song, & Shang, 2012). However, there are some
20
limitations of VBM such as imperfect registration and tissue segmentation, which is
particularly a concern in abnormal brains. In addition, the statistical assumptions result in the
potential statistical errors (Duncan, Firbank, O'Brien, & Burn, 2013).
1.4.2.2 Cortical thickness analysis
Instead of measuring cortical GM in volume, an automated technique to accurately measure
cortical thickness (CTh) was developed using FreeSurfer software (Fischl & Dale, 2000),
i.e., CTA. CTA allows for the measure of CTh at any point in the submillimeter accuracy.
The detection of subtle cortical changes is important to study neuroplasticity due to
development and aging, neuropathology and treatment effects. CTA allows for the
quantification of GM in human cerebral cortex that is highly folded and has large regional
variations of the thickness by modeling the gray/white and pial surfaces. The identification of
these surfaces is needed for an accurate measure of CTh because the cortex contains many
folds that cannot be aligned with any of the cardinal axes along which slice data are viewed.
Figure 1-6 illustrates measuring CTh of one point from the coronal and axial slices and an
overestimation of the measurement using the coronal slice, resulting from the fact that the
surface is locally parallel to the coronal slice (Figure 1-7).
To measure CTh, the border between the WM and the GM (WM surface) is identified to
provide an estimate of the inner boundary of the cortex using a triangular tessellation over
this surface. This inner border is constructed first because resolution limitations innate to
MRI make it difficult to directly compute the outer border between the GM and the
meningeal tissue/CSF (pial surface). Once the inner cortical surface is identified, it is refined
for submillimeter accuracy and deformed outward to the pial surface. The corners of each
triangle on the surface become vertices. Thickness can then be computed as the
perpendicular distance measured between the inner and outer cortical boundaries at every
vertex in each hemisphere (Figure 1-8).
In addition to the measurement of CTh, FreeSurfer provides an automated technique to
measure subcortical volume, allowing for the investigation of whole brain GM changes.
21
Figure 1-7. Estimations of CTh using slice data. Coronal (left) and horizontal (right) views of
a T1-weighted MR image of the left hemisphere with gray/white (yellow) and pial (red)
surfaces overlaid. The green crosses indicate a point at which would result in a dramatic
overestimation of the thickness of the cortex using the coronal view only. Reproduced with
permission from (Fischl & Dale, 2000), Copyright (2000) National Academy of Sciences.
22
Figure 1-8. Measurement of CTh. The intersection of the tessellated WM surface in blue and
pial surface in red on MRI volume (left). CTh is the perpendicular distance measured
between the inner and outer cortical boundaries at every vertex (right). Reproduced with
permission from (Malpass, 2011).
1.4.3 White matter imaging
WM imaging for quantitative analyses is based on diffusion-weighted imaging (DWI).
Diffusion refers to the random movement of molecules due to their thermal energy (Le
Bihan, 2003; Mukherjee, Chung, Berman, Hess, & Henry, 2008; R. Watts, Liston, Niogi, &
Ulug, 2003), known as Brownian motion (Beaulieu, 2002; Le Bihan, Poupon, Amadon, &
Lethimonnier, 2006). In a free medium, during a given time interval (diffusion time),
molecules travel randomly in space over a distance depending on the size of the molecules,
the temperature and the viscosity of the medium, which is statistically well described by a
diffusion coefficient. Water diffusion in biological tissues, however, significantly differs
23
from this Brownian motion because of the presence of obstacles such as cell membranes,
fibers and macromolecules (Beaulieu, 2002; Le Bihan et al., 2006; Mukherjee et al., 2008;
Neil, 2008). This reflects apparent diffusion coefficient (ADC), and allows us to infer
microstructure and geometric organization of neural tissues and importantly changes in these
features with pathological states (Le Bihan, 2003).
1.4.3.1 Diffusion-weighted imaging/Diffusion tensor imaging
DWI also relies on a proton signal from water. However, its contrast is based on water
displacements in a given time. In DWI, a pair of strong magnetic field gradient pulses
(diffusion-sensitizing gradients) is applied. The first pulse labels the spatial location of 1H
nuclei in water molecules, and the second pulse detects changes in the location due to
diffusion that occurred in the time in between the two pulses. The amplitude, duration, and
the interval between the onsets of the diffusion gradients, described by a parameter known as
the ‘b’ factor influences the signal intensities obtained in images (Mukherjee et al., 2008).
Multiple images are collected so that the signal can be sensitized to diffusion in many
different directions, building up multiple measurements for each voxel in the brain
(Johansen-Berg, 2009).
Water mobility in brain WM is anisotropic, reflecting the non-equal diffusivities depending
on direction (Basser, 1995; Pierpaoli & Basser, 1996; Watts et al., 2003). In WM, the
presence of structures in intracellular and extracellular space (Sullivan & Pfefferbaum, 2006)
including axonal membranes, myelin sheath (Beaulieu, 2002) and glial cells creates an
anisotropic environment (Assaf & Pasternak, 2008). Water diffuses mostly along the
direction of the axons rather than perpendicular to the axons (Figure 1-9). Under such
microstructure, diffusion tensor model is required to fully characterize diffusion including
the magnitude of diffusion, the degree of anisotropy, and the direction of diffusion
anisotropy (Alexander, Lee, Lazar, & Field, 2007). Diffusion tensor MR imaging or DTI is a
mathematical model of the three dimensional (3D) pattern of anisotropic diffusion of WM
tracts (Mukherjee et al., 2008) and was first introduced by Basser, Mattiello, & LeBihan
(1994). DTI estimates an effective diffusion tensor (D) with 3D measurements characterized
by a 3 x 3 matrix of nine components from DW images (Basser, 1995).
24
Dxx Dyx Dzx
D = Dxy Dyy Dzy
Dxz Dyz Dzz
Each component denoted by Dij reflects the direction, i, j = {x, y, z}. Because of symmetry,
paired off-diagonal diffusion components represent the same (i.e., Dxy = Dyx, Dxz = Dzx,
and Dyz = Dzy). Thus, the six diffusivities need to be determined (Gulani & Sundgren,
2006) with diffusion gradients along at least six different non-collinear directions (Bammer,
2003; Basser et al., 1994).
In practice, the image acquisition with a large number of non-collinear directions (at least 30
directions, but 60-80 directions are recommended) is required for accuracy in the
reconstruction of fiber orientations and for reduction in the statistical rotational variance (i.e.,
the extent to which the variance in an estimate of a given parameter depends on the
orientation of the structure) as fibers are not all aligned along the same axis (Jones, Knosche,
& Turner, 2013).
From the measured tensor, principal diffusivities or eigenvalues, annotated 1, 2, and 3 in
order from largest to smallest magnitude of diffusivities (Neil, 2008; Nucifora, Verma, Lee,
& Melhem, 2007) (Figure 1-10) as well as eigenvectors of the system for the orientation
(Pierpaoli & Basser, 1996) of the eigenvalues can be calculated by a mathematical
transformation or diagonalization relating to the measured tensor to the principle
diffusivities. The eigenvector associated with the largest eigenvalue is oriented parallel to the
fiber direction while the two other eigenvectors are oriented perpendicular to it (Basser et al.,
1994; Neil, 2008).
1 0 0
Dprin = 0 2 0
0 0 3
25
Once the diffusion tensor has been calculated, a range of quantities can be derived (Watts et
al., 2003). The mean diffusivity (MD) and trace of D are directionally averaged diffusivity of
water within a voxel, described as the mean of the three eigenvalues and the sum of the three
eigenvalues, respectively.
MD = (1 + 2 + 3 )/3 = trace (D)/3.
Fractional anisotropy (FA) characterizes the amount of diffusion anisotropy, and express the
anisotropic degree to which the diffusion tensor deviates from the directionally averaged
diffusion (Mukherjee et al., 2008).
When the primary eigenvalue is much larger than the other two eigenvalues, FA will be high,
indicating preferred direction of diffusion along the axon. FA rages from zero to one where
zero represents complete isotropic diffusion, and one represents complete anisotropic
diffusion (Le Bihan et al., 2001) (Figure 1-11).
The degree of orientational coherence of the axons is an important determinant of diffusion
anisotropy (Pierpaoli & Basser, 1996). Axotomy and loss of axons due to degeneration can
greatly contribute to decreased anisotropic diffusion. Myelin partially contributes to diffusion
anisotropy demonstrated in studies on multiple sclerosis (MS) or on neuronal development
(Horsfield, Larsson, Jones, & Gass, 1998; Huppi et al., 1998; Mukherjee et al., 2001;
Werring, Clark, Barker, Thompson, & Miller, 1999).
26
Figure 1-9. Anisotropic diffusion in WM. Water diffuses mostly along the direction of the
axons rather than perpendicular to the axons due to the presence of microstructures such as
axonal membranes and myelin sheath. Reproduced with permission from (Mukherjee,
Berman, Chung, Hess, & Henry, 2008), AJNR. American Journal of Neuroradiology, 29,
636).
Figure 1-10. Spherical tensor shape with three orthogonal eigenvectors and their associated
eigenvalues (λ1, λ2, λ3). Reproduced with permission from (Johansen-Berg & Rushworth,
2009).
27
Figure 1-11. Spherical tensor shapes illustrating various anisotropic degrees. FA = 0.1
indicates highly isotropic (left) and FA = 0.8 indicates with highly anisotropic (right).
Reproduced with permission from (Johansen-Berg & Rushworth, 2009).
1.4.3.2 Diffusion tensor tractography
There are two tractography techniques to characterize WM connectivity: (1) streamline
tractography (Mori, Crain, Chacko, & van Zijl, 1999) and (2) probabilistic tractography
using FSL software (Behrens et al., 2003; Parker & Alexander, 2005). The former follows
principal diffusion directions along adjacent voxels in WM to delineate WM connectivity.
Therefore, the only one streamline is created from one seed location (Mukherjee et al., 2008).
It uses arbitrary criteria to terminate tractography such as a low FA threshold that reflects
non WM, and a high curvature threshold, which is anatomically implausible. This approach
is often used to visualize WM connectivity. However, there are a few disadvantages of this
approach. First, modeling a single diffusion tensor in each voxel is problematic for voxels
that contain crossing fibers, which is often the case in most WM regions. Streamliners
typically terminate when fibers approach GM or areas of crossing fibers (i.e., areas of low
FA).
To address these limitations, probabilistic tractography has been developed. This technique
accounts for uncertainty in the likelihood that a WM pathway should continue and for the
possibility of crossing fibers within voxels (Behrens et al., 2003; Parker & Alexander, 2005).
In each voxel, a probability distribution of possible fiber orientations is created. From a seed
28
location (e.g., one or many voxels in a region of interest (ROI)), many streamline samples
(typically thousands) are sent to trace pathways through the voxel probability distributions.
This results in a map of many streamlines, where the density at a given voxel putatively
represents the strength of connectivity between that voxel and the seed location. This
approach can fit multiple fiber populations in a voxel, allowing for modeling crossing fibers
in a voxel (Behrens, Berg, Jbabdi, Rushworth, & Woolrich, 2007; Parker & Alexander,
2005). It also allows for tractography to GM targets (Johansen-Berg, 2009). Lastly, it can
provide quantitative information about the likelihood of connections (percentage of
streamline samples from seed location reaching target voxel), which can be used to make
inferences about connectivity strength and to perform statistical analyses between groups.
However, the resulting quantified connectivity strength should be interpreted with caution,
given the lack of knowledge about the biological significance of tracts obtained with
probabilistic tractography (Jones et al., 2013).
1.4.3.3 Tract-based Spatial Statistics
Tract-based Spatial Statistics (TBSS) is a fully automated method that allows for the whole-
brain voxel-wise comparison of DTI measures between groups using FSL software (Smith et
al., 2006). It is developed to overcome the limitations of VBM analyses such as inaccurate
alignment as well as smoothing that involves the arbitrary choice of smoothing extent and
that can be another source of partial volume effect. For this purpose, a group mean WM
‘skeleton’ is generated in TBSS. It is one voxel in thickness and contains the voxel with
highest FA being the center of WM tracts common to the subjects included in a study (Figure
1-12). Once the individual FA values are mapped onto the skeleton, group-level statistics can
be performed on the skeleton-space FA data.
29
Figure 1-12. Mean FA skeleton projected on a skull-stripped DWI image.
1.4.4 Functional MRI
1.4.4.1 Blood oxygen level-dependent signal
The blood oxygen level-dependent functional MRI (BOLD fMRI) was first described by
Ogawa and Lee (1990) in rat brains where deoxygenated venous blood resulted in vessels
appearing darker due to the reduction in signal. It is based on T2* signal and capitalized on
the fact that hemoglobin (Hb) in blood vessels in brain has different magnetic properties
depending on whether it is oxygenated or deoxygenated. While oxygenated Hb is
diamagnetic (weak repulsion from a magnetic field) and has little magnetic susceptibility,
deoxygenated Hb is paramagnetic (attracted to a magnetic filed). Deoxygenated Hb results in
greater field inhomogeneities and a faster decay and thus, reduction of the T2* signal
compared to oxygenated Hb (Heeger & Ress, 2002). Brain activation is reflected by an
increase in blood flow supplying neuronal activity in the active regions, and a decrease in
deoxygenated Hb, resulting in an increase in signal and image intensity. The decrease in
deoxygenated Hb is thought to be due to an oversupply of oxygenated blood when neuronal
activity increases, which could occur for a number of reasons including inefficient oxygen
30
extraction from blood to surrounding tissue and/or non-oxidative metabolic processes
(Heeger & Ress, 2002).
Changes in blood flow following neuronal activity are thought to be related with synaptic
activity of Glu and recycling of astrocytes (Raichle & Mintun, 2006). At the
electrophysiological level, BOLD activity increases in anesthetized monkeys corresponded
best with changes in local field potential (LFP) signals, representing the compound activity
of neuronal populations compared with multiunit and single unit spiking (Logothetis, Pauls,
Augath, Trinath, & Oeltermann, 2001). As LFPs reflect synchronized neural activity, not
necessarily linked with firing rate, it has been proposed that the BOLD signal reflects
synaptic input to and local processing within a given area, rather than spiking output.
However, the underlying mechanisms for BOLD signal changes are still not fully understood
despite demonstrated metabolic and neuronal correlates of the BOLD signals.
1.4.4.2 Resting state functional MRI
Resting state functional MRI (rsfMRI) measures low frequency (0.01-0.1 Hz) BOLD signal
fluctuations in the resting brain. BOLD fluctuations lower than 0.01 Hz occur due to scanner-
related noise and greater than 0.1 Hz are contaminated with respiratory and cardiovascular
signals. Participants undergo a typically 5-10 min scan where they rest, and are usually
instructed to think about nothing in particular and to let their minds wander. rsfMRI is
analyzed by using the entire time course of a scan to investigate FC with an assumption that
FC does not change over the course of the scan. FC is a measure of temporal correlations of
physiological signal in anatomically distinct brain regions or voxels (Friston, Frith, Liddle, &
Frackowiak, 1993). Regions demonstrating stronger correlations are thought to have greater
FC. Functionally connected regions form patterns of synchronous activity, which make up a
number of uniquely identifiable intrinsic connectivity networks (see section 1.4.4.4). While
rsfMRI can provide a whole brain representation of FC and avoids the potential cofounder of
differential task performance in clinical populations (Greicius, 2008), it does not have
controlled behavior, making it difficult to associate the data with cognitive functions.
1.4.4.3 Seed-based functional connectivity analysis
31
Seed-based analysis was the first and most straightforward method used to measure
correlations between resting state time-courses of different brain regions (Biswal, Yetkin,
Haughton, & Hyde, 1995). A time series extracted from a seed, which is a predefined ROI, is
correlated with that of every other voxel, resulting in correlation maps representing FC of
that particular seed with each individual voxel. This analysis method can also be applied to
many regions or all regions of the brain where the FC between all pairs of the regions is
calculated, generating an FC matrix. The resulting map or matrix provides information about
which regions the seed is functionally connected with and to what extent. Seed-based FC
analysis is most suitable when a researcher has a hypothesis-driven question about a specific
region. One advantage of this method is that the results are relatively easy to interpret, as
measured FC corresponds directly to the strength of correlated time-courses between specific
regions (Buckner & Vincent, 2007).
1.4.4.4 Independent Components Analysis
Independent Components Analysis (ICA) is a data-driven approach and separates
components of spatiotemporally coherent patterns of activity. Unlike the seed-based FC
analysis, ICA does not require any predefined ROIs as it uses multivariate data in all voxels
to separate the brain into groups of spatial regions with signals that have common properties.
This allows for ICA to detect structured noise components within the data, which can be
removed. This is one of the advantages of ICA over the seed-based FC analysis. ICA can
also generate multiple spatially independent networks of coherent activity, or resting-state
networks (RSNs). RSNs closely correspond to the organization of functional network derived
from co-activation analysis (Spreng, Sepulcre, Turner, Stevens, & Schacter, 2013). Examples
of these networks include default mode network (DMN), sensorimotor network (SMN),
visual network, and fronto-parietal network (FPN), and are largely consistent across healthy
people, and persist even during sleep and under anesthesia.
1.4.4.5 Graph theoretical analysis
Graph theory is a relatively old branch of mathematics that started back in the 1700 (Stam,
2014). The idea of graph theory is to represent a complex set of relationships and entity with
32
a set of nodes and their connections. In the application of graph theory to
neuroscience/neuroimaging, a graph (i.e., brain network) is defined by a set of nodes (i.e.,
brain regions) and the edges (i.e., anatomical, functional or effective connections) between
them. Anatomical connections typically correspond to WM tracts connecting pairs of brain
regions. Functional connections correspond to the magnitudes of temporal correlations in
brain activity between pairs of regions. Effective connections represent direct or indirect
causal influences of one region on another (Friston, Harrison, & Penny, 2003). Graph
theoretical analysis can model and quantify topological properties of structural and
functional brain networks at the level of the entire graph, modules (subnetworks), or
individual nodes, which makes this approach one of the most powerful and flexible analytical
methods to represent a complex brain system (Power et al., 2011). Graph theoretical analysis
can be applied to MRI, Electroencephalogram (EEG) or Magnetoencephalography (MEG)
data (Rubinov & Sporns, 2010).
1.4.4.5.1 Processing and analysis steps
Here, the processing and analysis steps for undirected graphs used in study 2 are discussed.
First, nodes are defined (e.g., anatomically defined regions of MRI, functionally defined
regions of fMRI activation data). Second, an association matrix is generated by compiling
allpairwise associations between nodes (e.g., Pearson correlation, partial correlation). Third,
the association matrix is thresholded and binarized to produce a binary adjacency matrix or
binary undirected graph where links are either connected or unconnected over the certain
threshold values. The threshold values are often arbitrarily determined and thus network
properties are often explored over a range of plausible thresholds. Instead of binary graphs,
weighted graphs can be generated, which does not require the thresholding and contain more
information than the simpler binary graphs. Lastly, the network parameters of interest in the
graph are computed. Statistical testing of network parameters may be best conducted by non-
parametric permutation methods, as there is the lack of statistical theory concerning the
distribution of network measures (Rubinov & Sporns, 2010).
In our study, we used a binary adjacency matrix or undirected graph as it is simpler to
characterize and has a more easily defined null model for statistical comparison than a
33
weighted network. In addition, weak and non-significant links of a weighted network tend to
obscure the topology of strong and significant connections and may also be spurious
connections (Rubinov & Sporns, 2010).
1.4.4.5.2 Graph measures
The most fundamental measure is the degree of a node, which is the number of connections
that a node has with the rest of the network. Nodes with high degree are often called hubs.
Hubs are also determined by their centrality. The centrality of a node or betweeness
centrality (BC) is measured by the number of shortest paths (i.e., direct connections) passing
through the node to connect with all other nodes in the network. A node with a high level of
centrality is often a bridging node connecting disparate parts of the network (Rubinov &
Sporns, 2010), and plays a key role in overall communication efficiency or integration of a
network (van den Heuvel & Hulshoff Pol, 2010). Creating a ‘lesioned’ network where a node
is deleted can assess the importance of an individual node to network efficiency.
If the nearest neighboring nodes of one node are also directly connected, they form a cluster
(Bullmore & Sporns, 2009). The clustering coefficient or local efficiency can be quantified
by the ratio of the number of connections that exit between the nearest neighbours of a node
and the maximum number of possible connections between the neighbouring nodes of a node
(Watts & Strogatz, 1998). High clustering coefficient indicates high local efficiency of
information transfer and robustness of the cluster/local function.
When a subset of nodes is densely connected to one another, they form a module. The nodes
in one module are sparsely connected to nodes in other modules of the network. It has been
demonstrated modules in graph correspond to RSNs of ICA (Crossley et al., 2013). The
degree to which the network is subdivided into distinct groups is quantified by modularity
(Newman, 2006). Hubs can also be defined in terms of their role in modules (Guimera &
Amaral, 2005). Nodes primarily connected to other nodes in the same module are referred to
as provincial hubs. They can be measured using the within-module degree z-score. On the
other hand, nodes linking nodes in other modules measured by participation coefficient are
referred to as connector hubs. They can be calculated by the proportion of edges linking it to
nodes in other modules (Power & Petersen, 2013). Thus, while provincial hubs play an
34
important role in functional specialization in the module (local integration), connector hubs
are important for interactions or functional integration between modules (Meunier,
Lambiotte, & Bullmore, 2010). Both structural and functional neuroimaging demonstrated
that the connector hubs are typically located at the junctions between anatomically
segregated cortices and in regions of multimodal association cortex while the provincial hubs
are located within functionally specialized areas of cortex such as primary or unimodal
association areas (Meunier et al., 2010).
Brain function of a network can be assessed using several measures. A degree distribution
represents the degrees of all nodes in a network (Amaral, Scala, Barthelemy, & Stanley,
2000), which is a marker of network resilience. Another measure of network resilience is
assortativity, which is the correlation between the degrees of connected nodes. Positive
assortativity indicates that high-degree nodes tend to connect to each other. The mean
clustering coefficient for a network reflects on average the prevalence of clustered
connectivity around individual nodes (Rubinov & Sporns, 2010). Path length is the minimum
number of edges a node must traverse to reach another node. Characteristic path length of
the network is the mean shortest path length between all pairs of nodes in the network and is
related to global efficiency. Low/short mean path length or high global efficiency indicates a
highly integrated system. Connection density or cost is the actual number of edges in the
graph as a proportion of the total number of possible edges. Lastly, small-worldness is
characterized by high levels of both local clustering and short paths and can be quantified by
the ratio of the clustering coefficient to the path length. It presents an intermediate network
organization between the organization of random networks characterized by a short overall
path length and low level of local clustering, and that of regular networks characterized by a
high-level of clustering and a long path length (Watts & Strogatz, 1998). Complex networks
such as brain generally presents small-world properties with non-Gaussian degree
distribution with a long tail towards high degree to achieve an optimal balance between
regional segregation (specialization) and global integration (Bullmore & Sporns, 2012;
Tononi, Sporns, & Edelman, 1994). Major graph measures are illustrated in Figure 1-13.
35
Figure 1-13. Graph measures. a. Representation of a graph with the collection of nodes and
the collection of edges (connections), describing the interactions between the nodes; b.
Clustering-coefficient of node i is given by the ratio of the number of connections between
the direct neighbors of node i in darker green (i.e., 2) and the maximum number of possible
connections between the neighbors of node i (i.e., 3), providing information about the level
of local connectedness in the graph; c. Path length of node i is given by the number of
connections that have to be crossed to travel from node i to node j in the graph, providing
information about the level of global communication efficiency of a network; d. BC of a
node i indicates how many of the shortest paths between the nodes of the network pass
g
36
through node i. A high BC indicates a potential hub role of this node in the overall network;
e. Degree of node i is defined as its total number of connections; f. Modularity of a graph
describes the possible formation of communities in the network; g. Network topologies:
regular network (left) has a rather local character, characterized by a high clustering-
coefficient (C) and a high path length (L); random network (right) has a more global
character, with a low C and a much shorter path length L than the regular network; and
small-world network (middle) with both a high C and a low L, combining a high level of
local and global efficiency. Reproduced with permission from (van den Heuvel & Hulshoff
Pol, 2010).
1.5 Positron Emission Tomography
1.5.1 Principles of Positron Emission Tomography imaging
PET imaging is a nuclear medicine imaging technique to quantify the chemical/biological
processes of molecules in vivo by injecting radiolabelled molecules (i.e., radioligand)
(Venneti, Lopresti, & Wiley, 2013). The radioligands typically resemble endogenous
biological molecules with specific biological targets to which they bind. This allows for
mapping the distribution of the protein in the brain. The radioligand is synthesized by
labeling a precursor molecule with short-lived radionuclides such as carbon-11 (t1/2 = 20.4
min) and fluorine-18 (t1/2 = 109.8 min) (Figure 1-14). After the radioligand is injected
intravenously, it enters the bloodstream, crosses the blood-brain barrier (BBB) and binds to
target receptors or proteins in the brain.
When the radionuclide decays, positron is emitted. The positron quickly loses its kinetic
energy and reaches thermal equilibrium with its surroundings and then combines with a free
electron before they mutually annihilate. The annihilation produces two 511 kilo-electron
volts (keV) gamma ray or photons that travel along a line in opposite directions (Figure 1-
13). These 511 keV gamma emissions are detected by rings of radiation detectors of a PET
scanner (Figure 1-14). Detection of two photons within a coincidence window (~ 6 nsec)
results in a coincidence event that is plotted on a sinogram Upon the completion of scan, the
37
sonogram is corrected for photon attenuation and scatter based on a transmission scan
acquired prior to the data acquisition. The PET image is reconstructed using a mathematical
algorithm based on the sinogram plots, radioligand dead time and decay for data
quantification.
Figure 1-14. Principles of PET. The synthesis of PET radioligand by labeling a precursor
38
molecule with short-lived radionuclides (top). PET radionuclides decay by positron emission
to a stable nuclide. Once the positron reaches thermal equilibrium with its surroundings
(thermalization), it interacts with a free electron, briefly forming a bound state called
positronium before both particles mutually annihilate. The annihilation of the electron and
positron results in the emission of two 511 keV gamma ray photons, traveling in opposite
directions along a line (middle). PET scanner detects the coincident annihilation events with
the surrounding rings of radiation detectors, which is used for the data reconstruction
(bottom). Reproduced with permission from (Venneti et al., 2013).
1.5.2 PET radioligands
PET radioligand should binds with ≤5% of target proteins (Innis et al., 2007) that reflects the
entire population of target proteins without significantly perturbing the total number of
available target proteins. To image targets in the brain, a PET radioligand must meet a
number of criteria such as high affinity, selectivity for the target protein, low non-specific
binding, and suitable brain pharmacokinetics related with radiolabel half-life (i.e., observable
brain uptake and washout kinetics). They also must penetrate the BBB and reach their target
in vivo. Radioligands with high lipophilicity are considered undesirable because of slow
brain entry and high non-specific binding to brain fats and proteins, resulting in attenuation
of specific signal. Furthermore, they must lack of radiometabolites in the brain because PET
cannot discriminate between the chemical sources of detected radioactivity. The metabolism
should ideally occur outside of the brain and produce less lipophilic radiometabolites with
poor brain entry and little or no interaction with the target protein.
1.5.3 Quantification of PET radioligand binding in vivo
1.5.3.1. Kinetic modeling of time-activity curve
Kinetic modeling is a mathematical model to quantify specific radioligand binding to a target
protein from the time-activity curve of the radioligand. Kinetic modeling often uses
compartment models. The compartments can include those of the blood plasma (CP),
specifically bound to a target protein (CS), nonspecifically bound (CNS) or free in the tissue
(CF). A compartment model is based on an assumption that radioligands may move freely
39
between compartments until they reach equilibrium. The movement of radioligand molecules
between the compartments is described in terms of rate constants. The rate constants are
estimated to fit the measured time-activity data to a compartment model using the
metabolite-corrected radioligand concentration in the plasma as an input function. The input
function is derived from estimations of a concentration of radioligand in the whole blood and
plasma, as well as fractions of radiolabeled metabolites in the plasma from a sequential series
of arterial blood samples obtained during the PET scan. Iterative non-linear fitting is
typically used to estimate the rate constants that best fit the measured time activity data. The
identifiability of the estimate of the rate constants can also be assessed using the percent
coefficient of variation (%COV), which is the standard error of the estimated parameters. An
estimate with a low %COV indicates a better fit of the chosen model and thus more accurate
parameter estimates. The goodness of fit and the appropriateness of a particular compartment
model for a given dataset can be quantitatively determined using the Akaike Information
Criterion (AIC) (Akaike, 1974) and Bayesian Information Criterion (BIC) (Schwarz, 1978).
Lower AIC and higher BIC values indicates better fit, respectively.
1.5.3.2. Two-tissue compartment model and its outcome measures
The two-tissue compartment model (2TCM) consists of three compartments: CP, CS, and
nondiplaceable compartment (CND) including both CF and CNS (CND = CF + CNS). (This
model is typically called three-compartment model in clinical pharmacology, but two-tissue
compartment in radioligand imaging) (Innis, et al., 2007). It is based on an assumption that
the CNS and CF have fast kinetics and reach equilibrium very rapidly, and thus they are
considered one compartment. (Figure 1-15). Rate constants include K1 from CP to CND, k2, in
a reverse direction, as well as k3 from CND to CS and k4, in a reverse direction. K1 uses upper
case whereas others use lower case. This distinction is primary because K1 uses different
units (mL* cm-3 mL-1 *min-1 representing a volume of blood per volume of tissue per minute)
than the other rate constants (min-1).
Using the 2TCM, three binding outcome measures can be estimated based on the kinetic
parameters including the total distribution volume (VT), distribution volume of the specific
compartment (VS), and the binding potential non-displaceable (BPND). VT refers to the ratio
40
at equilibrium of total concentration of radioligand in the brain tissue (CT) to the
concentration of radioligand in CP (Innis et al., 2007). Therefore, VT = CT / CP where CT =
CND + CS. In 2TCM, VT value can be estimated by VT = K1/k2 (1+ k3/k4). VS refers to the
ratio at equilibrium of the specifically bound radioligand to that of total parent radioligand in
plasma and estimated by VS = K1/k2 k3/k4. BPND refers to the ratio at equilibrium of
specifically bound radioligand to that of nondisplaceable radioligand in tissue and estimated
by BPND= k3/k4.
Figure 1-15. 2TCM. The two compartments are located within the tissue including
nondisplaceble (free and nonspecific) and specific compartments. Input to the tissue derives
from the plasma compartment. The rate constants representing the exchange of radioligand
between compartments.
1.5.3.3. One-tissue compartment model
A one-tissue compartment model (1TCM) includes two compartments: one is CP and the
other has CF, CNS, and CS, which cannot be kinetically distinguished from one another and
therefore, it describes the bidirectional flux of radioligand between the plasma and tissue
compartments (Figure 1-13).
CP
CND
= CF + C
NS
CS
K1
k2
k3
k4
CP
CF + C
NS +
C
S
K1
k2
41
Figure 1-16. 1TCM. The tissue has just one compartment, that is the free and nonspecifically
bound ligand cannot be kinetically distinguished from the specifically bound.
1.5.3.4 A simplified reference tissue model
A simplified reference tissue model (SRTM) does not require the arterial input function but
instead relies on the presence of a reference region (Lammertsma & Hume, 1996) (Figure 1-
17). It is based on four assumptions. First, the reference region is devoid of specific binding.
Second, target tissue and reference tissue compartments can be described as one
compartment based on an assumption that both compartments reach equilibrium rapidly.
Third, the level of nondisplaceble binding is the same in both reference and target tissues and
fourth, the blood volume contribution to both reference and target tissue is negligible
(Salinas, Searle, & Gunn, 2015). This model allows for an estimation of BPND, which refers
to the ratio at equilibrium of a specifically bound radioligand to that of a nondisplaceable
radioligand in tissue.
CREFERENCE
CP
K1
k2
K1
’
k2
’
CTARGET
CF + C
NS +
C
S
42
Figure 1-17. SRTM. The free, non-specific and specifically bound radioligand reach
equilibrium rapidly and can be condensed into one compartment. CREFERENCE represents a
brain region that is devoid of target receptors.
1.6 Neuroimaging studies in Parkinson’s disease
PD is a complex disease with different clinical phenotypes, rates of progression, and
pathologies. Current trends toward use of data driven analytic methods can facilitate to
expand our knowledge about PD beyond the classic BG motor model (Weingarten,
Sundman, Hickey, & Chen, 2015) and DA dysfunction. Structural and functional changes
beyond the BG have been investigated in vivo using different neuroimaging techniques. My
thesis projects employed MRI and PET to identify PD-related brain changes accounting for
the clinical manifestations.
1.6.1 Structural changes in Parkinson’s disease
MRI is a non-invasive and safe approach with high spatial and temporal resolution. It is
widely available, leading to its widespread applications for structural investigations of PD
(Weingarten et al., 2015). Whole brain GM and WM changes can be quantified using VBM
and CTA combined with subcortical ROI analysis from an anatomical T1-weighted image
and TBSS from a DWI image, respectively.
Whole brain GM changes have been extensively investigated using VBM in PD patients.
One meta-analysis study including 17 studies involving 498 idiopathic PD patients and 375
healthy control (HC) subjects reported significant regional GM volume loss in the left
inferior frontal gyrus (Brodmann Area (BA)47) extending to the left superior temporal gyrus
(BA38) and the left insula (Ins) (BA13) (Pan et al., 2012). As suggested from these findings,
many of VBM studies found cortical changes but failed to find subcortical changes (Biundo
et al., 2011; Burton, McKeith, Burn, Williams, & O'Brien, 2004; Hong, Lee, Sohn, & Lee,
43
2012; Kostic et al., 2010; Messina et al., 2011; Song et al., 2011; Tessitore et al., 2012;
Weintraub et al., 2011).
It is also noted that six of the 17 studies in the meta-analysis did not find any significant
group differences. While volume measurements like VBM often result in overestimations of
GM changes, they may underestimate subtle cortical changes occurring in submillimeter
(Fischl & Dale, 2000) (see section 1.4.2.2). Four PD studies using both CTA and VBM
suggested that CTA be more sensitive to detect changes in cortical regions in PD (Gerrits,
van Loenhoud, van den Berg, Berendse, Foncke, Klein, et al., 2016; Ibarretxe-Bilbao et al.,
2012; Pereira et al., 2012; Tessitore, Santangelo, De Micco, Vitale, Giordano, Raimo, et al.,
2016). In one study, faster rate of cortical thinning but not that of volume reduction was
detected in PD patients compared to HC subjects in a longitudinal study (Ibarretxe-Bilbao et
al., 2012), and in anther study CTA detected cortical thinning in a number of cortical regions
in PD patients compared with HC subjects whereas VBM found volume decrease only in one
cortical region (Pereira et al., 2012). In addition, CTA but not VBM differentiated PD
patients from HC subjects (Gerrits, et al., 2016) as well as PD patients with Impulse Control
Disorders (ICD) from those without ICD and HC subjects (Tessitore, et al., 2016).
CTA revealed that PD patients showed cortical thinning in various regions compared to HC
subjects. Cortical thinning occurred in the frontal regions including the motor (Pereira et al.,
2012), premotor (Zarei et al., 2013) and supplementary motor areas (SMA) (Jubault et al.,
2011; Zarei et al., 2013) as well as the orbitofrontal (Biundo et al., 2013; Hanganu et al.,
2013; Lyoo, Ryu, & Lee, 2010), superior (Zarei et al., 2013), middle (Lyoo et al., 2010),
dorsolateral (Zarei et al., 2013) and inferior frontal regions (Pereira et al., 2012; Tinaz,
Courtney, & Stern, 2011); in the temporal regions including the superior (Lyoo et al., 2010;
Pereira et al., 2012), middle (Pellicano et al., 2012; Pereira et al., 2012), and inferior
temporal or angular regions (Pereira et al., 2012; Zarei et al., 2013), fusiform (Lyoo et al.,
2010; Pellicano et al., 2012; Tinaz et al., 2011; Zarei et al., 2013) and lingual gyri (Hanganu
et al., 2013; Lyoo et al., 2010; Tinaz et al., 2011), and temporal pole (Lyoo et al., 2010;
Pereira et al., 2012); in the parietal regions including the precuneus (Zarei et al., 2013),
postcentral (Pereira et al., 2012) and the inferior parietal (Lyoo et al., 2010; Pereira et al.,
2012) or supramarginal (Lyoo et al., 2010; Zarei et al., 2013) regions; and in the occipital
44
regions including the lateral occipital (Li et al., 2013; Pereira et al., 2012; Tinaz et al., 2011)
and cuneus regions (Tinaz et al., 2011).
Taken together, the most reported region of cortical thinning is the frontal region followed by
the temporal and parietal regions. A longitudinal study also supported these findings by
demonstrating that PD patients showed a faster rate of cortical thinning than HC subjects in
most cortical regions except in the occipital region and the frontal cortex was most affected
(Ibarretxe-Bilbao et al., 2012). The frontal regions included the superior, caudal middle
frontal and inferior gyri, as well as precentral and paracentral regions. The temporal regions
included the superior temporal and middle temporal gyri. The parietal region included the
supramarginal and inferior parietal regions.
Cortical thickening in small cortical regions was also reported in subtypes of PD patients.
For example, PD-MCI patients showed cortical thickening in a small region of the left
middle temporal gyrus compared with PD patients with normal cognition (Hanganu et al.,
2013). In addition, PD patients with Levodopa-induced dyskinesia also showed cortical
thickening in the inferior frontal region compared to HC subjects (Cerasa et al., 2013).
WM changes also occurred in PD patients (Hattori et al., 2012; Kim et al., 2013; Melzer et
al., 2013; Theilmann et al., 2013; Zhan et al., 2012; Zheng et al., 2014). They become severe
as the disease progresses (Agosta et al., 2013) and cognition declines (Hattori et al., 2012).
The early stage of patients and patients with normal cognition may spare WM changes
(Agosta et al., 2013; Agosta et al., 2014). However, one study identified changes in limited
frontal WM including the anterior corpus callosum (CC), superior corona radiata (SCR), and
cingulum in a larger sample of PD patients with normal cognition (Melzer et al., 2013). This
study employed the Movement Disorders Society (MDS) diagnostic criteria for PD-MCI
(Litvan et al., 2012) to group PD patients into those with normal cognition and those with
MCI. WM changes only in the frontal region were also found in PD-MCI such as the anterior
corona radiata (ACR) and SCR (Agosta et al., 2014). Other studies tend to show diffuse WM
changes in non-demented PD patients including the studies where PD group showed no
deficits on neuropsychological tests (Gallagher et al., 2013; Theilmann et al., 2013).
45
CTh, subcortical volume, and WM measures were also investigated for correlations with
clinical/cognitive measures. CTh was associated with the duration (Hanganu et al., 2013;
Jubault et al., 2011; Lyoo, Ryu, & Lee, 2011), and stage of disease (Pereira et al., 2012;
Zarei et al., 2013), motor severity (Lyoo et al., 2011; Zarei et al., 2013), Mini-mental status
examination (MMSE) scores (Zarei et al., 2013), neuropsychological test scores (Biundo et
al., 2013; Pellicano et al., 2012), and facial emotion recognition (Baggio et al., 2012). GM
atrophy in subcortical regions was associated with poor memory, DA non-responsive motor
signs and executive function (Camicioli et al., 2009), global cognitive measure (aggregated
cognitive z score) (Melzer et al., 2012), as well as olfactory dysfunction (Wattendorf et al.,
2009). WM changes were associated with MMSE scores (Agosta et al., 2013; Gallagher et
al., 2013; Hattori et al., 2012; Melzer et al., 2013; Theilmann et al., 2013), cognitive
performance in different cognitive domains (Gallagher et al., 2013; Melzer et al., 2013;
Theilmann et al., 2013; Zhan et al., 2012).
Despite the evidence of structural abnormalities and variability of GM and WM measures
associated with clinical and cognitive manifestations in PD patients, whether the reported
structural abnormalities account for specific clinical sequelae in PD patients is still unclear.
This is largely because studies that used the whole brain approach typically investigated
group differences and correlations independently (Camicioli et al., 2009; Hanganu et al.,
2013; Jubault et al., 2011; Tinaz et al., 2011). For example, one study found group
differences in WM measured by increased MD in diffuse brain regions (Agosta et al., 2013).
However, correlations were found only between FA and MMSE scores. These findings make
the interpretations difficult and questions structural makers for PD pathology. Even though
brain regions demonstrating significant group differences have been shown to correlate with
cognition, PD patients did not show impairment on cognitive tasks (Theilmann et al., 2013).
The authors argued that WM changes might precede cognitive impairment in these patients,
which warrants further confirmation. Furthermore, only correlations were investigated in PD
patients (Lyoo et al., 2011; Zheng et al., 2014). Thus, whether or not these correlations were
unique to PD patients and whether or not these regions showed any structural differences
compared to HC subjects are unknown. Studies are needed to elucidate whether the structural
changes would account for specific clinical manifestations of PD by interrogating the regions
showing significant structural changes for correlations.
46
For this purpose, we aimed to determine whether structural changes occurred in PD patients
and whether these structural abnormalities were associated with the motor and NMS by
investigating the whole brain GM and WM using CTA and TBSS (see section 2.1).
1.6.2 Brain network changes in Parkinson’s disease using resting
statefMRI and graph theoretical approach
To date, there are approximately 20 studies published using graph theoretical analyses to
fMRI, EEG, and MEG data in PD patients. In general, these studies included PD patients in
the earlier stage of disease.
On the global level, PD patients showed more segregated (Fogelson et al., 2013; Gottlich et
al., 2013; Ma et al., 2016; Sang et al., 2015; Skidmore et al., 2011) and less integrated
network (Fogelson et al., 2013; Gottlich et al., 2013; Ma et al., 2016; Sang et al., 2015;
Skidmore et al., 2011). These topological patterns were also observed in PD-MCI (Baggio et
al., 2014). On the contrary, one study showed that PD patients showed less segregated
network with preserved overall integration (Luo et al., 2015). A longitudinal study showed
that earlier pathology was limited to reduced local clustering/integration, which later became
more prominent and widespread, leading to a reduction in global integration (Olde
Dubbelink et al., 2014).
More segregated network measured by clustering coefficient (Lebedev et al., 2014; Tinaz,
Lauro, Hallett, & Horovitz, 2016) or less integrated network measured by BC (Olde
Dubbelink et al., 2014; Tinaz et al., 2016) were associated with more severe motor symptoms
as well as the overall disease severity measured by UPDRS total scores. In addition,
increased network segregation was associated with worse cognitive performance in
visuospatial/visuoperceptual and memory domains in PD-MCI patients (Baggio et al., 2014).
On the modular (subnetwork) level, PD patient did not differ from HC subjects (Ma et al.,
2016; Sang et al., 2015). However, PD-MCI patients showed increased modularity compared
with PD patients with normal cognition and HC subjects (Baggio et al., 2014), suggesting
more segregated brain system in PD-MCI. One study investigated intra- and inter-modular
connectivity and found increased local integration in medial PFC (mPFC) network, salience
47
network (SAL), SMN, and FPN, as well as reduced interactions between mPFC network and
(1) SAL (2) SMN and (3) visual network in PD patients compared with HC subjects (Ma et
al., 2016). Furthermore, higher local integration within the mPFC network was associated
with better cognitive performance on MMSE (Ma et al., 2016).
On the local level, as expected, PD patients showed nodal changes (both increased and
decreased FC) within SMN (Gottlich et al., 2013; Luo et al., 2015; Nagano-Saito, Martinu, &
Monchi, 2014; Tinaz et al., 2016; Wei et al., 2014; Wu, Wang et al., 2009) and the changes
correlated (both positive and negative) with UPDRS scores (Cao, Xu, Zhao, Long, & Zhang,
2011; Wei et al., 2014; Wu et al., 2009). In addition, nodal changes occurred in the fronto-
parietal (Fogelson et al., 2013; Ma et al., 2016; Tinaz et al., 2016), visual or occipital
(Gottlich et al., 2013; Luo et al., 2015), orbitofrontal (Gottlich et al., 2013; Luo et al., 2015;
Olde Dubbelink et al., 2014; Tinaz et al., 2016), and temporal regions (Luo et al., 2015; Olde
Dubbelink et al., 2014). Increased connectivity in the fronto-parietal nodes was associated
with better executive function (Lebedev et al., 2014).
Hub reorganization was also observed in PD patients (Baggio et al., 2014; Sang et al., 2015).
Notable alternations in hub organization included lost hubness in nodes of temporal-limbic
regions as well as in the putamen in PD patients (Sang et al., 2015) and increased hubness in
nodes of the prefrontal region in PD-MCI patients (Baggio et al., 2014). It is noted that the
definition of hubness is arbitrary and these studies used different measures for hubness: The
former study used BC and tested the values against the null hypothesis using a non-
parametric one-tail test across different thresholds and the latter study defined nodes as hubs
when they ranked at 20% highest sum of BC and degree out of 90 ROIs.
These findings suggest that in the earlier disease, functional changes occur locally and a loss
of centrality of hub nodes leads to disconnection of interacting modules. As the disease
progresses, the functional changes on inter- and intra-modular levels become more prominent
and widespread, affecting global integration of the network.
Using the graph theoretical approach, an investigation of hub regions has become the focus
of attention in pathological brains because of its widespread effects on network structure and
function (Zalesky, Fornito, Cocchi, Gollo, & Breakspear, 2014). Empirical studies also
48
suggest that hub regions were disproportionately affected by brain disorders (Buckner et al.,
2009; Crossley et al., 2014). This may be related to the high baseline activity and metabolic
activity of hub regions such as high rates of cerebral blood flow, aerobic glycolysis, and
oxidative glucose metabolism (Buckner et al., 2009; Liang, Zou, He, & Yang, 2013; Saxena
& Caroni, 2011; Tomasi, Wang, & Volkow, 2013; Vaishnavi et al., 2010), which make them
particularly vulnerable to a wide range of pathologic factors such as metabolic (oxidative)
stress or oxidative stress (Crossley et al., 2014). Network hubs are almost certainly
vulnerable in PD (Stam, 2014) with dysfunction of multiple neurotransmitters (Gratwicke,
Jahanshahi, & Foltynie, 2015; Pagonabarraga et al., 2015), deposition of α-synuclein (Braak,
Rub, Jansen Steur, Del Tredici, & de Vos, 2005), and neuroinflammation (Gao & Hong,
2008; Hirsch & Hunot, 2009), and the selective deterioration of hubs may account for the
distributed abnormalities across the brain in PD (McColgan et al., 2015) with distinct clinical
manifestations. Therefore, the Study 2 hypothesized selective node changes (see section 2.2)
and their association with motor and cognitive symptoms.
1.6.3 Neuroinflammation in PD using Translocator protein 18 kDa
imaging
1.6.3.1.Translocator protein 18 kDa
TSPO (18 kDa) has been investigated as a in vivo biomarker for neuroinflammation (Chen &
Guilarte, 2008; Venneti et al., 2013). TSPO was first discovered as an alternative binding site
for the benzodiazepines in 1977 (Braestrup & Squires, 1977). It was initially called
Peripheral Benzodiazepine Receptor (PBR) and renamed TSPO that reflects its molecular
role in cholesterol synthesis and the molecular weight of the protein (Papadopoulos et al.,
2006). TSPO consists of 169 amino acids and highly hydrophobic and rich in tryptophan
(Casellas, Galiegue, & Basile, 2002). There is an 80% sequence homology among various
species such as rodents, bovines and humans. The TSPO gene is located in the q13.3 regions
of the long arm of human chromosome 22 (Riond et al., 1991).
TSPO is distinct from the central benzodiazepine receptor (CBR) in pharmacological profile,
structure, subcellular localization (located on the plasma membrane of GABAergic (γ-
49
aminobutyric acid) neurons), and tissue distribution (Gatliff & Campanella, 2012).
Benzodiazepines such as clonazepam binds with high affinity to CBR, but with extremely
low affinity to TSPO (Benavides et al., 1983) although diazepam binds to both TSPO and
CBR with high affinity.
Endogenous molecules that bind to TSPO include the diazepam binding inhibitor (DBI) or
endozepine (Guidotti et al., 1983), cholesterol and porphyrins. DBI is widely distributed in
the CNS (predominantly in glial cells) and in peripheral organs, especially steroidogenic
cells.
TSPO is located primarily in the outer mitochondrial membrane (Antkiewicz-Michaluk,
Mukhin, Guidotti, & Krueger, 1988; Bribes et al., 2004; Garnier et al., 1994). It shows high
levels in glandular and secretory tissues such as the pineal gland, adrenal gland, salivary
glands, olfactory epithelium and gonads, intermediate levels in renal and myocardial tissue
and relatively low levels in the liver and brain. In the brain parenchyma, TSPO is located in
the outer mitochondrial membrane of glial cells (Chen & Guilarte, 2008). However, it was
also detected in the nucleus and perinuclear area in glial cells in the CNS (Kuhlmann &
Guilarte, 2000). Additionally, TSPO is expressed in circulating blood cells with the highest
concentrations in monocytes and neurotrophils (Canat et al., 1993).
TSPO are thought to be involved in a lot of physiological functions including cell growth and
proliferation, steroidogenesis, mitochondrial respiration, and apoptosis. The fact that TSPO
knockout mice die at an early embryonic stage (Papadopoulos et al., 1997) suggests that
TSPO is involved in basic house-keeping functions and is essential for embryonic
development (Chen & Guilarte, 2008). Among many potential physiological functions of
TSPO, steroidogenesis is best characterized. Abundantly expressed in steroidogenetic tissues,
TSPO mediates cholesterol transport into mitochondria (Papadopoulos et al., 1997). TSPO
can bind cholesterol and facilitate the transport of cholesterol from the outer to the inner
mitochondrial membrane, the rate-determining step of steroidogenesis. The side chain
cleavage cytochrome P-450 enzyme, located in the inner mitochondrial membrane, can
convert cholesterol to pregnenolone (steroid precursor), and initiate steroidogenesis such as
neurosteroid production in glial cells (Rupprecht et al., 2009; Rupprecht et al., 2010).
50
1.6.3.2 Prototypical TSPO radioligand: [11C]-PK11195
[11C]-PK11195 was the first and has been most commonly used PET radioligand for the
investigation of TSPO density (Chauveau, Boutin, Van Camp, Dolle, & Tavitian, 2008).
Normal brain expresses minimal [3H]-PK11195 binding (Banati et al., 2000) and TSPO
expression (Cosenza-Nashat et al., 2009). In various pathological post mortem brains,
elevated [3H]-PK11195 binding was associated primarily with activated microglia (Banati et
al., 2000; Venneti, Wang, & Wiley, 2008). Microglial activation in vitro increased the
number of TSPO without changing binding affinity of PK11195 (Banati et al., 2000).
Although increased [11C]-PK11195 binding in vivo has been reported in various neurological
diseases such as MS (Banati et al., 2000), ischemic stroke (Gerhard et al., 2000),
amyotrophic lateral sclerosis (Turner et al., 2004), Human immunodeficiency virus (HIV)
dementia (Hammoud et al., 2005), frontotemporal dementia (Cagnin, Rossor, Sampson,
Mackinnon, & Banati, 2004), PD (Gerhard et al., 2006), and AD (Cagnin et al., 2001), [11C]-
PK11195 binding was not different between HC subjects and patients with mild cognitive
impairment (Schuitemaker et al., 2004; Schuitemaker et al., 2006), suggesting that the
radioligand may not be sensitive to detect mild forms of neuroinflammation (Venneti et al.,
2008).
In fact, this radioligand has several important limitations including a high nonspecific
binding (Chen, Baidoo, Verina, & Guilarte, 2004), low total brain uptake (Maeda et al.,
2004; Petit-Taboue et al., 1991), and a poor signal-noise ratio (Banati et al., 2000; Kropholler
et al., 2006; Pappata et al., 1991; Venneti et al., 2013). These limitations of [11C]-PK11195
have led to developing new radioligands with greater sensitivity and specificity to detecting
TSPO binding since neuroinflammatory processes are common pathologies in various brain
diseases and disorders (James, Selleri, & Kassiou, 2006; Okubo, Yoshikawa, Chaki,
Okuyama, & Nakazato, 2004).
1.6.3.3 Second generation TSPO radioligands
To date, there are a number of second generation TSPO radioligands have developed
including [11C]PBR28 (Imaizumi, Kim et al., 2007), [18F]-FEPPA (discussed in section
51
1.6.3.5) (Wilson et al., 2008), [11C]-DAA1106 (Maeda et al., 2004), [11C]-DAA1106 (Zhang
et al., 2003), [18F]-FEDAA1106 (Zhang, Maeda et al., 2003), [11C]-DPA 713 (Boutin,
Chauveau, Thominiaux, Gregoire et al., 2007), [11C]-CLINME (Boutin, Chauveau,
Thominiaux, Kuhnast et al., 2007), [18F]-PBR06 (Imaizumi, Briard et al., 2007), [18F]-GE180
(Wadsworth et al., 2012), [18F]-DPA 714 (James et al., 2008), [18F]-F-VUIIS1008 (Tang,
Nickels, Tantawy, Buck, & Manning, 2014) and [18F]DPA-C5yne (Medran-Navarrete et al.,
2014).
Compared with [11C]-PK11195, [11C]-DAA1106 (Venneti, Lopresti et al., 2007; Venneti,
Wagner et al., 2007; Venneti, Wang, Nguyen, & Wiley, 2008), [18F]-FEDAA1106 (Fujimura
et al., 2006), [11C]-PBR28 (Parente et al., 2016), [18F]-PBR06 (Imaizumi et al., 2007a), [18F]-
DPA 714 (James et al., 2008), [18F]-GE180 (Wadsworth et al., 2012), and [18F]-DPA-C5yne
(Medran-Navarrete et al., 2014) showed higher affinity; [11C]-PBR28 (Parente et al., 2016)
and [18F]-DPA 714 (James et al., 2008) showed higher brain uptake; as well as [11C]-DPA
713 (Boutin et al., 2007a; Doorduin et al., 2009), [11C]-CLINME (Boutin et al., 2007b),
[18F]-PBR111 (Van Camp et al., 2010), [18F]-DPA714 (Corcia et al., 2012) and [18F]-GE180
(Boutin et al., 2015; Liu et al., 2015) demonstrated a better contrast between lesioned and
non-lesioned sites of the brain.
Some of these radioligands have been evaluated in vivo in healthy human individuals.
Compared to [11C]-PK11195, [11C]-PBR28 and [11C]-DPA-713 showed a larger brain uptake
(Endres et al., 2009; Fujita et al., 2008) and signal (Endres et al., 2009). However, shorter
half-life of [11C] radiolabelled ligands limits their dissemination and thus, [18F] radiolabelled
ligands are preferred for wide clinical applications. [18F]-DPA 714 demonstrated the
favorable pharmacology and biodistribution (Arlicot et al., 2012). However, it was quantified
using the cerebellum as a reference region, which is likely to produce the underestimation of
specific binding (Arlicot et al., 2012), as there is no brain region devoiding TSPO (see
section 1.6.3.5). Despite of improvements of sensitivity and specificity for TSPO, other
second-generation TSPO radioligands also showed some limitations such as a slow
accumulation of radiometaboliltes of [18F]-PBR06 in the brain, which compromises the
accurate estimation of TSPO (Fujimura et al., 2006; Fujimura et al., 2009), and slow kinetic
52
behavior of [18F]-FEDAA1106 (Fujimura et al., 2006), which likely requires a longer scan
and thus may not be suitable for clinical populations.
Interestingly, the [11C]-PBR28 study reported “non-binders” who appeared to lack the
binding site or lack of the TSPO proteins in both brain and peripheral organs (Fujita et al.,
2008). In this study, there were two such individuals out of 10 participants. The authors also
reported that since the time of the study, there were four such participants out of 29. This
study led to finding of three binding affinity patterns of second-generation TSPO
radioligands depending on the TSPO polymorphism. The null findings of the studies in
clinical populations using [18F]- FEDAA1106 (Takano et al., 2013; Varrone et al., 2013) and
[18F]-DPA714 (Golla et al., 2015) may have been due to the lack of information about the
binding affinity of study participants.
1.6.3.4 TSPO polymorphism and binding affinity of second generation
TSPO radioligands
Following the finding of “non-binders” of [11C]-PBR28 (Fujita et al., 2008), binding
characteristics of PBR28 was compared with PK11195 in autoradiographic studies using MS
human brain tissue (Owen et al., 2010) and revealed that 5 of the 22 donors (23%) did not
show a measurable binding signal of [3H]PBR28 while they showed strong binding with
[3H]PK11195. These non-binding tissue samples showed a significant reduction in the
affinity of [3H]-PBR28 compared to those showing normal signal. In addition, 40% of those
with normal signal presented both high-affinity and low-affinity sites. It is noted that there
was no difference in TSPO density (Kd or Bmax) for [3H]-PK11195 among the study
samples. These binding patterns were characterized by high-affinity, low-affinity, and
mixed-affinity (two-site) binding. The subsequent study extended the findings to further
confirm that the three binding affinity patterns existed (1) in normal brain tissue of 20
subjects, (2) using other TSPO radioligands such as PBR06, DAA1106, DPA713, and
PBR111 in the brain tissue although differences in affinity between high affinity binders
(HABs) and low affinity binders (LABs) varied, (3) in the platelets of healthy subjects and
(4) in the outcome measures of [11C]-PBR28 PET scans of healthy subjects (Owen, Gunn,
Rabiner, Bennacef, Fujita, Kreisl, Innis, Pike, Reynolds, Matthews, & Parker, 2011).
53
These different binding affinity patterns can be predicted by the TSPO polymorphism (Owen
et al., 2012). Twenty polymorphisms in TSPO gene were analyzed for association with the
binding phenotype of PBR28 in platelets of 41 healthy subjects (Owen et al., 2012). HABs
and LABs were defined as subjects with a single binding site with Ki <15 or > 100 nmol/L,
respectively because there were no HABs with Ki >15 or no LABs with Ki > 100 (Owen et
al., 2011). Mixed- affinity binders (MABs) were defined as subjects with two binding sites.
Only one polymorphism (rs6971) survived the multiple testing P value threshold, which is
located in exon 4 of the TSPO gene. It results in a nonconservative amino-acid substitution at
position 147 from alanine to threonine (Ala147Thr) in the fifth transmembrane domain of the
TSPO protein. This study also showed the complete agreement between rs6971 genotype and
PBR28 binding phenotype in the platelets. This polymorphism also predicted PBR28 binding
affinity class in the leukocytes membranes as well as [11C]-PBR28 binding affinity class in
brain where HABs showed significantly higher mean VT values than MABs in the whole
brain and cortical regions including the medial temporal, cingulate, temporal, frontal, parietal
and occipital regions (Kreisl et al., 2013). Similarly, significant group differences in [18F]-
PBR111 VT values between HABs and MABs were reported in the global cortical region,
cerebellum, hippocampus, thalamus, brain stem and CN in healthy individuals (Guo et al.,
2013). Furthermore, [3H]-PBR28 binding in postmortem brain tissues from patients with
schizophrenia (SCZ) and HC subjects showed significant group difference after binding
affinity was included as a covariate in the statistical model. Taken together, accounting for
the TSPO polymorphism likely increases the chance to detect group differences.
The prevalence of HAB, MAB, and LAB depends on ethnicity. According to the HapMap
project, HAB, MAB, and LAB account for 46%, 38%, and 15% in Caucasian population; in
African American population, 78%, 20%, and 2%; in both Japanese and Han Chinese
population, 98%, 2%, and less than 0.1%, respectively. Thus, this binding affinity variation
has the greatest impact on studies of Caucasians.
The interindividual variability in binding affinity observed in second-generation TSPO
radioligands was not evident in [11C]-PK11195. This may be because second-generation
TSPO radioligands and [11C]-PK11195 bind to different sites (Kreisl et al., 2010; Owen et
al., 2010; Owen, et al., 2011).
54
1.6.3.5 [18F]-FEPPA
[18F]-FEPPA is also one of these second generation TSPO radioligands developed to
overcome the limitations of the prototypical [11C]-PK11195 at the Centre for Addiction and
Mental Health (CAMH) (Wilson et al., 2008). The study 3 presented the preliminary data
using this radioligand to investigate neuroinflammation in the disease-associated regions in
PD patients.
1.6.3.5.1 Evaluations in vitro and ex vivo of [18F]-FEPPA
Evaluation studies in vitro and ex vivo demonstrated that [18F]-FEPPA is a promising TSPO
radioligand (Wilson et al., 2008). The affinity of [18F]-FEPPA (Ki = 0.07 nM) was threefold
more potent than [11C]-PBR28 (Ki = 0.22 nM) and an order of magnitude more potent than
[11C]-DPA713 (Ki = 0.87 nM) and the prototypical PK11195 (Ki = 1.29 nM). A
lipophilicity of [18F]-FEPPA (logD = 2.99) was suitable for brain penetration. Unmetabolized
[18F]-FEPPA accounted for 95.4% while unmetabolized [11C]-PBR28, only 84.6% at 40
minutes post injection. The metabolites were less lipophilic than [18F]-FEPPA in plasma and
brain extracts of rats as well as in human brains (Rusjan et al., 2011). The biodistribution of
[18F]-FEPPA study demonstrated a reasonable BBB penetration, brain uptake and slow
washout, as well as ubiquitous distribution of TSPO with high levels of radioactivity in the
olfactory bulb and hypothalamus, which is consistent with previous literature (Benavides et
al., 1983; Chaki et al., 1999; Venneti et al., 2007). Furthermore, in pig brains, [18F]-FEPPA
showed good and rapid brain uptake as well as appropriate regional distribution with the
highest uptake found in the thalamus while the lowest, in the cerebellum and frontal cortex.
Pretreatment with PK11195 resulted in a significant reduction in binding, confirming the
specific binding and reversible kinetics of [18F]-FEPPA (Bennacef et al., 2008). Lastly, in rat
model of PD, [18F]-FEPPA demonstrated greater sensitivity and selectivity to detect
inflammation compared with [11C]-PK11195. [18F]-FEPPA uptake was higher in the lesion
site, which was also more closely correlated with inflammatory cytokines, and lower in the
non-lesion site than [11C]-PK11195 (Hatano et al., 2010).
55
1.6.3.5.2 Evaluation of quantification in vivo of [18F]-FEPPA in human
gray matter
Quantification of [18F]-FEPPA in human GM was evaluated to determine the appropriate (1)
kinetic modeling (1TCM or 2TCM), (2) outcome measures (VT, VS, or BPND), and (3) scan
length, as well as to determine the capacity for the quantification with increased specific
binding (Rusjan et al., 2011). Twelve healthy individuals consisting of four men and eight
women aged from 24 to 72 years old underwent a 180-min [18F]-FEPPA scan with arterial
sampling that generates a metabolite corrected plasma curve to be used as input function for
kinetic analyses. ROIs included the cerebellum, CN, putamen, frontal, temporal occipital,
insular, and ACC, and thalamus. %COV estimated the identifiability of the outcome
measures. Monte Carlo simulations were performed by increasing k3 in the thalamus that
showed the highest biding to determine the capacity of [18F]-FEPPA quantification.
Consistent with known distribution of TSPOs in human brain (Doble et al., 1987), the
radioactivity was widespread, which excludes reference region modeling for the kinetic
analysis of [18F]-FEPPA data. The 2TCM model provided better fit than the 1TCM model in
all of the ROIs, demonstrated by significantly lower AIC and higher Model Selection
Criterion. A two-hour scan appeared to be appropriate from the perspective of both
quantification of the outcome measures and the tolerance of subjects.
Using the two-hour scan data with 2TCM model, VT was the most appropriate binding
outcome measure as it showed better identifiability and less variability of the identifiability
across the ROIs compared to VS and BPND. It can also measure high level of TSPO density
(up to 500% increase in specific binding) with better identifiability than other outcome
measures. The highest VT binding was found in the thalamus followed by the temporal,
occipital, insular, and frontal cortex, cerebellum, ACC, putamen, and the lowest values were
found in the CN.
The study also demonstrated that radioactivity of skull was negligible, excluding a partial
volume effect in cortical ROIs. Furthermore, changes in blood blow did not affect the VT
quantification. This is important to study aging and pathological brains because increased
56
blood flow can occur due to increased vascular permeability of disrupted BBB by
neuroinflammation in diseased brains (Winkeler, Boisgard, Martin, & Tavitian, 2010) or
because reduced blood flow can occur by aging (Iwata & Harano, 1986; Takahashi,
Yamaguchi, Kobayashi, & Yamamoto, 2005).
1.6.3.5.3 Evaluation of quantification in vivo of [18F]-FEPPA in human
white matter
Unlike the quantification of TSPO density in human GM (Rusjan et al., 2011), that of the
WM is challenging. This is due to the low level of TSPO in WM and its very slow washout,
making the identification of the small k4 value difficult (Rusjan et al., 2011). All of these
make the TSPO quantification in WM susceptible to noise. However, given the importance
of studying neuroinflammation in WM in various pathological brains, the quantification of
[18F]-FEPPA binding to TSPO in WM was investigated under several noise levels in 32HC
subjects (Suridjan, Rusjan, Kenk et al., 2014). The WM ROIs included the CC, cingulum,
superior longitudinal fasciculus (SLF), and posterior limb of internal capsule (PLIC). Using
the 2TCM, identifiability for VT was reported to be moderate ranging from %COV 15-19%.
Approximately 10% of the subjects showed very low VT identifiability (%COV > 30 %)
although these data did not affect the results. The effects of noise increased VT variability
and decreased its identifiability. However, the noise-induced bias in VT was relatively small.
Under the highest noise level, 6% of the data did not fit reliably. A simulation of increased
TSPO density showed minimal effect on VT variability and identifiability regardless noise
level, supporting the potential utility of [18F]-FEPPA in the WM (Suridjan et al., 2014).
1.6.3.5.4 TSPO polymorphism and [18F]-FEPPA VT
The relationship between the TSPO polymorphism and VT was investigated in 20 healthy
individuals (Mizrahi et al., 2012). ROIs included the frontal, temporal, occipital, insular, and
ACC, cerebellum, putamen, CN, and thalamus. Genotyping the rs6971 polymorphism
revealed that the study sample consisted of 13 HABs, six MABs, and one LAB. There were
significant ROI and genetic effects on VT without significant interactions. MABs showed a
57
reduction of 27% (23-29%) in VT values on average across the ROIs compared to HABs.
However, there was overlap in VT values between MABs and HABs across the ROIs.
In contrast, there were no significant genetic effects on VT in any WM ROIs although MABs
showed on average 15% reduction in VT values than HABs (Suridjan et al., 2014). However,
excluding one MAB outlier, significant group differences appeared in the SLF, PLIC, and
cingulum, but not in CC. Considering differences in microstructure (e.g., myelin sheath in
WM), TSPO density, and kinetics of [18F]-FEPPA between GM and WM, a larger sample
size may be needed to test the polymorphism effects on VT (Suridjan et al., 2014).
1.6.3.5.5 [18F]-FEPPA study on neuroinflammation and aging
There was no significant age effect on [18F]-FEPPA VT values n 33 healthy individuals aged
from 19 to 82 years (22 HABs and 11 MABs) in the prefrontal cortex (PFC), dorsolateral
prefrontal cortex (DLPFC), and temporal cortex or hippocampus (Suridjan, Rusjan,
Voineskos et al., 2014) or in 32 healthy individuals aged from 19 to 78 years in the CC,
cingulum, SLF, and PLIC adjusted for the TSPO polymorphism (Suridjan et al., 2014). There
were also no interactions between age and the GM ROIs or between age and binding affinity
status in any of the GM ROIs.
1.6.3.5.6 [18F]-FEPPA studies in clinical populations
In addition to a case study of a patient with tumor (Ko et al., 2013), clinical populations such
as patients with SCZ (Kenk et al., 2015), AD (Suridjan et al., 2015), and major depression
episodes (MDE) (Setiawan et al., 2015) were investigated using [18F]-FEPPA. In two later
studies, the patients showed elevated TSPO binding in the disease/disorder related ROIs
compared to HC subjects and the TSPO binding was associated with clinical measures in
these patients (cognitive performance in AD patients and depression severity in MDE
patients). These two studies demonstrated the promise of [18F]-FEPPA to detect
neuroinflammation in clinical populations although there was the lack of evidence of
neuroinflammation in patients with SCZ, possibly due to the effects of antipsychotic
medications that can exert inhibitory effects on proinflammatory processes, and thus future
studies in medication-naïve patients are warranted (Kenk et al., 2015).
58
1.6.3.5.7 TSPO imaging in Parkinson’s disease
To date, eight studies have been published for the investigation of neuroinflammation in PD.
All of these studies used [11C]-PK11195 and the results were inconclusive. Some studies
found that drug-naïve, early stage PD patients showed increased neuroinflammation in the
midbrain/SN (Iannaccone et al., 2013; Ouchi et al., 2005) and putamen (Iannaccone et al.,
2013) compared to HC subjects. The elevated neuroinflammation in the midbrain was also
associated with more severe motor symptoms as well as with more DA dysfunction in the
putamen (Ouchi et al., 2005). On the other hand, another study did not find significantly
elevated neuroinflammation in a similar sample size of PD patients with the clinical
characteristics (Bartels et al., 2010).
Two studies included more advanced stages of PD patients (Hoehn and Yahr stage 2 and 2-3)
did not find increased neuroinflammation in the nigro-striatal regions (Bartels et al., 2010;
Kobylecki et al., 2013). Instead, one of the studies found increased neuroinflammation in the
pons in the PD patients (Kobylecki et al., 2013). Another study including PD subjects with
moderate severity showed significant increase in neuroinflammation in the cortical regions
including the temporal and occipital regions using the ROI analysis (Edison et al., 2013) . A
voxel-based whole brain analysis further revealed that the PD patients showed increased
neuroinflammation in the frontal, temporal and occipital regions.
In a relatively larger sample size of 18 PD patients with varied severity, duration, and stages
of disease (including four drug naïve patients), neuroinflammation occurred in the striatum,
palllidum, thalamus, frontal and cingular regions, and pons while a whole brain region
analysis additionally found neuroinflammation in the hippocampus and cerebellum in these
patients (Gerhard et al., 2006). Interestingly, neuroinflammation decreased with longer
duration of disease in the cingular regions and thalamus.
Two of the studies mentioned above followed up some of the PD patients. In the study by
Ouchi, et al (2005), four PD patients were followed up four years later and showed the
consistent results (i.e., elevated neuroinflammation level in the midbrain, and the correlation
with motor severity and presynaptic DA dysfunction in the putamen). Additionally, higher
neuroinflammation was detected in the putamen and in the thalamus compared to the HC
59
subjects and compared to their first assessment, suggesting a spread of neuroinflammation
within and beyond the nigrostriatal pathway (Ouchi et al., 2005). On the other hand, in the
other study where eight of the patients were followed up 18-28 months later, there were no
changes in neuroinflammation whereas the disease progressed, evidenced by an increase in
motor severity and further reduction in presynaptic DA dysfunction in the putamen in these
patients (Gerhard et al., 2006).
In more severe cases of PD patients with dementia, neuroinflammation primarily occurred in
cortical regions, which was also associated with MMSE scores (Edison et al., 2013; Fan,
Lindemann, Feuilloley, & Papadopoulos, 2012) suggesting that this outcome measure be
used to assess cognitive symptoms in PD. Interestingly, however, not all PD with dementia
(PDD) showed increased neuroinflammation (Edison et al., 2013). The authors argued that
the finding might be due to insufficient sensitivity of [11C]-PK11195.
As mentioned earlier, neuroinflammation has been studied in PD patients using [11C]-
PK11195 only. Studies using a second generation TSPO radioligand warrant elucidating
whether neuroinflammation occurs in PD patients. For this purpose, we used [18F]-FEPPA
(see section 2.3) to investigate the striatal regions in PD patients accounting for the TSPO
polymorphism.
60
2.0 Aims and hypotheses
The general aim of this thesis was to use multimodal imaging tools to investigate structural,
functional, and molecular changes in PD and the associations with clinical manifestations.
The specific aim of each study was (1) to detect structural changes using CTA, which is
more sensitive for subtle cortical changes than the most commonly used VBM for gray
matter changes, and TBSS for white matter changes to elucidate the relationships between
structural abnormalities and clinical manifestations in PD, (2) to investigate changes in brain
network properties using the most advanced FC analysis method, graph theoretical analysis,
and (3) to first assess whether neuroinflammatory changes occur in PD using a second-
generation TSPO radioligand, developed to overcome the limitations of the prototype TSPO
radioligand.
2.1: Study1: Imaging changes associated with cognitive
abnormalities in Parkinson’s disease
The study 1 aimed to determine whether structural changes occurred in PD patients and
whether observed structural abnormalities were associated with the motor and cognitive
symptoms by investigating the whole brain GM and WM using CTA and TBSS. We
hypothesized that PD patients would show GM and WM abnormalities compared to age-
matched HC subjects and these changes would be correlated with clinical and cognitive
measures in PD patients.
61
2.2 Study 2: Disrupted nodal and hub organization account for
brain network abnormalities in Parkinson’s disease
The study 2 aimed to determine whether the disease-associated nodal connections and
network hubs were affected in PD patients, and whether these changes were associated with
motor and cognitive symptoms using the graph theoretical approaches to rsfMRI data. We
hypothesized that PD would affect the nodal organization of regions such as SMA, Ins,
DLPFC, as well as the striatum, and the nodal and hub changes would account for motor and
cognitive symptoms of PD.
2.3: Study 3: Imaging striatal microglial activation in patients
with Parkinson’s disease
The study 3 aimed to evaluate whether a second generation TSPO radioligand, [18F]-FEPPA
could be used to detect neuroinflammation in PD patients. It would determine whether there
would be significant differences in VT values between HABs and MABs in PD patients like
those observed in HC subjects and in other clinical populations, and whether [18F]-FEPPA
VT values could differentiate PD patients from HC subjects. Using the striatum as our main
ROI, we hypothesized that PD-HABs would show higher VT values than PD-MABs, PD
patients would show increased neuroinflammation than HC subjects accounting for the
TSPO polymorphism, and the neuroinflammation would correlate with clinical measures of
PD.
62
3.0 Study 1: Imaging changes associated with cognitive
abnormalities in Parkinson’s disease
This study was published in Brain Structure and Function.
This chapter has been modified from the following:
Koshimori, Y., Segra, B., Christopher, L., Lobaugh, N., Duff-Canning, S., Mizrahi, R.,
… Strafella, A. (2015). Imaging changes associated with cognitive abnormalities in
Parkinson’s disease. Brain Structure and Function, 220(4), 2249-61.
3.1 Introduction
PD is a multisystem neurodegenerative disease, affecting not only dopaminergic nerve cells
of the SN, but also other brain regions and neurotransmitters (Braak et al., 2006). PD
presents with the cardinal motor symptoms such as tremor, rigidity, bradykinesia, and loss of
postural stability along with a set of NMS (Bonnet, Jutras, Czernecki, Corvol, & Vidailhet,
2012) such as cognitive impairment, depression, sleep disturbances, and autonomic
dysfunction (Barnum & Tansey, 2012; Chaudhuri, Odin, Antonini, & Martinez-Martin,
2011; Ferrer, Martinez, Blanco, Dalfo, & Carmona, 2011). For this reason, PD research has
expanded its investigation beyond the nigrostriatal region to the whole brain in order to
characterize the different symptoms. In particular, neuroimaging investigation of cognitive
impairment is a topic of a growing interest (Christopher & Strafella, 2013). It is prevalent
and present regardless of the disease stage (Litvan et al., 2012). It ranges from mild deficits
demonstrable by means of comprehensive neuropsychological testing, to dementia (Jellinger,
2012), and MCI increases risk of developing dementia (Janvin, Larsen, Aarsland, &
Hugdahl, 2006). The neural substrates of motor and cognitive symptoms has been
investigated using non-invasive brain imaging such as structural MRI, which can offer the
opportunity of identifying early biomarkers. In fact, different MRI techniques are providing
63
amounting evidence of both GM and WM changes in PD (Cochrane & Ebmeier, 2013; Pan et
al., 2012).
Whole brain GM changes have been investigated using VBM as well as surface-based
analyses including CTh and surface analyses combined with subcortical volumetric analysis.
PDD typically show bilateral diffuse GM changes (Beyer, Janvin, Larsen, & Aarsland, 2007;
Compta et al., 2012; Melzer et al., 2012; Song et al., 2011; Zarei et al., 2013). On the other
hand, non-demented PD patients show regional GM changes in areas such as frontal (Biundo
et al., 2011; Burton et al., 2004; Ibarretxe-Bilbao, Ramirez-Ruiz et al., 2010; Jubault et al.,
2009), temporal and/or limbic regions (Feldmann et al., 2008; Ibarretxe-Bilbao et al., 2009;
Pellicano et al., 2012; Tinaz et al., 2011; Wattendorf et al., 2009), or posterior regions
including the parieto-occipital cortex (Pereira et al., 2012). Furthermore, non-demented PD
patients show faster progression of GM changes including atrophy and cortical thinning than
healthy controls (HCs) in diffuse areas including frontal, temporal and parietal regions (Hu et
al., 2001; Ibarretxe-Bilbao et al., 2012), restricted regions to cortical motor areas and
cerebellum (Ibarretxe-Bilbao, Junque et al., 2010), or limbic, paralimbic, and temporo-
occipital regions (Ramirez-Ruiz et al., 2007). Among PD patients, variability in CTh and/or
subcortical volume is also associated with cognitive measures (Biundo et al., 2011; Camicioli
et al., 2009; Ibarretxe-Bilbao et al., 2009; Melzer et al., 2012; Pellicano et al., 2012; Zarei et
al., 2013), facial emotion recognition (Baggio et al., 2012), duration of disease (Hanganu et
al., 2013; Jubault et al., 2011; Lyoo et al., 2011), motor severity (Lyoo et al., 2011; Melzer et
al., 2012; Zarei et al., 2013) and stage (Zarei et al., 2013) as well as DA non-responsive
symptoms (Brenneis et al., 2003; Camicioli et al., 2009).
WM changes have been investigated using DTI. The two most common indices derived from
DTI are FA and or MD (Cochrane & Ebmeier, 2013). FA estimates the degree of anisotropic
directionality of water diffusion while MD estimates the magnitude/size of water diffusion.
FA and MD in the whole brain WM can be assessed using voxel-based analysis (VBA) or
TBSS. Contrary to GM findings, in which regional changes are more common in non-
demented PD, WM changes are reported in more diffuse brain areas even in non-demented
PD (Hattori et al., 2012; Kim et al., 2013; Melzer et al., 2013; Theilmann et al., 2013; Zheng
et al., 2014), and the frontal region was one of the most consistently reported regions for
64
WM changes (Agosta et al., 2014; Deng et al., 2013; Gattellaro et al., 2009; Rae et al., 2012;
Zhan et al., 2012; K. Zhang et al., 2011). Moreover, variability in FA and/or MD was
primarily associated with cognitive measures (Agosta et al., 2014; Gallagher et al., 2013; Rae
et al., 2012; Theilmann et al., 2013).
Despite the evidence of structural abnormalities and the variability associated with clinical
and cognitive manifestations in non-demented PD, whether structural abnormalities account
for specific clinical sequelae in PD patients is still unclear. This is largely due to the lack of
imaging studies investigating both structural group differences, and relationships between
these differences and various clinical and cognitive measures. For example, the anatomical
regions displaying significant group differences in structure have not been specifically
investigated for correlation analyses (Camicioli et al., 2009; Hanganu et al., 2013; Jubault et
al., 2011; Tinaz et al., 2011). Even though brain regions demonstrating significant group
differences have been shown to correlate with cognition, PD patients did not show
impairment on cognitive tasks (Theilmann et al., 2013). Furthermore, only correlations were
demonstrated in PD without comparing structural data of PD with those of HCs (Lyoo et al.,
2011; Zheng et al., 2014). To date, only a handful of studies have addressed which structural
abnormalities account for different PD symptoms by investigating whole brain GM and WM
changes (Agosta et al., 2013; Agosta et al., 2014; Hattori et al., 2012). Using both VBM and
TBSS analyses, these studies have consistently concluded that WM and not GM changes
underlie cognitive impairment in PD (Agosta et al., 2013; Agosta et al., 2014). This
discrepancy may be due to the fact that VBM may not be highly sensitive for detecting subtle
cortical atrophy in early stages of PD (Agosta et al., 2013; Jubault et al., 2011). In fact, CTA
appeared more prone to detect cortical GM changes in PD than VBM when both analysis
methods were compared (Pereira et al., 2012). Thus based on these preliminary observations,
the current study aimed at (1) investigating both GM and WM changes using CTA and
TBSS, and thereby also demonstrating which MRI technique can be a promising biomarker
for PD and (2) further elucidating the relationships between observed structural
abnormalities and clinical and cognitive manifestations in non-demented PD patients.
65
3.2 Methods
3.2.1 Participants
Twenty-six patients meeting UK Brain Bank criteria for the diagnosis of idiopathic PD
(Defer, Widner, Marie, Remy, & Levivier, 1999; Langston et al., 1992) and 15 HCs
participated in the study. PD patients were recruited from the Movement Disorders Clinic of
the Toronto Western Hospital. The HCs were recruited from friends and spouses of the
patients or through advertisements posted at the hospital and the affiliated university.
Exclusion criteria included (1) history of a head injury, psychiatric, neurological or major
medical diseases; (2) dementia assessed by a modified Disability Assessment for Dementia
(DAD) with an additional question regarding whether any reported impairment was related to
cognitive difficulties or the physical impairments of PD; (3) contraindications for MRI
scanning; and (4) for HCs evidence of cognitive impairment as assessed by a
neuropsychological test battery. All participants underwent a cognitive assessment by means
of an extended neuropsychological test battery and the Montreal Cognitive Assessment
(MoCA) (Nasreddine et al., 2005; Tison et al., 1995) as well as the Beck Depression
Inventory (BDI) (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) that assesses levels of
depression. PD patients were additionally evaluated for motor severity of the disease using
the motor subset of the UPDRS (UPDRS-III). All participants underwent structural MRI
scans and 35 participants had additional DTI scans. Among the 35 participants, five subjects
(four PD and one HC) were excluded due to motion artifacts resulting in 16 PD patients and
14 HCs included in the DTI analysis. PD patients underwent all the study procedures in an
“on-medication” state”. All participants gave informed consent following full explanation of
the study procedures. This study was approved by the institutional ethics committee of the
CAMH and the University Health Network.
3.2.2 Neuropsychological Assessment
Cognitive function in the domains of executive function, attention/working memory,
language, visuospatial function and memory were assessed using the following
neuropsychological tests (Litvan et al., 2012). For executive function: Visual Verbal Test
66
total number of shifts (Wicklund, Johnson, & Weintraub, 2004), Delis-Kaplan Executive
Function System (D-KEFS) Color-Word Interference (Stroop task), time to complete
condition 3: inhibition, and D-KEFS Verbal Fluency, total score for category fluency
(Delis, Kaplan, & Kramer, 2001). For attention and working memory: Wechsler Memory
Scale-III (WMS-III) Digit Span and Letter-Number Sequencing total score (Wechsler,
1997), D-KEFS Color-Word Interference (Stroop task), and time to complete condition 1:
color naming (attention). For language: the category fluency from the Verbal Fluency
subtest (Delis et al., 2001). For visuospatial function: Judgment of Line Orientation (JLO)
total score (Benton, Sivan, Hamsher, Varney, & Spreen, 1994). For memory: California
Verbal Learning Test-II (CVLT-II) long delay free recall score (Delis, Kramer, Kaplan, &
Ober, 2002). Global composite z-scores were calculated for each domain of cognition.
When the cognitive domain had only one test, the individual test score was used. For
executive function the average of the z-scores from the visual verbal test, inhibition
segment of the Stroop task, and verbal fluency was used, while for attention/working
memory, the average of the z-scores from the digit span, letter-number sequencing, and
color-naming segment of the Stroop was used.
3.2.3 MRI Acquisition
Whole-brain T1-weighted and DWI images were acquired using a 3.0 T General Electric
(GE) Signa HD x MRI system (General Electric, Milwaukee, WI) equipped with an eight-
channel phased array head coil. For GM analysis, a high-resolution 3D anatomical scan was
acquired with a T1-weighted 3D IR-Fast Spoied Gradient-Recalled-Echo sequence
(TR/TE/inversion time (TI), 7.8/min full/450 ms; matrix, 256 x 256; voxel size, 1 x 1 x 1
mm; field of view (FOV), 256 x 256 mm; flip angle, 15°; 180 axial slices). For WM analysis,
a DWI scan was acquired with spin-echo single-shot echo planar imaging with diffusion
encoding in 60 noncolinear directions (TR/TE, 17,000/min ms; FOV, 230 x 230 mm; matrix,
128 x 128, voxel size, 1.8 x 1.8 x 2.4 mm; b value, 1000 s/mm2; 64 slices). Parallel imaging
was employed using the Array Spatial Sensitivity Encoding Technique (ASSET) with an
acceleration factor of two. DWI images were acquired in the axial plane. Additionally, ten
non-DWI scans were acquired at the beginning of each scan. The DWI scans were repeated
67
three times to increase signal-to-noise ratio. Two PD patients and two HCs underwent only
two DWI scans due to inability to lie down in the scanner for an extended period of time.
3.2.3.1 Cortical Thickness Analysis
CTA was performed using the FreeSurfer image analysis suite (version 5.1; available at
http://surfer.nmr.mgh.harvard.edu). The processing of T1 high-resolution images for the
cortical surface reconstruction involved several steps (Dale, Fischl, & Sereno, 1999):
automated Talairach transformation, intensity normalization (Sled, Zijdenbos, & Evans,
1998), skull stripping, and WM segmentation, tessellation of the GM/WM boundary,
automated topology correction (Fischl, Liu, & Dale, 2001; Segonne, Pacheco, & Fischl,
2007), and surface deformation following intensity gradients to optimally place the
gray/white and gray/CSF borders at the location (Dale & Sereno, 1993; Dale et al., 1999;
Fischl & Dale, 2000). All surface models were visually inspected for accuracy. CTh was
calculated as the closest distance from the gray/white boundary to the gray/CSF boundary at
each vertex on the tessellated surface (Fischl & Dale, 2000).
3.2.3.2 Subcortical Volume Analysis
Segmentation of brain volume was obtained based on the automatic procedure included in
FreeSurfer (version 5.1; available at: http://surfer.nmr.harvard.edu) (Fischl & Dale, 2000).
The labels were created using an automated subcortical labeling algorithm based on a
probabilistic atlas obtained from a manually labeled training set. The image was rigid-body
registered to the probabilistic brain atlas, followed by non-linear morphing to the atlas. Then,
an automated segmentation procedure assigned a label to each voxel in a dataset based on
signal intensity information and the spatial relationship of the subcortical labels in the
training sets. Volumetric measures from 18 structures in each hemisphere as well as
intracranial volume (ICV) were automatically obtained. Among these structures, we selected
the thalamus, caudate, putamen, pallidum, NA, hippocampus, amygdala and brainstem for
our ROIs.
68
3.2.3.3 TBSS
Individual FA and MD images were generated to conduct TBSS analysis using FSL tools
from the FMRIB software library (FSL version 4.1.5, http://www.fmrib.ox.ac.uk/) (Smith et
al., 2004; Woolrich et al., 2009). First, each volume upsampled to create isotropic voxel size
(2.4 x 2.4 x 2.4 mm) was affine registered to the 9th b0 volume using FSL Linear
Registration Tool (FLIRT) to correct motion and eddy current distortion (Jenkinson & Smith,
2001; Jenkinson, Bannister, Brady, & Smith, 2002). Then, an averaged DWI image was
generated from the three sets of DWI images. Non-brain tissue of the averaged DWI image
was removed using BET, brain extraction tool (Smith, 2002), and FA and MD images were
derived using DTIfit from FMRIB’s Diffusion Toolbox. The FA maps of all participants
were warped to the FMRIB58_FA template using FSL non-linear registration tool (FNIRT)
(Andersson, Jenkinson, & Smith, 2007a; Andersson, Jenkinson, & Smith, 2007b). A mean
FA map of all subjects was created and thinned to generate a mean FA skeleton, which
represents the centres of WM tracts common to all subjects included in the present study.
The mean skeleton was thresholded and binarized at FA value of 0.2 to minimize partial
voluming. The individual FA and MD maps were projected onto the mean skeleton resulting
in a skeletonised FA and MD maps.
3.2.3.4 Statistical Analysis
The CTA was performed using Freesurfer’s QDEC application that fits a general linear
model (GLM) at each surface vertex. CTh maps were smoothed using a circularly symmetric
Gaussian kernel across the surface with a full width at half maximum (FWHM) of 15 mm. Z
Monte Carlo simulations with 10,000 iterations were applied to CTh maps to provide
clusterwise correction for multiple comparisons, and the results were thresholded at a
corrected P value of 0.05 (Z = 1.3). Mean thickness was calculated for each significant
cluster.
Voxelwise statistics were performed on the skeletonized FA and MD maps using the
Threshold Free Cluster Enhancement (TFCE) (S. M. Smith & Nichols, 2009). A P < 0.05
voxelwise correction for multiple comparisons was considered significant. The Johns
69
Hopkins University (JHU) DTI-based WM atlases including the JHU WM tractography atlas
and the International Consortium for Brain Mapping (ICBM)-DTI WM labels (Hua et al.,
2008; Mori, Wakana, Nagae-Poetscher, & van Zijl, 2005; Wakana et al., 2007) were used to
identify WM tracts that showed significant group differences and correlations in FA and MD
values.
Age, years of education, MoCA, BDI and neuropsychological tests were compared between
groups using independent samples t tests. Gender and handedness were compared using
Pearson’s chi-square tests. Subcortical volume in a priori ROIs (thalamus, putamen, caudate,
nucleus accumbens, globus pallidus (GP), hippocampus, amygdala, and brainstem) in each
hemisphere was compared between groups using independent t test or Analysis of Variance
(ANOVA) after testing the effects of ICV on the subcortical volume in each ROI.
From the clusters displaying significant group differences mean individual values were
extracted to investigate the relationships between structural changes (CTh, subcortical
volume, and DTI) and (1) cognitive data (MoCA and neuropsychological test scores) and (2)
clinical data (UPDRS-III scores and duration of disease). Prior to the correlation analyses,
bivariate correlation was performed among all relevant covariates including demographic
(age, gender, years of education, handedness), clinical (symptom-dominant side, UPDRS-III
scores, disease duration, Levodopa equivalent daily dose (LEDD), BDI), cognitive (MoCA,
visual verbal test, JLO, and global and executive composite z) and MRI findings from the
group analysis using Pearson’s correlation tests to determine the effects of nuisance
variables, which would be then controlled for in the correlations, if necessary. All statistical
analyses were two-sided and statistical significance was set at P < 0.05 corrected for multiple
comparisons. All of the statistical analyses were conducted using Statistical Package for the
Social Sciences (SPSS 13.0).
3.3 Results
3.3.1 Demographic, Clinical, and Cognitive Characteristics
70
Table 3-1 shows the demographic and clinical characteristics as well as neuropsychological
data of all of 26 PD patients and 15 HCs. There was no difference between the two groups
regarding age, gender, education, BDI, and handedness (P > 0.05). However, they were
significantly different on MoCA scores (t = 2.88, P = 0.006), global composite z (t = 2.82, P
= 0.007), executive composite z scores (t = 3.48, P = 0.001), visual verbal test (t = 3.6, P <
0.001), and JLO (t = 2.3, P = 0.027).
Table 3-1. Demographic, clinical, and cognitive characteristics of PD patients and HCs.
PD (n = 26) HC (n = 15)
Age (years) 70.5 (5.6) 67.13 (5.1)
Sex (male/female) 13/13 4/11
Handedness (right/left) 24/2 14/1
Education (years) 15.6 (2.1) 17.0 (2.5)
MoCA 25.2 (2.8) 27.6 (2.2)**
BDI 6.5 (5.5) 3.8 (3.6)
Disease duration (years) 6.7 (4.2) -
Symptom-dominant side (right/left) 17/9
UPDRS-III (on-medication) 25.3 (15.3) -
Total LEDDa (mg/day) 731.3 (459.8) -
Neuropsychological tests
Global composite z -0.22 (0.60) 0.30 (0.50)**
Attention/working memory
composite z
0.12 (0.63) 0.27 (0.70)
Executive composite z -0.64 (0.89) 0.25 (0.60)**
71
Digit span forward 0.31 (0.73) 0.46 (1.02)
California verbal test 0.19 (1.06) 0.23 (0.92)
Letter-number sequencing 0.20 (0.77) 0.15 (0.68)
Visual verbal test -2.21 (1.93) -0.380 (0.48)**
Judgment of line orientation -0.78 (1.57) 0.24 (0.88)*
STROOP
Color naming -1.44 (0.86) 0.19 (1.17)
Inhibition 0.13 (0.72) 0.29 (1.00)
Category fluency 0.21 (1.21) 0.86 (0.83)
Data are presented in mean (standard deviation) and neuropsychological data are presented in
group based mean Z-scores (standard deviation). aLEDD: Levodopa dose + 1-Levodopa-CR
x 0.75 + pramipexole (mg) x 67 (Evans et al., 2004). * P < 0.03, ** P < 0.01.
3.3.2 Cortical Thickness
We found that the PD patients showed significant reduction in CTh in the left superior-
frontal gyrus (SFG) (cluster size: 1585 mm2; P = 0.03, Talairach coordinates of maxima: x =
-18, y = 35.2, z = 39.3; HC = 2.72 ± 0.03 and PD = 2.51 ± 0.04) and the left precentral gyrus
extending back into the postcentral gyrus (cluster size: 2555.42 mm2; P = 0.01, Talairach
coordinates of maxima: x = -58.8, y = -5.4, z = 9.5; HC = 2.76 ± 0.03 and PD = 2.6 ± 0.03)
compared with the HCs (Figure 3-1). The PD patients did not show significant increase in
CTh in any brain regions compared with the HCs.
We further investigated whether the CTh in these two significant clusters was correlated with
clinical and cognitive measures that showed significant group difference (i.e., MoCA, visual
verbal test, JLO, and global and executive composite z scores) in PD patients. To this end,
72
first, we ran bivariate correlations among different covariates including the CTh values in the
SFG and precentral gyrus clusters and demographic, clinical and cognitive measures to
determine the effects of nuisance variables. In particular, age showed significant negative
correlation both with SFG CTh (r = -0.497, P = 0.01) and cognitive scores including MoCA
(r = -0.423, P = 0.031), visual verbal tests (r = -0.499, P = 0.009), and executive z scores (r =
-0.541, P = 0.004). Thus controlling for the effect of age, we found significant positive
correlations between the SFG thickness and global composite z (r = 0.597, P = 0.002), and
executive composite z (r = 0.430, P = 0.032) scores (Figure 3-2). Similarly, we performed
the partial correlations between the SFG CTh and the cognitive scores, controlling for age in
HCs. In HCs, we found a significant positive correlation between the SFG thickness and
visual verbal test scores (r = 0.607, P = 0.016). These findings suggest that the SFG thinning
was uniquely associated with poor executive and global cognitive performance in PD
patients.
73
Figure 3-1. a. Cortical areas showing significant cortical thinning in PD patients compared
to HCs. Color bar indicates the significance levels in the clusters in Z values. b. Bar
graphs on extracted CTh values (mm) from the significant clusters in the left SFG and left
precentral gyrus between HCs and PD patients. Error bars represent s.e.m. Figure
reproduced with permission from (Koshimori, Segura et al., 2015).
74
Figure 3-2. Partial correlations with age as a covariate between SFG thickness (extracted
from the significant cluster) and global composite z (left) and executive composite z (right)
in 26 PD patients showing the significant positive correlations. Figure reproduced with
permission from (Koshimori, Segura et al., 2015).
3.3.3 Subcortical Volume
We found no significant group differences between PD patients and HCs in any subcortical
ROIs and therefore, no correlation analysis was pursued.
3.3.4 White Matter
We found that PD patients showed WM changes with significantly higher MD values
compared to HCs (PD: 0.86 ± 0.01 and HC: 0.81 ± 0.006). The significant cluster with
increased MD was detected in widespread cortical WM regions including the frontal,
temporal, parietal and occipital regions as well as subcortical white matter regions (Figure 3-
3). These changes were located primarily in a larger area of bilateral frontal and temporal
regions and smaller areas of the left parietal and occipital regions. More specifically, WM
tracts with MD changes included forceps minor, cingulum, anterior thalamic radiation
(ATR), SCR, external capsule (EC), body of CC, uncinate fasciculus (UF), inferior fronto-
occipital fasciculus (IFO), SLF and inferior longitudinal fasciculus (ILF), and forceps major.
PD patients did not show significantly lower MD values or any differences in FA values
compared with HCs.
75
We extracted mean MD values from the significant cluster resulting from the group analysis
for each subject and investigated correlations between the MD values and (1) cognitive
measures that showed significant group differences (i.e., visual verbal test, JLO, global
composite and executive composite z scores) in both PD patients and HCs as well as clinical
measures (duration of disease and UPDRS-III scores) in PD patients. We first ran bivariate
correlations among all relevant covariates. In PD patients we found a significant negative
correlation between age and executive z scores (r = -0.622, P = 0.01) and a positive
correlation between age and the mean MD values in the significant cluster (r = 0.525, P =
0.037). In this group of patients, we also observed a significant correlation between gender
and cognitive measures (global composite z score (r = 0.667, P = 0.005) and executive
composite z score (r = 0.583, P = 0.018)). Thus, controlling for the effects of age and gender,
we found significant negative correlations between the MD values and global composite z
score (r = -0.39, P < 0.05) and executive composite z score (r = -0.44, P = 0.018) in PD
patients, suggesting that WM damage (mainly in frontal and temporal regions) was
associated with cognitive impairment. In HCs, we found no correlations among any of the
considered variables.
76
77
Figure 3-3. a. Clusters of significantly increased MD in 16 patients with PD compared
with 15 HCs in TBSS. Result images are overlaid on the MNI152 template. b. Bar graphs
on mean MD derived from the significant clusters in TBSS between HC and PD groups.
Error bars represent s.e.m. Reproduced with permission from (Koshimori, Segura et al.,
2015).
3.4 Discussion
The present study corroborated structural changes in PD patients and demonstrated specific
relationships between GM and WM damages and cognitive deficits in PD, suggesting that
these abnormalities may represent a sensitive biomarker for detecting brain changes
associated with cognitive changes.
Our results provided evidence of GM changes at the cortical level in PD. PD patients showed
significant cortical thinning in the left superior frontal, caudal middle frontal, precentral and
postcentral gyri compared to HCs. GM changes in these regions have previously been
reported in non-demented PD patients (Kostic et al., 2010; Melzer et al., 2012; Pereira et al.,
2012; Zarei et al., 2013). In addition, the superior frontal, caudal middle frontal gyrus (MFG)
and precentral sulcus were among the areas showing significantly greater progression of
cortical thinning in PD patients compared to HCs (Ibarretxe-Bilbao et al., 2012).
We further investigated whether these cortical abnormalities would explain the clinical
sequelae including motor and cognitive manifestations in PD and found that cortical thinning
in the dorsolateral SFG was associated with global and executive cognitive measures. The
dorsolateral SFG including BA8 and BA9 that showed significant thinning, is involved in a
variety of cognitive functions including WM (Levy & Goldman-Rakic, 2000; Owen et al.,
1998), attention (Corbetta, Patel, & Shulman, 2008), episodic memory (Desgranges, Baron,
& Eustache, 1998), and spatial cognition (Courtney, Petit, Maisog, Ungerleider, & Haxby,
1998; du Boisgueheneuc et al., 2006). In particular, the left SFG is involved in higher levels
of cognitive processing (du Boisgueheneuc et al., 2006) or executive functions. This may
well explain the significant correlation between the SFG thinning and global composite and
78
executive z scores. We also found a significant positive correlation between the left SFG
thickness and visual verbal test scores in HCs but not in PD patients. This is most likely due
to a lack of variability in the scores of the PD patients.
We did not find any correlation between the reduced CTh in the sensorimotor area and
UPDRS-III scores or duration of disease. These observations seem to be consistent with
previous studies where cortical thinning was found in motor areas including the left medial
SMA and right dorsal pre-SMA without any correlation with UPDRS-III scores or disease
duration in those regions (Jubault et al., 2011). On the other hand, correlations between
cortical GM measures in the sensorimotor area and motor symptoms (Lyoo et al., 2011;
Rosenberg-Katz et al., 2013) as well as between CTh in several cortical areas and duration of
disease (Lyoo et al., 2011; Rosenberg-Katz et al., 2013) have been reported in other studies.
However, it is unknown whether PD patients had cortical abnormalities in those areas
compared to HCs, as those studies did not include control data. The lack of a significant
correlation between cortical abnormality in the sensorimotor area and UPDRS-III scores in
our patients may well be due to the fact that UPDRS evaluations were performed during an
on-medication state (instead of during an off-medication state), and this could have very
likely diminished the possibility of detecting a significant relationship with the cortical
thinning in those regions.
We did not find any changes in subcortical volume between groups. Previous studies using
the same method also failed to find significant group differences (Tinaz et al., 2011; Zarei et
al., 2013). These consistent observations suggest that current imaging analysis may lack
sensitivity for detecting subtle subcortical GM changes.
Our TBSS analysis revealed that PD patients showed WM damage in multiple WM tracts in
widespread areas, more extensively in the bilateral frontal and temporal regions compared to
HCs. WM changes in these regions were consistently reported in previous TBSS findings
(Agosta et al., 2013; Agosta et al., 2014; Deng et al., 2013; Hattori et al., 2012; Matsui,
Nishinaka, Oda, Niikawa, Kubori et al., 2007; Rae et al., 2012). We detected group
differences with an increase in MD (but not in FA) in our patients. MD appears to be more
sensitive in detecting subtle WM changes than FA as suggested in other studies in early PD
79
(Melzer et al., 2013), early AD patients (Acosta-Cabronero, Williams, Pengas, & Nestor,
2010) and individuals with concussion (Cubon, Putukian, Boyer, & Dettwiler, 2011).
We further found that the fronto-temporal WM damage significantly correlated with the
global and executive cognitive measures. Our findings combined with previous studies
(Agosta et al., 2013; Gallagher et al., 2013; Hattori et al., 2012; Melzer et al., 2013) suggest
that frontal WM damage is a core pathological substrate of mild cognitive deficits in PD.
WM damage did not correlate UPDRS-III scores. This is very likely because UPDRS-III
evaluation was representative of the on-medication state as the same explanation provided
above.
Our GM and WM data consistently showed structural changes in the frontal region in PD.
The frontal cortex includes part of the fronto-striatal loops (Alexander, DeLong, & Strick,
1986), which have important implications for motor and NMS of PD. Prefronto-striatal
dysfunction is thought to underlie the basis for the most prominent executive impairment in
PD (Nagano-Saito et al., 2013; Owen, 2004; Pagonabarraga & Kulisevsky, 2012; Zgaljardic
et al., 2006). Abnormality in the SFG in particular can be an early indicator for further
decline of cognitive function. For example, PD patients who converted to dementia showed
cortical thinning in the frontal regions including the SFG, precentral gyrus, and anterior
cingulate at a baseline assessment and showed wider areas of cortical thinning in temporal,
parietal and occipital regions at a follow-up assessment (Compta et al., 2013). The WM
underlying the PFC contains projections from striatum via thalamus, to the striatum via
thalamus and directly to striatum. Although the TBSS does not allow for specific
identification of the fronto-striatal pathway, it reveals that the WM comprising the fronto-
striato-thalamic loop was affected in our PD patients. For example, the ATR includes WM
tracts from thalamus to PFC and vice versa. The lateral anterior ventral and medial dorsal
nuclei of thalamus, in particular, receive input from the BG. WM damage found in our study
extended beyond the frontal region and thus, it is still to be determined whether WM changes
in the prefrontal region alone can contribute to cognitive impairment. However, prefrontal
WM alone appears to be able to contribute to executive cognitive functions in PD (Gallagher
et al., 2013).
80
Our PD patients also showed extensive WM changes in bilateral temporal regions. A few
studies consistently showed that PD with MCI may be associated with GM atrophy in limited
regions of frontal and temporal regions (Hanganu et al., 2013; Melzer et al., 2012; Song et
al., 2011) while one of them additionally showed parietal volume loss (Melzer et al., 2012)
and others, occipital volume loss (Hanganu et al., 2013; Song et al., 2011). A longitudinal
study also demonstrated higher rate of cortical thinning in the frontal and temporal regions,
extending to parietal cortex (Ibarretxe-Bilbao et al., 2012). Although our PD patients did not
show GM abnormalities in the temporal region, the WM changes may have preceded GM
changes. The topography of WM changes in our study is similar to the previous GM findings
mentioned above with extensive frontal and temporal abnormalities as core changes and
unilateral focal parietal and occipital changes. Abnormality in the temporal region may be
associated with developing dementia in PD patients. For example, compared to the PD with
normal cognition, PD-MCI and PDD had significantly smaller hippocampal volumes, and
PDD additionally showed the medial temporal lobe atrophy (Weintraub et al., 2011).
Furthermore, PDD showed significant atrophy in the entorhinal cortex compared with PD
with normal cognition (Goldman et al., 2012).
Our findings in GM and WM changes in PD are in line with previous studies showing
widespread WM abnormalities but limited (Agosta et al., 2014) or absent (Agosta et al.,
2013; Hattori et al., 2012) cortical GM changes. However, differently from those studies
using VBM, the significant GM abnormalities reported here seem to suggest that CTA may
be a better approach. Thus, the combination of DTI and CTA appear to be sensitive
approaches for detecting subtle WM and GM changes associated with PD.
Although the MRI techniques used in the present study are validated methods to assess
structural changes, the biological underpinnings of these changes are not fully understood.
Subtle cortical thinning may reflect changes in size of cell bodies, dendritic arborisation,
and/or presynaptic terminals (Morrison & Hof, 1997; Pellicano et al., 2012). Changes in DTI
indices can result from a number of processes including neuronal loss and gliosis, as well as
disturbances in axonal membranes, myelin sheath, microtubules, and neurofilaments
(Shenton et al., 2012). PD has been associated with cytoskeletal damage of various neuronal
cells including dopaminergic, glutamatergic, cholinergic, tryptaminergic, GABAergic,
81
noradrenergic and adrenergic neurons (Braak et al., 1994; Braak et al., 1995; Braak, de Vos,
Jansen, Bratzke, & Braak, 1998; Foley & Riederer, 1999; Jellinger, 1991). The cytoskeletal
damage leads to Lewy pathologies including Lewy bodies and Lewy neuritis mostly located
in presynaptic terminals and in axons of affected nerve cells respectively (Braak,
Ghebremedhin, Rub, Bratzke, & Del Tredici, 2004). The major components of Lewy
pathologies include aggregations of misfolded α-synuclein (Braak et al., 2004) and
abnormally phosphorylated neurofilaments (Braak & Braak, 2000). Thus, MRI changes may
reflect the Lewy pathologies and/or neuronal degeneration secondary to the Lewy
pathologies.
The current study has a few potential limitations. First, PD patients included in the present
study had been taking parkinsonian medications for quite some time. To date, the effects of
chronic dopaminergic medication on brain structures remain to be determined. Second, our
PD patients underwent all study procedures in a on-medication state. It is well known that
dopaminergic medications in general can influence cognition (Kehagia, Barker, & Robbins,
2010). For example, while they can ameliorate certain cognitive deficits (e.g. executive
functions), dopaminergic medications can also worsen other cognitive abilities (Kehagia et
al., 2010; MacDonald et al., 2013; Ryterska, Jahanshahi, & Osman, 2013). Our decision not
to study them in an off-medication state was justified by the risk of worsening their motor
symptoms, increasing the risk of motion artifacts during MRI acquisitions.
In conclusion, the current study employed the CTA and TBSS to demonstrate that both GM
and WM are affected in PD patients and these anatomical changes may represent the neural
substrates underlying mild cognitive deficits in non-demented PD patients. Our findings
suggest that the dorsolateral SFG and fronto-temporal WM integrity play an important role in
global and executive cognitive performance, which are potential biomarkers for earlier
cognitive impairment in PD patients. In addition, distributed WM pathology (in contrast to
localized GM pathology in the frontal region) suggests that WM changes may precede
cortical GM changes and that WM changes be a potential biomarker for the early stage of
pathology. Taken together, the structural changes in the frontal region in particular may be an
early pathological substrates of cognitive impairment of PD, and may represent a sensitive
biomarker for brain changes in PD patients.
82
4.0 Study 2: Disrupted nodal and hub organization
account for brain network abnormalities in Parkinson’s
disease
Yuko Koshimori, Sang-Soo Cho, Marion Criaud, Leigh Christopher, Mark Jacobs, Christine
Ghadery, … Antonio P. Strafella.
4.1 Introduction
The clinical phenotypes of PD include cardinal motor symptoms such as tremor, rigidity,
bradykinesia, and loss of postural stability, along with a set of NMS such as depression, sleep
disturbances, autonomic dysfunction (Bonnet et al., 2012), and cognitive impairment
(Barnum & Tansey, 2012; Bonnet et al., 2012; Chaudhuri et al., 2011; Ferrer et al., 2011).
Cognitive impairment ranging from MCI in different cognitive domains (e.g., attention,
executive, visuospatial, and memory) to dementia is derived from dysfunction of different
neurotransmitter systems/brain networks (Gratwicke et al., 2015). The recent application of
graph theory to brain networks (Bullmore & Sporns, 2009) may shed light on complex
diseases such as PD.
The graph theoretical approaches have revealed that human brains show the following
topological characteristics (Bullmore & Sporns, 2009; Bullmore & Sporns, 2012). First,
individual brain regions termed ‘nodes’ typically have disproportionate connections with
other nodes. Highly connected nodes are considered as hubs, which play a crucial role in
integrating information. The anterior and posterior cingulate cortices, Ins, and superior
frontal and parietal cortex are identified as both structural and functional hubs (Sporns, 2014;
van den Heuvel & Sporns, 2013). These are heteromodal areas that are involved in a broad
range of cognitive processes (Achard, Salvador, Whitcher, Suckling, & Bullmore, 2006;
Mesulam, 1998). Second, some nodes are more closely connected to one another and form
modules/subnetworks, which promote specialized processing and functional segregation.
Large-scale modules correspond to well-established resting-state functional networks (Engel,
Gerloff, Hilgetag, & Nolte, 2013; Stam, 2014). Third, these modules are also interconnected
83
via hubs with long-distance pathways and thereby promote functional integration. Brain
networks display small-world properties of efficient global and local parallel information
transfer (i.e., integration and segregation) at relatively low connection cost (Achard &
Bullmore, 2007), which reflects an optimal organization through evolution and development
(Bullmore & Sporns, 2012; Gong et al., 2009). The coexistence and balance of segregation
and integration of subnetworks are fundamental for brain function (Bullmore & Sporns,
2012; Sporns, 2013).
There is evidence that the pathological course of neurodegenerative diseases, such as AD,
target network hubs (Seeley, Crawford, Zhou, Miller, & Greicius, 2009; Stam et al., 2009),
and selective damage to hubs can significantly change the integration of brain networks
(Bullmore & Sporns, 2012; Gong et al., 2009). With the dysfunction of multiple
neurotransmitters (Gratwicke et al., 2015; Pagonabarraga et al., 2015) and deposition of α-
synuclein (Braak et al., 2005), network hubs are almost certainly vulnerable in PD (Stam,
2014), and the selective deterioration of these network hubs may account for the distributed
abnormalities across the brain in PD (McColgan et al., 2015) with distinct clinical
correlations.
The premotor area and pre-supplementary motor area (SMA) have been identified as key
regions connecting multiple networks in healthy adults (Spreng et al., 2013). These regions
are important for self-initiated movements and preparation for actions (Passingham,
Bengtsson, & Lau, 2010). The SMA consisting of SMA proper and pre-SMA, is part of the
BC thalamo-cortical motor circuit associated with DA (Alexander et al., 1986; Krack, Hariz,
Baunez, Guridi, & Obeso, 2010; Obeso & Lanciego, 2011) and the changes in FC have been
well documented in PD patients, along with other cortical motor areas such as the primary
motor cortex (M1) and premotor area (Haslinger et al., 2001; Sabatini et al., 2000; Wu, Long
et al., 2009; Yu, Sternad, Corcos, & Vaillancourt, 2007).
DLPFC has been implicated as an anatomical and a functional node associated with
executive functions (Gong et al., 2009; Spreng et al., 2013; van den Heuvel & Sporns, 2013).
This brain region is part of FPN or executive control network (Dosenbach et al., 2007;
Seeley et al., 2007; Trujillo et al., 2015), as well as the fronto-striatal DA network
84
(Gratwicke et al., 2015). PD patients often show executive dysfunction due to a disruption of
the fronto-striatal DA system (Gratwicke et al., 2015). In addition, structural (Koshimori,
Segura et al., 2015) and functional (Li, Liang, Jia, & Li, 2016; Trujillo et al., 2015; Wu et al.,
2015) changes in this cortical area and their association with cognitive impairment have been
reported consistently in these patients (Koshimori, Segura et al., 2015).
Another important disease-associated hub in PD is the insular cortex, which connects and
integrates several functional systems: the anterior area is involved in cognitive and
behavioral/emotional functions, and the mid to posterior area involved in sensorimotor
functions (Kurth, Zilles, Fox, Laird, & Eickhoff, 2010). The insular cortex has extensive
connections with the striatum in a connectivity gradient from posterior to anterior
(Christopher, Koshimori, Lang, Criaud, & Strafella, 2014). In post-mortem PD brains, α-
synuclein depositions become evident throughout the Ins at Braak’s stage V (Braak et al.,
2005). In addition, the level of CSF α-synuclein concentration was significantly associated
with anterior insular network disruption (Madhyastha et al., 2015). The anterior insula (AIns)
is a core region of the SAL that guides behaviour (Menon & Uddin, 2010), and this brain
network can affect other cognitive subnetworks such as the DMN (Bonnelle et al., 2012;
Chiong et al., 2013). Graph theoretical analyses have revealed a decreased hub role in the left
dorsal AIns in PD patients compared to HCs (Tinaz et al., 2016). Furthermore, recent data
from our group have demonstrated that dopamine D2 receptor (D2 receptor) availability in
the AIns was associated with executive and memory dysfunction in PD-MCI (Christopher,
Duff-Canning et al., 2014; Christopher et al., 2015). Thus, the AIns, in particular, seems to
be a critical “hub” for cognitive impairment in parkinsonian patients (Christopher, Koshimori
et al., 2014; Criaud et al., 2016).
The present study aimed to investigate whether nodal connections and network hubs were
affected in PD patients, and whether these changes were associated with motor and NMS
employing graph theoretical approaches to rs-fMRI data. We hypothesized that (1) PD would
affect the nodal organization of regions such as the SMA, Ins, DLPFC, as well as the
striatum, and (2) the nodal and hub changes would account for some of the motor and
cognitive symptoms of PD.
85
4.2 Methods
4.2.1 Participants
Forty-five patients meeting the UK Brain Bank criteria for the diagnosis of idiopathic PD and
25 HCs participated in the study. During the screening session, all participants provided
written informed consent following a full explanation of the study procedures. Subsequently,
they were assessed for their global cognitive performance using the MoCA and their level of
depression using the BDI II. Additionally, patients were assessed for their motor severity of
the disease using the motor subset of the UPDRS (UPDRS-III) and for disease stage using
the Hoehn and Yahr. All the participants underwent a structural and rsfMRI scan. Patients
underwent the study procedures in an on-medication state. We reasoned that (1) scanning
patients in an on-medication state would minimize participants’ motion effects on rs-fMRI
data (Mowinckel, Espeseth, & Westlye, 2012; Tahmasian, Bettray, van Eimeren, Drzezga,
Timmermann, Eickhoff, et al., 2015) and (2) studying chronically medicated patients would
permit us to capture the complex neural substrates of cognitive impairment because cognitive
impairment in patients is mostly likely to be derived from dysfunction of multiple
neurotransmitters (Gratwicke et al., 2015). Exclusion criteria for the participants included:
(1) history of a head injury, psychiatric or neurological diseases (except PD for the patients),
(2) alcohol or drug dependency or abuse, (3) contraindications for MRI scanning, (4)
dyskinesia and dystonia for PD patients, and (5) for HCs, MoCA scores < 26 and BDI II
scores > 13. The study was approved by the CAMH Research Ethics Board.
4.2.2 MRI image acquisition
MR images were acquired with a GE Discovery MR 750 3T scanner with 8-channel head
coil. The protocol included whole-brain anatomic T1-weighted MRI images (Fast Spoiled
Gradient Echo pulse sequence; 200 sagittal slices; matrix of 256 x 256; TR: 6.7 ms; TE: 3.0
ms; slice thickness: 0.9 mm; FOV: 23 cm; Inversion Time: 650 ms; flip angle = 8°) and rs-
fMRI images (Gradient Echo/Fast Gradient Echo pulse sequence, 31 axial slices; matrix of
64 x 64; TR: 2000 ms; TE: 30 ms; flip angle = 60°, FOV: 22 cm; slice thickness: 5 mm;
duration: approximately 8min). During the rsfMRI scan, the participants were instructed to
86
keep their eyes open, let their mind wander, not to think about anything in particular, and not
to fall asleep. The first two volumes of the rsfMRI images were removed to allow for
magnetization equilibrium, resulting in the preprocessing of 240 volumes.
4.2.3 rsfMRI preprocessing
The functional data were preprocessed for each subject using the Conn FC toolbox version
15a (Whitfield-Gabrieli & Nieto-Castanon, 2012) including slice timing correction, motion
correction using a 6 degree rigid spatial transformation, which provided the spatial deviation
for each time point for translational (x, y, z) and rotational (roll, pitch, yaw) directions of
movement, normalization into standard MNI space using non-linear transformations, and
smoothing with a Gaussian smoothing kernel of 6 mm FWHM. Outliers in the global signal
of brain activation and movement were examined using the artifact detection toolbox. Time
points were considered as outliers if the global signal of brain activation exceeded three
standard deviations of the mean or if movement exceeded 0.5 mm across translational and
rotational directions of scan-to-scan deviation. Participants were excluded from the analyses
if outliers accounted for greater than 20 % of the entire dataset. Among the 70 participants,
five participants (two HCs and three patients) were excluded due to excessive head motion
and artifacts, resulting in 23 HCs and 42 patients included in the graph theoretical analysis.
Nuisance signals including six motion parameters as well as the signal from WM and CSF
voxels were regressed using aCompCor, the component–based noise correction method
(Behzadi, Restom, Liau, & Liu, 2007). The residual datasets were then temporally filtered
(0.008 < f < 0.09).
4.2.4 Network nodes
Among 160 regions of interest, of six different neural networks, derived from a series of five
meta-analyses of fMRI activation studies combined with cognitive control nodes identified
during error-processing (Dosenbach et al., 2007; 2008), we selected 120 regions of interest in
spheres with a radius of 5 mm in four networks (Figure. 4-1 A-D and Table 4-1) to
investigate the hub regions associated with motor and cognitive symptoms of PD. These
87
included 34 nodes of the DMN, 21 nodes of the FPN, 32 nodes of the cingulo-opercular
network (CON), and 33 nodes of the SMN.
Figure 4-1. Medial and lateral views of brain images with 120 regions of interest of four
subnetworks (Dosenbach et al., 2010). 34 nodes in red are affiliated with DMN (default
mode network); 21 nodes in blue, with FPN (fronto-parietal network); 32 nodes in green,
with CON (cinculo-opercular network); and 33 nodes in purple, with SMN (sensorimotor
network). Brain image and nodes are visualized using the BrainNet Viewer (NKLCNL,
Beijing Normal University).
DMN FPN
CON SMN
R R L L
88
Table 4-1 Node labels of four networks and their MNI coordinates
Network Node Coordinate#
X Y Z
Default mode
network
R ventromedial PFC 6 64 3
Medial PFC 0 51 32
L anterior PFC -25 51 27
R ventromedial PFC 9 51 16
L ventromedial PFC -6 50 -1
L ventromedial PFC -11 45 17
R ventromedial PFC 8 42 -5
R anterior cingulate cortex 9 39 20
R ventrolateral PFC 46 39 -15
R superior frontal cortex 23 33 47
L superior frontal cortex -16 29 54
R inferior temporal cortex 52 -15 -13
L inferior temporal cortex -59 -25 -15
R posterior cingulate cortex 1 -26 31
R fusiform gyrus 28 -37 -15
L precuneus -3 -38 45
L posterior cingulate cortex -8 -41 3
L inferior temporal cortex -61 -41 -2
L occipital -28 -42 -11
L posterior cingulate cortex -5 -43 25
R precuneus 9 -43 25
R precuneus 5 -50 33
L posterior cingulate cortex -5 -52 17
R posterior cingulate cortex 10 -55 17
L precuneus -6 -56 29
L posterior cingulate cortex -11 -58 17
R angular gyrus 51 -59 34
L angular gyrus -48 -63 35
R precuneus 11 -68 42
L intraparietal sulcus -36 -69 40
L occipital cortex -9 -72 41
R occipital cortex 45 -72 29
L occipital cortex -2 -75 32
L occipital cortex -42 -76 26
Fronto-parietal
network
R anterior PFC 29 57 18
L anterior PFC -29 57 10
R ventral anterior PFC 42 48 -3
L ventral anterior PFC -43 47 2
R ventrolateral PFC 39 42 16
R dorsolateral PFC 40 36 29
L anterior cingulate cortex -1 28 40
89
R dorsolateral PFC 46 28 31
L ventral PFC -52 28 17
L dorsolateral PFC -44 27 33
R dorsal frontal cortex 40 17 40
R dorsal frontal cortex 44 8 34
L dorsal frontal cortex -42 7 36
L inferior parietal lobule -41 -40 42
R inferior parietal lobule 54 -44 43
L posterior parietal cortex -35 -46 48
L inferior parietal lobule -48 -47 49
L inferior parietal lobule -53 -50 39
R inferior parietal lobule 44 -52 47
L intraparietal sulcus -32 -58 46
R intraparietal sulcus 32 -59 41
Cingulo-opercular
network
R anterior PFC 27 49 26
R ventral PFC 34 32 7
L anterior cingulate cortex -2 30 27
R ventral FC 51 23 8
R anterior insula 38 21 -1
R dorsal anterior cingulate cortex 9 20 34
L anterior insula -36 18 2
L basal ganglia -6 17 34
Medial frontal cortex 0 15 45
L ventral frontal cortex -46 10 14
L basal ganglia -20 6 7
R basal ganglia 14 6 7
L ventral frontal cortex -48 6 1
R mid insula 37 -2 -3
L thalamus -12 -3 13
L thalamus -12 -12 6
R thalamus 11 -12 6
R mid insula 32 -12 2
L mid insula -30 -14 1
R basal ganglia 11 -24 2
L posterior insula -30 -28 9
R temporal cortex 51 -30 5
L posterior cingulate -4 -31 -4
R fusiform gyrus 54 -31 -18
R precuneus 8 -40 50
R parietal cortex 58 -41 20
R temporal cortex 43 -43 8
L parietal cortex -55 -44 30
R superior temporal cortex 42 -46 21
L angular gyrus -41 -47 29
L temporal cortex -59 -47 11
90
L: left; PFC: prefrontal cortex; R: right; # the Montreal Neurological Institute coordinates
describe the center of node area.
4.2.5 Computation of network measures
Network measures were computed with age as a covariate using a Graph-Theoretical
Analysis Toolbox (GAT) (Hosseini, Hoeft, & Kesler, 2012). This toolbox constructs a binary
L temporoparietal junction -52 -63 15
Sensorimotor
network
R frontal cortex 58 11 14
R dorsal frontal cortex 60 8 34
L ventral frontal cortex -55 7 23
R pre-supplementary motor area 10 5 51
R ventral frontal cortex 43 1 12
Supplementary motor area 0 -1 52
R frontal cortex 53 -3 32
R precentral gyrus 58 -3 17
L MIns -42 -3 11
L precentral gyrus -44 -6 49
L parietal cortex -26 -8 54
R precentral gyrus 46 -8 24
L precentral gyrus -54 -9 23
R precentral gyrus 44 -11 38
L parietal cortex -47 -12 36
R MIns 33 -12 16
L MIns -36 -12 15
R temporal cortex 59 -13 8
L parietal cortex -38 -15 59
L parietal cortex -47 -18 50
R parietal cortex 46 -20 45
L parietal cortex -55 -22 38
L precentral gyrus -54 -22 22
L temporal cortex -54 -22 9
R parietal cortex 41 -23 55
R PostIns 42 -24 17
R parietal cortex 18 -27 62
L parietal cortex -38 -27 60
L parietal cortex -24 -30 64
L poterior parietal cortex -41 -31 48
L temporal cortex -41 -37 16
L temporal cortex -53 -37 13
R superior parietal cortex 34 -39 65
91
undirected graph G that has a network degree of E equal to the number of edges, and a
network density (cost) of D = E/[N x (N-1)]/2 representing the ratio of existing edges relative
to all possible edges (Bruno, Hosseini, & Kesler, 2012). It also identifies network fragment.
Network measures were computed over a range of connection densities. We examined those
graphs where all nodes were fully connected in the networks for both groups, which ensured
that network comparison between groups was meaningful with the same number of nodes in
each subnetwork, and where small-world properties were displayed for both groups. Network
fragment was observed at the density of 40%. Animal research has indicated that densities
above 50% are unlikely to be biological (Kaiser & Hilgetag, 2006). Therefore, the densities
ranging from 41% to 50% with an increment of 1% were investigated. Both HCs and patients
displayed small worldness properties where normalized characteristic path length is close to
one and normalized clustering coefficient is greater than one. Our interest was to detect
changes in nodal characteristics including hubness. Therefore, we investigated the graph
measures of local efficiency, as well as of nodal degree and BC (Bullmore & Sporns, 2009;
Rubinov & Sporns, 2010). Local efficiency measures the efficiency of local communication
of a given node, taking account for the number of the shortest paths between its neighbouring
nodes (Latora & Marchiori, 2001). It measures the ability of functional integration to
combine information across distributed brain regions. Nodal degree is defined as the total
number of connections that a node has with other nodes in the network, and BC is defined as
the fraction of all shortest paths in the network that pass through a given node. Bridging
nodes that connect disparate parts of the network often have high BC, making a potential hub
role of these nodes in the network (van den Heuvel & Hulshoff Pol, 2010) that interact with
many other regions and facilitate functional integration (Rubinov & Sporns, 2010).
4.2.6 Seed-based functional connectivity analysis
The only nodes showing group differences were further investigated using seed-based FC
analyses to elucidate changes in connectivity with other nodes in the network using the Conn
FC toolbox version 15a.
4.2.7 Statistical analysis
92
Statistical analyses for demographic measures were performed using SPSS Statistics 20.0.0
(Chicago, IL, http://www-01.ibm.com/software/analytics/spss/). Statistical significance
threshold was set at P ≤ 0.05. Independent student’s t-test (two-tailed) was used to compare
the means in head motion parameters and demographic measures between patients and HCs.
Area under the curve (AUC) analyses was performed to test group differences in graph
measures using a two-tailed non-parametric permutation test with 1000 repetitions over the
density range without network fragmentation using GAT. The AUC analyses are less
sensitive for thresholding processes as they yield a summarized scalar independent of single
threshold selection. Statistical significance threshold was set at P ≤ 0.005.
As a hub is characterized both by high connectivity and centrality (Sporns, Honey, & Kotter,
2007), the nodes whose degree and BC were one standard deviation above the mean network
degree and BC were considered hubs. Spearman’s correlation analyses were also performed
between the only network measures that yielded significant group differences, and different
disease measures (UPDRS III total scores, rigidity, bradykinesia, and resting tremor sub-
scores, LEDD, MoCA and BDI II scores) using SPSS Statistics 20.0.0 (Chicago, IL,
http://www-01.ibm.com/software/analytics/spss/). Statistical significance threshold was set at
P ≤ 0.05. General linear model was used to obtain between-subjects contrasts of seed-based
FC analysis. Statistical significance threshold was set at P ≤ 0.005.
4.3 Results
4.3.1 Demographic and clinical characteristics
Table 4-2 shows the demographic and clinical characteristics of the patients and HCs. There
was a significant difference in MoCA (t = 3.216, P = 0.002) and BDI II scores (t = -4.27, P <
0.001) between the two groups. PD patients showed lower global cognitive performance and
higher depression level than HCs. There was no significant difference regarding age (t =
0.59, P = 0.56) or education (t = 0.18, P = 0.86) between HCs and patients. The LEDD was
calculated by Levodopa (mg/day) + Controlled Levodopa (Levodopa x 0.75 mg/day) +
93
Entacapone/Comtan (Levodopa x 0.33 mg/day) + Praipexole/Mirapex (mg/day) +
Ropinirole/Requiq x 20 (mg/day) + Rasagilline/Azilect x 100 (mg/day) + Selegillne x 10
(mg) + Amantadine (mg/day) (Tomlinson et al., 2010).
Table 4-2 Demographic and clinical characteristics of patients with PD and HCs
Patients
(n = 42)
Healthy controls
(n = 23)
Age (years) 65.4 (6.5) 64.3 (8.3)
Sex (male/female) 31/11 10/13
Handedness (right/left) 34/8 23/0
Education (years) 16.0 (3.2) 16.1 (3.1)
MoCA 25.8 (2.9)* 27.5 (1.4)
BDI II 8.0 (4.6) ** 3.2 (3.4)
Disease duration (years) 5.9 (4.5) -
Symptom-dominant side (right/left) 26/16 -
UPDRS III
(on-medication)
27.5 (10.1) -
Rigidity 5.1 (2.2) -
Bradykinesia 1.9 (0.6) -
Resting tremor 1.3 (1.3) -
Hoehn and Yahr 2 (median) -
LEDD (mg/day) 644.4 (298.5) -
Data are presented in mean (standard deviation) unless otherwise indicated; * P = 0.002. **
P < 0.001.
94
4.3.2 Head motion parameters and outliers
The mean of realignment parameters for head motion and outliers of patients and HCs are
presented in Table 4-3. Realignment parameters for head motion including x (t = 1.70, P =
0.1), y (t = 0.69, P = 0.49), z (t = 0.31, P = 0.76), pitch (t = 1.87, P = 0.07), roll (t = 0.06, P =
0.95), and yaw (t = 1.43, P = 0.16) as well as the total number of outliers (t = 0.040, P =
0.97) were not significantly different between HCs and patients.
Table 4-3 Head motion parameters and outliers of patients with PD and HCs
Patients
(n = 42)
Healthy controls
(n = 23)
Head motion parameters
Liner motion parameters (mm)
X 0.013 (0.005) 0.011 (0.005)
Y 0.018 (0.119) 0.016 (0.009)
Z 0.058 (0.035) 0.061 (0.036)
Rotational motion parameters (radians)
Pitch 0.0006 (0.0003) 0.0005 (0.0002)
Roll 0.0002 (0.0001) 0.0002 (0.0001)
Yaw 0.0003 (0.0001) 0.00002 (0.0001)
Outliers (# of volume) 7.6 (7.4) 7.7 (7.9)
Data are presented in mean (standard deviation).
95
4.3.3 Group comparisons of graph measures
Compared to HCs, patients showed increased numbers of connections in the right and left
DLPFC of the FPN measured by nodal degree (P = 0.005 and P = 0.003, respectively)
(Figure 4-2; Table 4-4). In contrast, patients showed a reduction in intra-connectivity of the
right mid-insula (MIns) of the SMN measured by local efficiency (P = 0.005), as well as a
reduction in bridging role of the right pre-SMA of the SMN (P = 0.005) measured by nodal
BC.
We interrogated whether the graph measures of these four nodes would account for any of
the clinical measures (i.e., disease duration, UPDRS III scores as well as rigidity,
bradykinesia, and resting tremor subscores, Hoehn and Yahr, and LEDD, as well as MoCA
and BDI scores) by conducting Spearman’s correlation analyses. We found that BC of the
right pre-SMA showed a significant negative correlation with bradykinesia scores
(Spearman’s rho = -0.307, P = 0.048), suggesting that the diminished bridging role (i.e.
impaired integration) of the right pre-SMA may contribute to more severe bradykinesia.
We then applied the seed-based FC analyses to explore where these four affected areas
changed their connectivity. The results are summarized in Table 4-5. First, nodes in the right
DLPFC of the FPN showed increased intra-connectivity with a node in the homologous
anatomical region (t = 2.95, P = 0.002) and increased inter-connectivity with nodes of the
CON including the right MIns (t = 4.01, P < 0.001) and left PostIns (t= 3.3, P < 0.001).
Furthermore, the left DLPFC of the FPN showed increased intra-connectivity with a node in
the contralateral anatomical region (t = 2.68, P < 0.005). These increased regional intra- and
inter-connectivity may have occurred as a compensatory mechanism. Second, the right pre-
SMA of the SMN showed reduced inter-connectivity with nodes in the DMN including the
left angular gyrus (t = 3.29, P < 0.001) and left inferior temporal cortex (t = 2.81, P = 0.003),
suggesting its bridging and integrative role between cognitive and motor functions. The right
MIns did not show any significant changes in functional connectivity with other nodes.
96
Figure 4-2. Dorsal view of brain image presenting four nodes showing group differences.
Increased degree in the bilateral DLPFC of FPN (fronto-parietal network), as well as
decreased between centrality in the right pre-SMA and decreased local efficiency in the right
mid-insula of SMN (sensorimotor network) in PD patients compared with healthy controls (P
≤ 0.005, uncorrected). The nodes are presented in equal size for visualization purposes. Brain
image and nodes are visualized using the BrainNet Viewer (NKLCNL, Beijing Normal
University). Bar graphs show the averaged values of each graph measure for HC (healthy
controls) and PD (patients with Parkinson’s disease). Error bars represent standard error of
the mean.
97
Table 4-4 Graph measure changes in patients with Parkinson’s disease compared with HCs
# the MNI coordinates describe the center of node area.
Graph measure Node Subnetwork BA
Coordinate#
X Y Z
HCs < patients Degree Right DLPFC FPN 9 46 28 31
Left DLPFC FPN 9 -42 7 36
HCs > patients Local efficiency Right MIns SMN - 33 -12 16
BC Right pre-SMA SMN 6 10 5 51
98
Table 4-5. FC of three nodes that showed group differences.
Graph
measur
e
Seed
region
Subnetwork Coordinate# Target region Subnetwork Coordinate# T (63) P
(uncorrecte
d) X Y Z X Y Z
HCs <
Patients
Degree Right
DLPFC
(BA9)
FPN 46 28 3
1
Right MIns CON 32 -12 2 4.01 < 0.001
Left PostIns CON -30 -28 9 3.23 < 0.001
Right DLPFC
(BA9)
FPN 40 17 40 2.95 0.002
Left putamen CON -20 6 7 2.7 0.004
Left
DLPFC
(BA9)
FPN -42 7 3
6
Right DLPFC
(BA9)
FPN 40 17 40 2.68 <0.005
HCs >
Patients
BC Right pre-
SMA
(BA6)
SMN 10 5 5
1
Left angular
gyrus (BA40)
DMN -48 -63 35 3.29 < 0.001
Left inferior
temporal cortex
(BA20)
DMN -59 -25 -15 2.81 0.003
# the MNI coordinates describe the center of node area.
99
4.3.4 Hub analysis
Ten hub nodes, their affiliated networks, and MNI coordinates of both patients and HCs are
presented in Figure 4-3 and Table 4-5. The disease affected hub organization: Patients
presented eight regions acting as hubs while HCs presented nine regions. Seven of these
regions were shared by both patients and HCs, which included the right ACC and left
precuneus of the DMN, the left ACC of the FPN, as well as the left ACC, left thalamus, and
two nodes in the right thalamus of the CON. The patients lost hub properties in the left AIns
of the CON and the left MIns of the SMN, while the CN of the CON acted as a new hub in
the patients. We explored whether any of these three hub regions were associated with the
clinical measures (i.e., disease duration, UPDRS III scores as well as rigidity, bradykinesia,
and resting tremor scores, Hoehn and Yahr, and LEDD, as well as MoCA and BDI scores) in
PD patients. We found a significant positive correlation between hubness (i.e., the mean of
degree and BC) in the left MIns of the SMN and dopaminergic medication (Spearman’s rho
= 0.303, P = 0.05), suggesting that dopaminergic medication may influence hubness in the
MIns in these patients.
100
Figure 4-3 Caudal view of brain image presenting 10 hub nodes. ACC: anterior cingulate cortex; AIns: anterior insula; CN: caudate
nucleus; CON: cingulo-opercular network; DMN: default mode network; FPN: fronto-parietal network; MIns: mid insula. Seven hubs
in red represent common hubs to both healthy controls and patients with Parkinson’s disease. Two in yellow represent hubs identified
only in healthy controls and one in green, only in patients. Brain image and hubs are visualized using the BrainNet Viewer (NKLCNL,
Beijing Normal University).
ACC (FPN)
ACC (CON)
Precuneus (DMN) ACC (DMN)
Thalamus
(CON)
CN (CON) AIns (CON)
L R
MIns (SMN)
Hub – Common
Hub – HCs
Hub – Patients with PD
101
Table 4-6. Hub regions in HCs and patients with PD.
# The MNI coordinates describe the center of node area.
Hub region Subnetwork Coordinate#
X Y Z
Common Right ACC DMN 9 39 20
Left precuneus DMN -3 -38 45
Left ACC FPN -1 28 40
Left ACC CON -2 30 27
Left thalamus CON -12 -12 6
Right thalamus CON 11 -12 6
Right thalamus CON 11 -24 2
HC only Left AIns CON -36 18 2
Left MIns SMN -36 -12 15
Patient only Right CN CON 14 6 7
102
4.4 Discussion
In the current study, we aimed to investigate functional changes in the SMN and cognitive
networks in PD with a particular focus on changes in the disease-associated nodes and hubs
using graph theoretical analyses. In general, patients showed weakened connectivity in nodes
of the SMN and enhanced connectivity in nodes associated with cognitive networks. Patients
showed functional changes in the right pre-SMA and right MIns, as well as the bilateral
DLPFC (BA9) compared with HCs. PD also affected hub functions and organization. In
particular, the left AIns of the CON and left MIns of the SMN lost their hub properties and a
node in the CN was identified as a new hub in these patients.
4.4.1 Changes in the sensorimotor network
As hypothesized, we found functional changes in the nodes of the SMN. More specifically,
patients showed a reduction in intra-connectivity of the right MIns and inter-connectivity of
the right pre-SMA with the nodes of the DMN. Furthermore, the diminished inter-
connectivity of the right pre-SMA was associated with more severe bradykinesia. In addition,
the left MIns lost its hub properties in the patients and showed a positive interaction with
dopaminergic medication.
Reduced FC in the SMA including the pre-SMA was consistently demonstrated in PD
patients compared to HCs using the graph theoretical approaches. (Nagano-Saito et al., 2014;
Sang, et al., 2015; Skidmore et al., 2011; Tinaz et al., 2016; Wei et al., 2014; Wu et al.,
2009). For example, nodal degree, efficiency, and BC were all reduced in the SMA in PD
patients (Sang, et al., 2015). Reduced connectivity of the SMA was also associated with
more severe motor symptoms (Cao et al., 2011; Wu, et al., 2009).
The pre-SMA is the anterior part of the SMA and is separated from the SMA proper by the
vertical anterior commissural line (Picard & Strick, 2001). These two areas are different in
terms of anatomical connections and functions. The SMA proper is connected to the M1,
projected to the spinal cord and is involved in movement generation, while the pre-SMA has
extensive connections with the PFC (Wang, Shima, Sawamura, & Tanji, 2001) and is
103
involved in cognition (Picard & Strick, 2001). For example, the pre-SMA has been
implicated to play an important role in cognitive motor control involving integration of
sensory information and decision-making about actions or motor selection (Ikeda et al.,
1999).
It is not surprising that the decreased interconnectivity of the pre-SMA with a cognitive
subnetwork can account for the slowness of movement, as it is postulated that bradykinesia
could arise from slowness in formulating the instructions to move before the onset of and
during the actions (Berardelli, Rothwell, Thompson, & Hallett, 2001). In the present study,
the weakened connectivity between the right pre-SMA and the nodes of the DMN, in the
context of DA deficiency, may affect the redirection of attentional processes from self-
reflection (i.e., internal reference) to goal-directed behavior, thus contributing to
bradykinesia.
4.4.2 Changes in the cognitive networks
We also found functional changes in the DLPFC (BA9) associated with the FPN and the left
AIns of the CON in patients compared with HCs. Patients showed an increased level of
connectivity in bilateral DLPFC (BA9). The subsequent FC analysis revealed that the
enhanced connectivity occurred within and between cognitive networks (i.e., FPN and
CON). The hub analysis revealed that while the left AIns lost its hub properties, a new hub
region was identified in the striatal region (i.e. CN) associated with the same network (i.e.
CON) in PD patients.
Literature showed mixed results in FC associated with the fronto-parietal network and the
DLPFC in PD patients compared with HC subjects using graph theoretical analysis. In line
with our findings, increased degree of DLPFC (Cao et al., 2011) and increased FC in the
fronto-parietal module (Baggio et al., 2014) were reported in PD patients. On the other hand,
reduced FC in the DLPFC (Nagano-Saito et al., 2014) and in the fronto-parietal regions
(Fogelson et al., 2013) was also reported. The discrepancy in these findings may be in part
due to capturing different mechanisms in response to the disease process in this anatomical
region (i.e., compensatory or pathological mechanism).
104
Increased FC of task-related regions (i.e. DLPFC) is often interpreted as a compensatory
mechanism (Grady, 2012) when cognitive/behavioural performance starts failing. Our
patients had lower MoCA scores compared with HCs, suggesting that the enhanced
connectivity and the new hub of the CN may be parallel “attempted compensation” (Grady,
2012) for trying to maintain an efficient utilization of neural resources (Grady, 2008), in the
context of DA deficiency. In fact, AIns of CON lost its hub properties in our PD patients.
The insular cortex is associated with dopaminergic dysfunction PD (Christopher, Duff-
Canning et al., 2014; Christopher et al., 2015) and deposition of aggregated α-synuclein
(Braak et al., 2005). Dopamine dysfunction in the AIns in particular was associated with
poor executive and memory impairment in PD patients (Christopher, Duff-Canning et al.,
2014; Christopher et al., 2015). The right DLPFC enhanced FC with MIns and PostIns of the
CON. Thus, enhanced FC in the DLPFC and CN might be counteracting the reduced or
reducing hubness of the AIns. It may also be explained by pathological cortical disinhibition
(de Haan, Mott, van Straaten, Shceltens, Stam, 2012). Another possible explanation is that
the brain changes may have resulted from the effects of dopaminergic medication although
there was no association between the changes and LEDD.
4.4.3 Hub reorganization
Hub nodes are characterized with high connectivity and centrality (Sporns et al., 2007).
Therefore, we used both BC and degree to identify hub regions. Our HCs showed nine hub
regions: two DMN nodes including ACC and precuneus, one FPN node - ACC, five CON
nodes in the regions of ACC, thalamus, and AIns, and one SMN node – MIns. As we
hypothesized, PD patients showed hub reorganization characterized by the loss of hub
properties in the insular regions and an emergence of a new hub in the CN. Consistent with
the literature, our findings further support that (1) the insular cortex is an important
integrative node in brain network communication (Menon & Uddin, 2010), and (2) it is
affected by PD, possibly due to aggregated alpha synuclein (Braak et al., 2005) and
dopamine dysfunction (Christopher, Duff-Canning et al., 2014; Christopher et al., 2015).
To date, there were two studies investigated hub regions in PD patients and HCs using
rsfMRI data (Baggio et al., 2014; Sang et al., 2015). One study identified 18 nodes as hub
regions that showed the 20% highest scoring of the sum of clustering coefficient and BC
105
(Baggio et al., 2014). Consistent with our data, the precuneus was one of the hub regions in
the DMN in both HCs and PD patients. This study also identified the medial SFG node as a
hub in HCs and PD-MCI patients; SFG in HCs and PD patients; and middle SFC in non-MCI
PD patients. The other study used BC and tested the values against the null hypothesis using
a non-parametric one-tail test across different thresholds and found 13 hubs in both PD
patients and HCs (Sang et al., 2015). There was no consistency in hub regions in HCs
between this study and our study while this study also identified the CN as a hub in PD
patients.
In addition to the functional hubs, two studies investigated structural hubs using cortical
thickness and subcortical volume (Pereira et al., 2015) and using DTI tractography (Nigro et
al., 2016). The first study employed the same classification to identify hub regions as our
study (i.e, one standard deviation higher than the mean of both degree and BC) (Pereira et al.,
2015). It identified 19 hubs in HCs, 13 hubs in non-MCI PD patients, and 16 hubs in PD-
MCI patients. The hubs identified across the three groups were located in distributed cortical
regions, primarily in the frontal, temporal and parietal regions. The nodes in the DLPFC
region were also identified as hubs in both HCs and non-MCI PD patients. Consistent with
our data, the precuneus was identified as a hub region in both HC and two PD groups.
Interestingly, only PD-MCI patients showed the insula region as a hub region. The second
study used four different graph measures to score each node from 0 to 4. If the node showed
the highest 20 % of one of these four graph measures, it gets one score and if the node had a
score of 2 or higher, it was identified as a hub. PD patients and HCs shared 14 hubs, which
were located in the cortical regions except the occipital region. The precuneus, DLPFC
region and insula were identified as hub regions in both groups, and the paracentral lobule
was identified as a hub region only in HCs.
Some of these studies identified the DLPFC and SMA regions as hubs in HCs, whose
function was altered in PD patients, which we hypothesized, but we did not find in our data.
The discrepancy in hub regions between our study and the previous studies largely arises
from the definition of hub regions and types of nodes (e.g., functional nodes derived from
tasks used in our study or anatomically parceled nodes used in all other studies) investigated.
Other possible factors contributing to the differences include types of MRI data (e.g.,
106
structural or function), the number of nodes investigated, as well as processing and analysis
methods.
4.4.4 Conclusions
Using graph theoretical approach, we further characterized PD-related functional changes
and highlight the diffuse changes in the nodal organization along with the regional hub
disruption, which accounts for the distributed abnormalities across brain networks and
clinical manifestations of PD. PD patient showed both increased and decreased FC, which
occurred within the same network and across different networks. The decreased FC found in
the pre-SMA of the SMN is most likely pathological changes evidenced by its association
with severe motor symptom. Increased FC and a new hub in cognitive networks such as FPN
and CON may be adaptive and counteracting the influence of neurotransmitter dysfunction
and deposition of α-synuclein associated with cognitive function. Lastly, the insula plays an
important integrative role in brain network communication, susceptible to PD pathology.
107
5.0 Study 3: Imaging striatal microglial activation in
patients with Parkinson’s disease
This study was published in PLoSOne.
This chapter has been modified from the following:
Yuko Koshimori, Ji-Hyun Ko, Romina Mizrahi, Pablo Rusjan, Rostom Mabrouk, Mark F.
Jacobs, … Antonio P. Strafella (2015). Imaging Striatal Microglial Activation in Patients
with Parkinson's Disease. PLoSOne, 10(9), e0138721.
5.1 Introduction
Neuroinflammation is considered to play an important role in the progression of Parkinson’s
PD. Neuroinflammation in PD was first evidenced in a postmortem study where activated
microglia were found in the SN (McGeer et al., 1988). A subsequent postmortem study
further identified activated microglia in the extended brain areas such as putamen,
hippocampus, as well as trans-entorhinal, cingulate, and temporal cortices (Imamura et al.,
2003). Neuroinflammatory processes were also confirmed by increased concentration of
inflammatory cytokines such TNF-α and IL-1β and IL-6 in the striatum at postmortem (Mogi
et al., 1994) as well as in vivo studies using the serum (Dobbs et al., 1999) and cerebrospinal
fluid (Blum-Degen et al., 1995) of PD patients.
TSPO has been studied as a potential in vivo biomarker of reactive gliosis and inflammation
associated with a variety of neuropathological conditions (Chen & Guilarte, 2008). TSPO is
located in the outer mitochondrial membrane of glial cells. While TSPO levels are very low
in healthy brains, they markedly increase co-localizing activated microglia in brains affected
by various diseases such as amyotrophic lateral sclerosis, AD, frontotemporal dementia and
MS (Venneti et al., 2013). This elevated TSPO expression was primarily quantified using
[11C]-PK11195 PET, the first and most commonly used TSPO radioligand.
108
To date, a few studies have investigated neuroinflammation in PD patients using [11C]-
PK11195 PET. However, the findings have been inconclusive. While some studies have
found elevated TSPO binding in the nigro-striatal regions (Ouchi et al., 2005), others did not
support these observations (Kobylecki et al., 2013). Additionally, an association between
elevated TSPO binding in the nigro-striatal regions and clinical measures has been
demonstrated in some studies (Ouchi et al., 2005) but not in others (Kobylecki et al., 2013).
One possible explanation for these inconsistent findings may be due to the limitations
inherent to this prototypical radioligand ([11C]-PK11195) such as low signal-to-noise ratio,
high non-specific binding, low brain penetration, and high plasma protein binding. These
limitations as well as its limited dissemination for wider clinical use due to the short-lived
carbon-11 labeling have prompted the development of new TSPO radioligands.
[18F]-FEPPA is a new, second-generation TSPO radioligand that has shown a high affinity
for TSPO, an appropriate metabolic profile, high brain penetration and good
pharmacokinetics (Wilson et al., 2008). Its quantification has been validated using the 2TCM
with VT as reliable outcome measure (Rusjan et al., 2011). [18F]-FEPPA demonstrated
superior specificity to the affected side of the striatum and stronger correlation with
inflammatory cytokines compared with [11C]-PK11195 using an animal model of PD
(Hatano et al., 2010). Further, our group recently demonstrated a three-fold higher [18F]-
FEPPA VT in a tumor site compared to the healthy contralateral area in an individual subject
(Ko et al., 2013), as well as a significant increase in individuals with MDE (Setiawan et al.,
2015) and patients with AD (Suridjan et al., 2015).
Second-generation TSPO radioligands have been known to present three patterns of binding
affinity based on genetic polymorphism: HABs, MABs, and LABs (Owen et al., 2012).
These different phenotype patterns can be predicted by a single-nucleotide polymorphism
(SNP), rs6971 located in the exon 4 of the TSPO gene resulting in a nonconservative amino-
acid substitution at position 147 from alanine to threonine (Ala147Thr) in the fifth
transmembrane domain of the TSPO protein. This polymorphism accounts for some of the
large inter-individual variability in the outcome measures (Owen et al., 2012). Therefore,
accounting for this polymorphism in the statistical analyses will likely increase sensitivity in
detecting neuropathological changes (Mizrahi et al., 2012).
109
In the present study, we investigated for the first time whether TSPO imaging with [18F]-
FEPPA, could be used as a potential biomarker for neuroinflammation in PD. Using the
striatum as our main ROI, we sought to determine whether (1) the polymorphisms would
reflect on TSPO binding affinity also in PD patients; (2) [18F]-FEPPA VT could differentiate
the PD group from the HC group; and (3) the levels of TSPO binding would correlate with
clinical measures of PD.
5.2 Methods
5.2.1 Subjects
Nineteen patients meeting UK Brain Bank criteria for the diagnosis of idiopathic PD and 17
HC subjects participated in the study. Exclusion criteria for all participants included (1)
history of a head injury, psychiatric or neurological (except PD for the patients) diseases, (2)
alcohol or drug dependency or abuse, (3) contraindications for MRI scanning, and (4) use of
nonsteroidal anti-inflammatory drugs (NSAID). PD patients were assessed for their cognitive
ability using the MoCA, the level of depression using the BDI II, as well as for motor
severity of the disease using the motor subset of the UPDRS-III. All participants underwent
PET and structural MRI scans. All participants provided written informed consent following
full explanation of the study procedures. The study was approved by the CAMH Research
Ethics Board and the University Health Network Research Ethics Board, and conformed to
the Declaration of Helsinki.
5.2.2 PET data acquisition
The synthesis of [18F]-FEPPA has been described in detail elsewhere (Wilson et al., 2008). It
can be reliably and quickly labeled with [18F] by nucleophilic displacement of a tosylate,
leaving group in a fast one-step reaction yielding a sterile and pyrogen-free product after
purification and formulation.
110
The PET images were obtained using a 3D High Resolution Research Tomograph (HRRT)
(CPS/Siemens, Knoxville, TN, USA), which measures radioactivity in 207 slices with an
inter-slice distance of 1.22 mm. A custom-fitted thermoplastic mask was made for each
subject and used with a head fixation system during the PET scans to minimize head
movement. Following a transmission scan, intravenous [18F]-FEPPA was administered as a
bolus. The mean injected amount, specific activity at the time of injection and mass injected
of all the participants were 4.926 ± 0.265 (mCi), 3127.723 ± 2026.481 (mCi/μmol), and 0.9 ±
0.676 (μg), respectively. The scan duration was 125 min. The images were reconstructed into
34 time frames: 1 frame of variable length until the radioactivity appears in the FOV, 5
frames of 30s, 1 frame of 45s, 2 frames of 60s, 1 frame of 90s, 1 frame of 120s, 1 frame of
210s, and 22 frames of 300s.
All PET images were corrected for attenuation using a single photon point source, 137Cs
(T50 = 30.2 years, Eγ = 662 keV) and were reconstructed by filtered back projection
algorithm using a HANN filter at Nyquist cutoff frequency. The reconstructed image has 256
× 256 × 207 cubic voxels measuring 1.22 × 1.22 × 1.22 mm and the resulting reconstructed
resolution is close to isotropic 4.4 mm, FWHM in plane and 4.5 mm FWHM axially,
averaged over measurements from the center of the transaxial FOV to 10 cm off-center in 1.0
cm increments. In addition, for frame realignment for head motion correction, each image
was reconstructed without attenuation correction using three iterations of iterative
reconstruction (Rusjan et al., 2011).
5.2.3 MRI acquisition
MRI images for all the subjects were acquired for co-registration with the corresponding
PET images and the anatomical delineation of the ROIs (i.e., CN and putamen). Proton
density weighted MR images were chosen for better identification of the striatum (Rusjan et
al., 2006). Two dimensional (2D) oblique proton density-weighted MR images were acquired
with a GE Discovery 3.0 T MRI scanner (slice thickness = 2 mm, TR = 6000 ms, TE = Min
Full, flip angle = 90°, number of excitations (NEX) = 2, acquisition matrix = 256 × 192, and
FOV = 22 cm).
111
5.2.4 Input function measurement
Dispersion- and metabolite-corrected plasma input function was generated as previously
described (Rusjan et al., 2011). Briefly, arterial blood was taken continuously at a rate 2.5
mL/min for the first 22.5 minutes after radioligand injection and the blood radioactivity
levels were measured using an automatic blood sampling system (Model # PBS-101 from
Veenstra Instruments, Joure, The Netherlands). In addition, 4 to 8 ml manual arterial blood
samples were obtained at 2.5, 7, 12, 15, 30, 45, 60, 90, and 120 min relative to time of
injection (Rusjan et al., 2011). A bi-exponential function was used to fit the blood-to-plasma
ratios. A hill function was used to fit the percentage of unmetabolized radioligand. The
dispersion effect was modeled as to the convolution with a mono-exponential with dispersion
coefficient of 16 seconds and corrected with iterative deconvolution (Rusjan et al., 2014).
5.2.5 Generation of ROI-based time activity curve
[18F]-FEPPA PET images were preprocessed and ROIs were automatically generated using
in-house software, ROMI (Rusjan et al., 2011). Briefly, ROMI fits a standard template of
ROIs to an individual PD-weighted MR image based on the probability of GM, WM, and
CSF. The individual MR images are then co-registered to each summed [18F]-FEPPA PET
image using the normalized mutual information algorithm so that individual refined ROI
template can be transferred to the PET image space to generate the time activity curve (TAC)
for each ROI. Our a priori ROIs included the CN and putamen, which are disease-affected
regions and whose quantification was validated (Rusjan et al., 2011). Dynamical series of
images of [18F]-FEPPA PET were visually checked for head motion and corrected using
frame-by-frame realignment. Low noise, nonattenuation-corrected images (created with
iterative reconstruction) were used to optimize the frame-by-frame realignment process. A
normalized mutual information algorithm was applied with SPM8 (Wellcome Trust Centre
for Neuroimaging, London, UK) to co-register each frame to the frame that showed a high
signal-to-noise ratio. Parameters from the normalized mutual information were applied to the
corresponding attenuation-corrected dynamic images to generate a movement-corrected
dynamic image.
112
To address the potential issues of bias from the volume loss in the older subjects, time
activity data for all subjects was corrected for the effect of partial volume error (PVE) using
the Mueller-Gartner partial volume error correction (PVEC) algorithm as implemented in
Bencherif, Stumpf, Links, & Frost (2004).
5.2.6 Kinetic analysis
VT values in each ROI were derived from a 2TCM using [18F]-FEPPA radioactivity in
arterial plasma as an input function and a 5% vascular contribution (Rusjan et al., 2011). VT
is a ratio at equilibrium of the radioligand concentration in tissue to that in plasma (i.e.
specific binding and non-displaceable uptake including non-specifically bound and free
radioligand in tissue) and can be expressed in terms of kinetic rate parameters as: VT = K1 /
k2 (1 + k3 / k4) where K1 and k2 are influx and efflux rates for radiotracer passage across the
blood brain barrier and k3 and k4 describe the radioligand transfer between the free and non-
specific compartments and the specific binding compartment. We also measured %COV
(100% x standard error/mean), where standard error was estimated from the diagonal of the
covariance matrix of nonlinear least-squares fitting (Rusjan et al., 2011). From the different
ROIs, we included VT with %COV of ≤ 20, which assured less data noise.
5.2.7 DNA extraction and polymorphism genotyping
Genomic DNA was obtained from peripheral leukocytes using high salt extraction methods
(Lahiri & Nurnberger, 1991). The polymorphism rs6971 was genotyped variously using a
TaqMan® assay on demand C_2512465_20 (AppliedBiosystems, CA, USA). The allele
T147 was linked to Vic and the allele A147 was linked FAM. Polymerase chain reactions
(PCR) were performed in a 96-well microtiter-plate on a GeneAmp PCR System 9700
(Applied Biosystems, CA, USA). After PCR amplification, end point plate read and allele
calling was performed using an ABI 7900 HT (Applied Biosystems, CA, USA) and the
corresponding SDS software (v2.2.2). Individuals with genotype Ala147/Ala147 were
classified as HABs, Ala147/Thr147 as MABs, and Thr147/Thr147 as LABs (Owen et al.,
2012).
113
5.2.8 Statistical analysis
Demographic and clinical measures were compared using factorial ANOVA, independent,
two-tailed student t tests, or Fisher’s exact tests. Group differences in VT values were
analyzed using factorial ANOVA with TSPO genotype and disease as fixed factors in the CN
and the putamen. A second level of analysis with student t tests were performed among four
groups: (1) between HC with MAB (HC-MAB) and HC with HAB (HC-HAB), (2) between
PD with MAB (PD-MAB) and PD with HAB (PD-HAB), (3) between HC-MAB and PD-
MAB, and (4) between HC-HAB and PD-HAB. Correlations were investigated between
clinical measures and VT values in the ROIs using Pearson’s correlation tests. All of the
statistical analyses were performed using SPSS Statistics version 20.0. The threshold for
significance was set at P < 0.05.
5.3 Results
5.3.1 Demographic and clinical characteristics
Based on the rs6971 polymorphism, there were 16 MABs consisting of 8 HC subjects and 8
PD patients, 16 HABs consisting of 8 HC subjects and 8 PD patients, as well as 4 LABs
consisting of 3 PD patients and 1 HC subject. [18F]-FEPPA signals are too low to be
quantifiable in these LABs and for this reason these subjects were not included in the
analyses. In general, LAB subjects in a Caucasian sample represent less than 5% (Mizrahi et
al., 2012), and only very few are identified in any given study. The demographic and clinical
characteristics of HC subjects and PD patients included in the analyses were presented in
Table 5-1 and 5-2.
The factorial ANOVA showed that there were no significant differences in age (F(3, 28) =
0.86, P = 0.47), as well as in the [18F]-FEPPA injected amount (F (3, 28) = 1.30, P = 0.30),
specific activity at the time of injection (F (3, 28) = 0.94, P = 0.44) or mass injected (F (3,
28) = 1.68, P = 0.19). Chi-square analysis showed that there was no significant difference in
the composition of gender (χ2 = 3.14, P = 0.08). In addition, the PD-MAB and PD-HAB
114
groups were not significantly different in years of education (t(14) = 0.50, P = 0.63), MoCA
scores (t(14) = 0.82, P = 0.43), BDI scores (t(14)= 0.65, P = 0.53), UPDS III (t(14) = 0.88, P
= 0.40), duration of disease (t(14) = 0.18, P = 0.87), total LEDD (t(14) = 0.03, P = 0.98) or
the composition of symptom dominant side (Fisher’s exact tests, P = 0.61).
Table 5-1. Demographic and clinical characteristics and PET measures of HC subjects and
PD patients
HC
(n = 16)
PD
(n = 16)
Age 62.1 (8.7) 64.3 (9.0)
Gender (M:F) 6:10 11:5
Genotype (HAB:MAB) 8:8 8:8
Injected amount (mCi) 4.85 (0.26) 5.00 (0.26)
Specific activity
at the time of injection (mCi/µmol)
3006.91 (2435.27) 3248.54 (1589.13)
mass injected (µg) 1.05 (0.83) 0.75 (0.45)
Data was presented in mean (standard deviation).
Table 5-2. Demographic and clinical characteristics of PD patients
PD-MAB
(n = 8)
PD-HAB
(n = 8)
Education (years) 15.5 (3.8) 16.4 (3.2)
MoCA 27.4 (2.3) 26.5 (1.9)
BDI 6.4 (4.1) 7.6 (4.3)
Symptom dominant side (R:L) 6:2 4:4
UPDRS III 28.4 (9.6) 25.0 (5.1)
Duration of disease (years) 5.6 (2.1) 5.9 (4.7)
Total LEDDa (mg/day) 623.8 (596.4) 525.1 (267.6)
Data was presented in mean (standard deviation). aLEDD: Levodopa dose + 1-Levodopa-CR
x 0.75 + pramipexole (mg) x 67 (Evans et al., 2004).
5.3.2 Genotype and disease effects on TSPO binding
115
Levene’s tests of equality of error variances were not significant in either CN (P = 0.449) or
putamen (P = 0.284). There was a significant main effect of genotype on VT values in the CN
(F(1, 28) = 14.62, P = 0.001) and in the putamen (F(1, 28) = 27.15, P < 0.001). There was no
main effect of disease in the CN (F(1, 28) = 0.37, P = 0.55) or in the putamen (F(1, 28) =
1.27, P = 0.269). There was also no disease x genotype interaction in the CN (F(1, 28) =
1.38, P = 0.25) or in the putamen (F(1, 28) = 0.86, P = 0.362).
Using two-tailed student t tests, when comparing the effect of genotype in HC groups, HC-
HAB group showed significantly higher VT values in in the putamen (10.05 ± 0.97; t(14) =
3.04, P = 0.009), but not in the CN (9.29 ± 0.94; t(14) = 1.86, P = 0.084) compared with HC-
MAB group (putamen: 6.31 ± 0.76; CN: 6.95 ± 0.91).
The effect of genotype in PD groups showed that PD-HAB had significantly higher VT
values in both putamen (11.84 ± 1.12; t(14) = 4.30, P = 0.001) and CN (11.02 ± 1.14; t(14) =
3.56, P = 0.003) and compared with PD-MAB (putamen: 6.49 ± 0.52; CN: 6.43 ± 0.61;)
(Figure 5-1A).
When testing the effect of the disease, neither PD-MAB (putamen: 6.49 ± 0.52; CN: 6.43 ±
0.61) nor PD-HAB (putamen: 11.84 ± 1.12; CN: 11.02 ± 1.14) group showed significant
increase in VT values in any ROI compared with HC-MAB (putamen: 6.31 ± 0.76; CN: 6.95
± 0.91) and HC-HAB (putamen: 10.05 ± 0.97; CN: 9.39 ± 0.94) groups, respectively (Figure
5-1B). The percentage differences of the mean VT values between PD-HAB and HC-HAB
groups were 16% in both putamen and CN, while they were 3% and -8%, respectively in the
MAB groups.
There was no correlation between LEDD, UPDRS scores or duration of disease with the VT
values in the striatum in either PD-HAB, PD-MAB, or PD groups, thus excluding any role of
these variables in the VT findings.
We conducted as well an additional analysis on the data without PVEC to investigate
whether this played any role in the results. The analysis confirmed the main effect of
genotype on VT values (Figure 5-2A), but no main effect of disease (Figure 5-2B), or disease
x genotype interaction.
116
Figure 5.1. Graphs of PVEC VT in the CN and in the putamen. A. HC-MAB and HC-HAB
groups as well as PD-MAB and PD-HAB groups. Asterisks indicate that the HAB groups
show significantly higher VT mean values compared with the MAB groups (** P < 0.01). B.
HC-MAB and PD-MAB groups as well as HC-HAB and PD-HAB groups. In the HAB
group, the percentage difference between PD and HC was 16% in both CN and putamen, and
in the MAB group, it was -8% and 3%, respectively. Reproduced with permission from
(Koshimori, Ko et al., 2015).
Figure 5.2. Graphs of VT in the CN and in the putamen. A. HC-MAB and HC-HAB groups
as well as PD-MAB and PD-HAB groups. Asterisks indicate that the HAB groups show
significantly higher VT mean values compared with the MAB groups (**P < 0.01). B. HC-
MAB and PD-MAB groups as well as HC-HAB and PD-HAB groups. In the HAB group, the
percentage difference between PD and HC was 11% in the CN and 14% in the putamen, and
in the MAB group, it was -3% and 5%, respectively. Reproduced with permission from
117
(Koshimori, Ko et al., 2015)
5.4 Discussion
This is the first study to investigate the potential use of [18F]-FEPPA as a new radioligand to
measure neuroinflammation in PD patients. As already observed in normal controls, we
found that in PD patients as well, the rs6971 polymorphism influenced TSPO binding
affinity. Although, we did not find a significant disease effect on TSPO expression in the
striatum, an interesting observation was a trend towards elevated TSPO binding in the PD-
HAB group (16% in both CN and putamen), but not in the PD-MAB group. These
observations imply that while the genotype of the rs6791 polymorphism certainly plays a
role in TSPO expression also in PD, a larger sample size may be needed to investigate the
interactions between rs6791polymorphism and neuroinflammation in PD patients, although
other studies with similar sample size did show a significant increase in [18F]-FEPPA VT in
individuals with MDE (Setiawan et al., 2015) and patients with AD (Suridjan et al., 2015).
In those studies, such differences in TSPO expression were reported mainly in the HAB
groups. In particular in AD patients (Suridjan et al., 2015), elevated microglia was detected
throughout the brain, in both GM and WM suggesting that neuroinflammation may certainly
play a role in the cognitive decline.
The same dissociation and potential role of the rs6971 polymorphism on TSPO binding has
also been suggested in studies using other second-generation TSPO radioligands. For
example using [3H]PBR28, postmortem brains of patients with SCZ classified as HABs
showed significantly greater specific binding in the DLPFC while those categorized as
MABs did not (Kreisl et al., 2013). Similarly using [18F]-FEMPA in AD, the disease effect
on TSPO expression was stronger in different cerebral regions in HAB patients (Varrone et
al., 2015). On the other hand, cocaine abusers did not seem to demonstrate the same
dissociation (Narendran et al., 2014).
The rs6791 polymorphism can be associated with the susceptibility, progression or disease
protection (Rone et al., 2012). For instance, healthy individuals with Ala147/Ala147 (i.e.,
HABs) have been associated with significantly higher level of low-density lipoprotein (LDL)
118
than those with Ala147/Thr147 (i.e., MABs) and with Thr147/Thr147 (i.e., LABs) (Costa et
al., 2009), and LDL has been associated with the risk of PD (Huang et al., 2015). While these
reports are highly suggestive, the clinical significance of this polymorphism in PD is yet to
be determined. Similarly, despite the well-documented increase in the TSPO expression as a
result of brain insult in ex-vivo brain tissue, the functional significance of the upregulated
TSPO is still unclear (Chen & Guilarte, 2008). One possibility is that this upregulation may
be related to glial proliferation, migration, and phagocytosis (Streit, Sammons, Kuhns, &
Sparks, 2004) or secretion of inflammatory cytokines. [11C]-PK11195 reduced the expression
of proinflammatory cytokines in cultured human microglia (Choi et al., 2002). Furthermore,
increased TSPO levels in microglia and astrocytes may possibly increase neurosteroid
synthesis at injury sites to promote neurotropic and neuroprotective activity (Schumacher et
al., 2000). While the observations regarding the potential role of the polymorphism on TSPO
binding may be quite intriguing, they are speculative and warrant further investigations.
We investigated microglial activation only in the CN and putamen as a first evaluation study
of a second generation TSPO radioligand, [18F]-FEPPA, and did not find a significant disease
effect on [18F]-FEPPA VT in the striatum. We selected these two regions because they are
most affected by dopaminergic dysfunction in PD and [18F]-FEPPA VT is quantifiable while
the SN is also severely affected in PD by DA neurons undergoing degeneration, and is most
studied and evidenced for microglial activation in the postmortem PD brains, but it is hard to
accurately quantify the VT value of the SN due to its low and unstable identifiability.
Postmortem and neuroimaging studies demonstrated evidence of elevated neuroiflammatory
processes in the striatal region although they yielded mixed results. In one study, the tissues
in the putamen from 12 PD brains showed significantly higher number of ramified microglia
compared with those from 4 HC brains (Imamura et al., 2003). On the other hand, two
postmortem studies showed no activated microglia in the putamen, comparing five PD brains
with five HC brains (Mira et al., 2000) and comparing brains of nine PD, six incidental LBD,
four progressive supranuclear palsy, and 10 HCs (Bradaric et al., 2012) although these two
studies found significantly higher level of microglial activation in the SN in PD brains. To
our knowledge, the CN has not been investigated for microglial activation in postmortem PD
brains. However, there is evidence for significantly increased reactive astrocytes in the
119
region in postmortem PD brains (Mythri et al., 2011). The expression of GFAP in the CN of
five PD brains was approximately two fold higher than that of six HC brains. In addition,
GFAP immunoreactive astrocytes were reported in the striatum of 12 PD brains although the
immunoactivity varied from slight in seven cases to moderate in five cases while two cases
showed no detectable immunoactivity (Braak et al., 2007).
Similarly, some of the neuroimaging studies using [11C]-PK11195 found microglial
activation in the putamen (Iannacone et al., 2012; Ouchi et al., 2009) and striatum (Gerhard
et al., 2006) while others did not in the striatum (Bartel et al., 2010; Edison et al., 2013;
Kobylecki et al., 2013).
We did not observe any relationships between clinical measures such as antiparkinsonian
medication, disease severity and duration of disease and [18F]-FEPPA VT in the striatum,
which is consistent with previous neuroimaging studies (Gerhard et al., 2006; Kobylecki et
al., 2013). These findings thus imply that these variables are not associated with TSPO
expression in PD patients.
Our findings did not reveal a significant disease effect on [18F]-FEPPA VT in the striatum.
Postmortem and neuroimaging studies showed evidence of neuroinflammatory processes in
PD while the degree of neuroinflammation might vary. Further investigations warrant to
determine whether [18F]-FEPPA could be used as a biomarker of neuroinflammation in PD
by investigating a large sample size and to understand the interactions between the
rs6791polymorphism, neuroinflammation, and clinical subtypes as well as disease prognosis.
120
6.0 General Discussion
6.1 Summary of Findings
The studies included in this thesis used multimodal imaging tools and techniques to
investigate structural, functional, and molecular changes and their associations with clinical
manifestations in PD. The specific aims of each study were (1) to elucidate the relationships
between structural abnormalities and clinical manifestations of PD, (2) to investigate changes
in brain network properties with focus on the disease-associated nodes and hubs using the
most advanced graph theoretical approach, and (3) to assess TSPO polymorphism and
neuroinflammation using a second generation TSPO radioligand, [18F]-FEPPA in PD.
Overall, the main findings were:
1. Compared to HC subjects, PD patients showed structural changes including cortical
thinning in the left SFG (BA8 and BA9) and in the left M1 extending anteriorly to the
MFG and posteriorly to the S1, as well as diffuse WM abnormalities in bilateral
frontal and temporal regions, in the left parietal and occipital regions as well as
subcortical regions in a subgroup of the patients.
2. The cortical thinning in the left SFG and the WM abnormalities were associated with
poor performance on the overall cognition as well as executive function measured by
neuropsychological tests in PD patients. The significant cortical thinning in the left
SFG solely accounted for the associations with cognitive function, suggesting that
this region play an important role in cognition and be a potential biomarker for earlier
cognitive impairment in PD.
3. Diffuse WM changes occurred in the subgroup of PD patients, suggesting that WM
changes precede cortical GM changes and that WM changes measured by TBSS be a
biomarker for the early stage of disease.
4. Within the SMN, PD patients showed weakened connectivity in key nodes including
the right pre-SMA and right MIns compared to HC subjects. Additionally, the left
MIns lost its hub properties in PD patients.
5. The reduced bridging role of the right pre-SMA was associated with more severe
121
bradykinesia, which specifically impaired network integration with DMN, suggesting
that the nodal change may affect the redirection of attentional processes from self-
reflection (i.e., internal reference) to goal-directed behavior. The finding also
suggests that pre-SMA be an important node possibly for both motor and cognitive
symptoms.
6. The hubness in the left MIns was positively correlated with dopaminergic medication,
suggesting the role of dopaminergic medication in the connectivity of the node.
7. Within the executive cognitive subnetworks, the left AIns lost its hub properties,
while a new hub region was identified in the CN in PD patients paralleled by
increased levels of inter- and intra-connectivity in the DLPFC compared to HC
subjects.
8. The increased connectivity in the cognitive subnetworks might reflect “attempted
compensatory” or dedifferentiation in our PD patients.
9. The insula plays an important integrative role in brain network communication,
which is susceptible to PD pathology.
10. Using [18F]-FEPPA, PD-HABs showed significantly higher VT values than PD-MABs
in the putamen and CN. However, there was no evidence of elevated levels of
neuroinflammation in the striatal regions in PD patients compared to HC subjects or
no correlations between TSPO density and clinical measures in PD patients.
6.2 Structural and functional changes in sensorimotor nodes in
Parkinson’s disease
In study 1, we demonstrated cortical thinning in the M1 in our patients. The cortical thinning
and faster rate of thinning in the M1 in PD has been reported in several studies (Compta et
al., 2013; Huang, Lou, Xuan, Gu, Guan, Xu, et al., 2016; Ibarretxe-Bilbao et al., 2012; Kim
et al., 2014; Pereira et al., 2012; Segura et al., 2014; Uribe et al., 2016). Similar to the
findings of four of these studies, we found cortical thinning in the face area of M1. This
atrophy may reflect impairment in speech activity as the ventral sensorimotor area is
122
associated with speech performance (Yeo et al., 2011). In addition, over 90% of PD patients
develop speech disorder (i.e. hypophonia, slurred speech), and PD patients had significantly
reduced brain activation during a speech task compared to HC subjects (Pinto et al., 2011).
In study 2, we found functional changes in the MIns and pre-SMA in the SMN in our
patients. (Note that in this section, “mid-posterior Ins” is used to discuss the findings of
different studies because different studies used the MIns or the posterior insula (PostIns) for
the similar MNI coordinates.) Insular cortex is an important area to PD. First, the Ins is
highly interconnected with the striatum in a connectivity gradient from posterior to anterior,
with PostIns projecting to the dorsal/posterior striatum, and AIns progressively towards
anterior and ventral regions of the striatum. This organization is highly consistent with the
functional roles of both dorsal/posterior Ins and striatum in sensorimotor processes, and
anterior/ventral regions of these structures in cognitive and affective processing (Chikama,
McFarland, Amaral, & Haber, 1997). Second, α-synuclein aggregates were found throughout
insular cortex starting at Braak stage 5. Third, PD-MCI patients showed reduced D2 receptor
availability in the bilateral mid-posterior Ins compared with HC subjects (Christopher, Duff-
Canning et al., 2014; Christopher et al., 2015). Fourth, the meta-analysis study confirmed
one of the important insular subregions associated with PD was the dorsal mid-posterior Ins,
which was associated with an off-medication state, suggesting the vulnerability of this
subregion without dopaminergic medication (Criaud et al., 2016). Lastly, the meta-analysis
of VBM studies in PD reported volume loss in the Ins (Pan, et al., 2012).
Study 2 further corroborated the insular involvement in PD. First, the right mid-posterior Ins
showed reduced FC within the SMN compared to HC subjects. Second, the left mid-posterior
Ins lost hub function and its reduced hub role was associated with lower dose of
dopaminergic medication. The mid-posterior Ins is implicated in the processing of position,
movement, and sensation of the body (Cerasa et al., 2006; Chang, Yarkoni, Khaw, & Sanfey,
2013). Thus, our findings of abnormalities in the sensorimotor Ins may be associated with
postural instability, which is one of the cardinal motor symptoms but typically appears at
more advanced stage.
123
Furthermore, we demonstrated diminished bridging role of the right pre-SMA with the DMN
nodes in PD patients compared to HC subjects. SMA or pre-SMA has been implicated as a
hub in a few studies in healthy adults (de Pasquale et al., 2013; de Pasquale, Della Penna,
Sporns, Romani, & Corbetta, 2015; Spreng et al., 2013; Yeo et al., 2011). SMA was
identified as a hub region along with other hubs in the DMN and dorsal attention network
(DAN) in healthy adults (de Pasquale et al., 2015). As hubs tend to connect with one another,
the authors speculated that these hubs allow for the integration of internal cognitive
processes such as memory and self-referential behavior medicated by the DMN, the selection
of environmental and body information mediated by the DAN, and the motor planning and
execution mediated by the SMN. Thus, it is possible that pre-SMA plays an important role in
funneling the internal cognitive processes and passing on to the SMA proper for self-initiated
movements.
Importantly, the diminished bridging role of the right pre-SMA measured by BC was
associated with more severe bradykinesia in the patients. Bradykinesia could arise from
slowness in formulating the instructions to move before the onset of and during the actions
(Berardelli et al., 2001), suggesting the cognitive involvement in the generation of the
symptom. The pre-SMA is involved in motor planning based on internally generated thought
(Chung, Han, Jeong, & Jack, 2005). Thus, compromised FC between the pre-SMA and
cognitive nodes might lead to slowness to move and slow movement. The DMN nodes
(BA40 and BA20) that had reduced connectivity with the pre-SMA in our study are
functionally connected to the posterior putamen and the posterior CN in PD patients and HC
subjects (Helmich et al., 2010). Thus, these DMN nodes are also involved in sensorimotor
function. In fact, imagery of movements activates pre-SMA as well as the inferior parietal
cortex (Macuga & Frey, 2011).
6.3 Structural and functional changes in cognitive nodes in
Parkinson’s disease
We found structural and functional changes in the DLPFC in addition to functional changes
124
in the AIns in our PD patients. DLPFC is a core region of FPN or executive control network
(Dosenbach et al., 2007; Seeley et al., 2007; Trujillo et al., 2015), as well as the fronto-
striatal DA network (Gratwicke et al., 2015), and has been also implicated as an anatomical
and a functional hub associated with executive functions in healthy brains (Gong et al., 2009;
Spreng et al., 2013; van den Heuvel & Sporns, 2013). The most common cognitive
impairment of PD patients is executive dysfunction due to a disruption of the fronto-striatal
DA system as well as the FPN possibly due to cholinergic and/or noradrenergic dysfunction
(Gratwicke et al., 2015).
Importantly, we demonstrated structural changes in the left SFG (BA8 and BA9) (Koshimori,
Segura et al., 2015) and functional changes in bilateral DLPFC (BA9) compared to HC
subjects in study 1 and study 2, respectively. In study 1, we also demonstrated that cortical
thinning in the left SFG was associated with poor executive and overall cognitive
performance on a neuropsychological test battery (Koshimori, Segura et al., 2015).
Supporting the important role of this region in cognition, the SFG is one of the regions
shown cortical thinning at the baseline assessment in PD patients who were later converted to
PDD (Compta et al., 2013). Similarly, multiple protein pathologies (tau, beta-amyloid, α-
synuclein) in BA9 along with the superior and middle temporal gyrus were associated with
the development of dementia in PDD and DLB (Howlett et al., 2015). The lateral SFG is
thought to be involved in executive processing such as monitoring and manipulation (Postle,
Stern, Rosen, & Corkin, 2000). However, the left SFC in particular is functionally connected
not only to a number of distributed FPN regions but also to most of DMN regions, making
this area a key region for internally-focused, goal-oriented cognition (Spreng et al., 2013).
Thus, abnormalities in this region could have substantial effects on cognitive function in PD
patients.
In addition, we demonstrated increased FC in DLPFC (BA9 in MFG) of the FPN within and
between cognitive subnetworks compared to HC subjects. Increased FC in the fronto-parietal
module was also reported in PD-MCI patients using graph theoretical analysis (Baggio et al.,
2014) and during cognitive tasks in PD patients (Gerrits et al., 2015; Madhyastha et al.,
2015; Trujillo et al., 2015). In aging or neuropathologic literature, hyper-activations are
interpreted as compensation when older adults or patients perform at the same level as
125
younger adults or HC subjects (Cabeza, Anderson, Locantore, & McIntosh, 2002; Clement,
Gauthier, & Belleville, 2013; de Vries et al., 2014), or when increased brain activity is
positively correlated with cognitive performance in older adults or patients, but not in
younger adults or HC subjects (Davis, Dennis, Daselaar, Fleck, & Cabeza, 2008; Della-
Maggiore et al., 2000; Madden et al., 1999). However, when hyper-activations are not related
to clinical or cognitive measures, this refers to as “attempted compensation” (Grady, 2012),
which applied to our findings. Our findings can be also interpreted as pathological responses
called dedifferentiation defined by inefficient use of neural resources or reduction in
specific/selective recruitment of brain regions (Fornito, Zalesky, & Breakspear, 2015; Grady,
2008). It may be possible that at early disease compensation occurs, but as the disease
progresses, brain changes such as dedifferentiation appear before a full-blown dysfunction
develops (Gratwicke et al., 2015).
One possible pathological mechanism for higher FC is due to GABA deficiency resulting in
glutamatergic disinhibition and excessive Glu release (Krystal et al., 2003; Lewis, Pierri,
Volk, Melchitzky, & Woo, 1999) or the impaired N-methyl-D-aspartate (NMDA)
conductance/Glu neurotransmission onto GABAergic interneurons and thereby induces
cortical disinhibition (Anticevic et al., 2012; Krystal et al., 2003). Glu overactivity in the BG
is well documented in PD (DeLong & Wichmann, 2015). Postmortem PD brains showed
impairment of GABAergic neurotransmission in BA9 (Lanoue, Dumitriu, Myers, &
Soghomonian, 2010). On the other hand, significantly reduced vesicular Glu transporters
were found in the BA9 and anterior part of the superior temporal gyrus of postmortem PD
brains (Kashani, Betancur, Giros, Hirsch, & El Mestikawy, 2007). To date, there were two
studies using proton magnetic resonance spectroscopy (1HMRS) at high magnetic field
strengths demonstrated reduced Glu (reductions in Glu/Cr) in the posterior cingulate cortex
in PD patients (Griffith, Okonkwo, O'Brien, & Hollander, 2008) and in PDD patients
(Griffith et al., 2008).
We also found that the left dorsal AIns lost its hub role, which was consistent with another
study using graph theoretical analysis in PD patients (Tinaz et al., 2016). The hub role of the
dorsal AIns is also supported from a meta-analysis study showing that it is the overlap region
for multimodal functions including cognition, interoception, empathy, emotion, pain,
126
olfaction, and gustation (Kurth et al., 2010). Furthermore, the AIns is known as a core region
of the SAL, which is involved in switching between cognitive subnetworks (i.e., central
executive network and DMN) (Seeley et al., 2007). In PD-MCI patients, reduced D2 receptor
availability in the right dorsal AIns was associated with executive and memory dysfunction
(Christopher, Duff-Canning et al., 2014; Christopher et al., 2015). The meta-analysis study
also showed that the bilateral AIns was associated with cognitive function in PD patients
(Criaud et al., 2016).
Furthermore, in the CON, the right CN emerged as a new hub while the left dorsal AIns lost
its hub role in our patients. This may be explained by the “hub failure” at the chronic stage of
disease proposed by Stam (2014). This model postulates that in the acute stage of brain
disease, hubs take up connections from nodes that redirect from the affected nodes. If this
abnormal redirection is severe and sustained, hubs themselves may deteriorate and diminish
their role of handling information traffic. In the chronic stage of disease, the affected hubs
reroute their traffic to nodes at lower levels of hierarchy, resulting in the decreased centrality
of the original hubs and local emergence of new connections, which may also be an
emergence of new hubs.
6.4 Dopaminergic modulation on resting-state functional
connectivity
Abnormalities in resting state FC have been found in PD patients both in ON and OFF states
compared with HC subjects (Tahmasian et al., 2015 for review). However, there are
significant effects of dopamine replacement therapy (DRT) on FC at rest, which was
demonstrated by contrasting ON and OFF states using various FC methods. In general,
administration of DRT normalizes abnormalities in FC in PD patients. For example, there is
one study that investigated functional changes in a motor network using graph theoretical
analysis in both ON and OFF states compared with HC subjects (Wu, et al., 2009). This
study included 18 nodes associated with a motor network and measured their nodal degree
(i.e., the number of connections each node has). There were six nodes showing significant
127
group differences after correcting multiple comparisons between PD patients in an OFF state
and HC subjects including the SMA, cerebellum, M1, parietal cortex, DLPFC, and putamen
(P ≤ 0.003) and group differences in the globus pallidum and premotor cortex to lesser extent
(0.003 < P < 0.05). After the administration of DRT, the five nodes except the left parietal
cortex were normalized. However, PD patients in an ON state still showed some decrease in
nodal degree in the SMA (P = 0.02), putamen (P = 0.016), and cerebellum (P = 0.042)
compared with HC subjects. The nodal degree in the SMA, putamen, and thalamus showed
negative correlations with UPDRS III scores while the nodal degree in the M1, cerebellum,
premotor cortex, and parietal cortex showed positive correlations with UPDRS III scores.
Whether or not these correlations persist in an ON state is unknown, as they were
investigated only in an OFF state.
Dopaminergic effects are also known to affect cognitive performance primarily in executive
function and modulate brain activation/deactivation in PD patients (Poletti and Bonuccelli,
2013 for review). However, the effects of dopaminergic medication on cognitive networks at
rest were little studied in PD patients. In one study, FC at rest as well as the spectral
compositions of rsfMRI oscillations were investigated before and after the administration of
DRT in drug-naïve PD patients and compared with those in a placebo group of drug-naïve
PD patients as well as HC subjects (Esposito et al., 2013). The levodopa-induced spectral
changes in low-frequency band were observed in visual network and bilateral FPN.
However, cognitive performance associated with the changes is unknown.
There were more studies on dopaminergic modulation on resting state FC in healthy subjects.
One study found liner effects of dopaminergic modulation (L-dopa > placebo > haloperidol)
on RSNs including between bilateral midbrain and DMN, between dorsal caudate and right
FPN, and between the ventro-medial thalamus and left FPN, as well as between the ventral
striatum and salience/executive network (Cole, Oei, Soeter, Both, van Gerven, Rombouts, et
al., 2013). However, differential effects of dopamine were also reported depending on nodes
within networks. For example, haloperidol group showed greater FC in the left precentral
and middle frontal gyri while it showed reduced FC in supramarginal gyrus/intraparietal
sulcus within the DMN (Cole, Beckmann, Oei, Both, Gerven, and Rombouts, 2013),
suggesting the complex effects of dopamine on cognitive networks.
128
Dopamine modulation on FC and cognitive performance may be more complicated in PD
patients than healthy subjects. It can depend on differential effects of different medication
such as levodopa, dopamine agonists, or different dopamine agonists when PD patients are in
an ON state, acute and chronic effects of the medication (Poletti and Bonucelli, 2013), and
the interactions of these factors with COMT Val58Met polymorphism of individual PD
patients.
Most studies including the current studies enrolled PD patients who already have been
medicated. The chronic effects of dopaminergic medication on cognitive status in PD
patients are largely unknown due to paucity of literature (Poletti and Bonucelli, 2013).
However, there were three randomized longitudinal studies (Kulisevsky et al., 2000;
Rektorová et al., 2005; Relja and Klepac, 2006). They consistently concluded while
dopaminergic medication significantly improved motor symptoms, it had differential or no
effects on cognitive performance, and was not sufficient to compensate their cognitive
impairment and decline.
Taken together, our findings of reduced FC in the SMN nodes including the pre-SMA and
MIns likely reflected the abnormalities that were not completely restored by dopaminergic
medication in our PD patients with moderate severity. Similarly, our patients showed
significant cognitive impairment even in an ON state and thus dopaminergic medication had
not compensated for the cognitive impairment, which is consistent with the findings of the
three studies above. We found increased FC in the DLPFC nodes in an ON state. As FC in
the DLPFC is generally affected by dopaminergic medication, it cannot totally exclude the
medication effects on these brain changes even though we did not find the correlation
between the changes and LEDD.
6.5 White matter changes in Parkinson’s disease
We found more extensive WM changes in the bilateral frontal and temporal areas and limited
changes in the left parietal and occipital areas. The altered WM tracts and ROIs in the frontal
region included (1) the forceps minor connecting the lateral and medial parts of the frontal
129
lobes including the anterior CC, (2) cingulum, (3) ATR, and (4) SCR containing motor WM
tracts, and body of CC connecting motor areas. The affected association fibers included (1)
the UF containing cholinergic fibers (Selden, Gitelman, Salamon-Murayama, Parrish, &
Mesulam, 1998) and connecting the orbital cortex and the anterior temporal lobe, (2) SLF in
the frontal, parietal, and temporal regions, (3) ILF in the temporal and occipital region, and
(4) IFO in the frontal, temporal and occipital regions. The Forceps major excluding the
splenium of the CC was affected in the occipital region. Subcortically, MD changes were
also detected in the anterior limb of internal capsule (ALIC) containing ATR, which lies
between the head of the CN medially and the putamen laterally as well as in the EC adjacent
to the putamen and Ins, which containing association fibers such as SLF and IFO,
commissural fibers as well as ascending cholinergic fibers (Selden et al., 1998).
Combined with the findings of previous studies, frontal WM tracts are vulnerable in PD
(Agosta et al., 2013; Agosta et al., 2014; Deng et al., 2013; Hattori et al., 2012; Matsui,
Nishinaka, Oda, Niikawa, Komatsu et al., 2007; Melzer et al., 2013; Rae et al., 2012). We
also found bilateral temporal WM changes. In postmortem PD brains, the reduction of
serotonergic axonal density was most severe in the temporal and prefrontal regions compared
to normal brains (Azmitia & Nixon, 2008). In addition, the axons in these regions displayed
enlarged vesicles, splayed ending, clustering and degenerating profile. We also found that
WM abnormalities were associated with executive and global cognitive dysfunction. This is
not surprising as association fibers connecting different cortical regions were disrupted in our
patients.
A majority of studies investigated WM changes using both FA and MD. Our study
demonstrated WM changes measured by MD. Only MD changes but not FA changes were
detected in PD patients with normal cognition (Melzer et al., 2013). Other studies also
showed changes in MD, but not in FA (Agosta et al., 2013; Kim et al., 2013). In addition,
MD showed a larger area of WM changes than FA (Gallagher et al., 2013; Theilmann et al.,
2013). Furthermore, MD values were more often correlated with neuropsychological test
scores in different domains than FA values (Melzer et al., 2013; Zheng et al., 2014). FA and
MD are thought to reflect different tissue characteristics where FA is more sensitive to tissue
directionality and organization (architecture) and MD to tissue density (Farina et al., 2000;
130
Wiltshire et al., 2010). Depending on WM region, disease stage, and the presence of
cognitive impairment, either MD or FA can be more sensitive detecting WM changes
(Agosta et al., 2014; Rae et al., 2012; Theilmann et al., 2013; Zhan et al., 2012). Thus
investigating both DTI indices are most likely to capture changes in WM microstructure.
6.6 Neuroinflammation in Parkinson’s disease
6.6.1 Interindividual variability in total distribution volume
It is now known that it is important to account for TSPO binding affinity classes based on the
rs6971 polymorphism in the statistical analysis to reduce interindividual variability in VT
values. Yet, within the same genetic group, there is still a large variability in VT values,
resulting in a considerable overlap in VT values between HABs and MABs. The previous
study reported that most MABs expressed their two binding sites in approximately equal
proportions with the mean fraction of high affinity sites was 58% ± 6.6% in vitro and 63% ±
9.8% in platelets (Owen et al., 2011). However, the fraction of high-affinity PBR28 binding
sites within MABs ranges from 38% to 83% in vitro. Thus, the authors discussed the
possibility of a continuum in the expression of the HAB and LAB sites, which may explain
some of the variability VT values within the same genetic group.
Age may be another factor accounting for the variability of VT values. As discussed in
section 1.3.4.3, there are age-related changes in microglia and astrocytes. So far, the effect of
age on TSPO outcome measures in healthy subjects has yielded inconsistent results using
second-generation TSPO radioligands. There was no significant age effect using [11C]-
DAA1106 (Yasuno et al., 2008), [18F]-FEPPA (Suridjan et al., 2014) or using [11C]-DPA713
(Yokokura et al., 2016). On the other hand, the age effect was detected in the whole brain
using [11C]-vinpocentine (Gulyas et al., 2011), in most GM brain regions using [11C]-
PBR111 (Guo et al., 2013), as well as frontal and temporal lobes using [11C]-PBR28
(Bloomfield et al., 2016). To avoid the potential age effect on TSPO outcome measures,
study groups should be age-matched and age can also be included in the analysis as a
covariate.
131
Another possible factor may be the variability of plasma free fraction (fp) (Bloomfield et al.,
2016). Although previous [18F]-FEPPA studies in AD and MDE demonstrated significant
group differences without accounting for the effect of the fp (Setiawan et al., 2015; Suridjan
et al., 2015), future studies are needed to determine whether the inclusion of fp would
improve the sensitivity detecting neuroinflammation in clinical populations (Suridjan et al.,
2015).
The reproducibility of the outcome measures may play a role in generating the variability.
One study tested the reproducibility of VT values in six HABs including three HC subjects
and three HIV patients. They underwent [11C]-DPA713 PET scans twice in the morning and
afternoon on the same day and showed significant systematic differences in VT values
between the two scans (Coughlin et al., 2014). VT values obtained from the afternoon scans
were consistently higher than those obtained from the morning scans in all of the 16 ROIs
investigated with interater correlation coefficient (ICC) being 0.028 ± 0.057. The authors
also reported the inclusion of fp did not change the results. The underlying mechanisms for
the systematic differences are unknown. However, the authors discussed possible
contributing factors such as hormone-medicated phasic changes in TSPO expression, tonic
changes due to anxiety associated with the study procedure, or changes in the level of blood
cholesterol due to food intake between the two scans (Coughlin et al., 2014). These factors
may also contribute to the variability and obscure the true group differences. The
reproducibility of VT values using [18F]-FEPPA needs to determine.
In an effort to reduce the interindividual variability, some recent studies using the second-
generation TSPO radioligands have used normalized outcome measures such as VT ratio (the
ratio of the VT in the ROI to VT in the whole brain or GM), and found significant group
differences between clinical populations and HC subjects (Bloomfield et al., 2016; Coughlin
et al., 2014). However, there are some issues about using the normalized outcome measure.
In such approach, VT value in the ROI is also included in the denominator to calculate the
ratio. In addition, normalization using the whole brain or GM can compromise the accurate
estimate of neuroinflammation in pathological brains where neuroinflammation can occur in
diffuse areas of brain, which may be the case of PD. Furthermore, as discussed in section
1.6.3.5.3, TSPO density is different between in GM and WM, and the pharmacological
132
properties of [18F]-FEPPA in WM are different from those in GM. Thus, normalization using
the whole brain including WM or WM poses a problem.
Taken together, there are some factors that can be taken into consideration in the study and
analysis procedures to reduce noise. This may lead to the reduction of the exiting
interindividual variability in VT values and thereby increase the sensitivity of [18F]-FEPPA
detecting neuroinflammation in PD patients.
6.6.2 [18F]-FEPPA binding
VT value is a reliable measure of [18F]-FEPPA and other second-generation TSPO
radioligands. However, specific binding of most of these radioligands is unknown. There was
no suitable pharmacological agent to perform blocking studies to estimate non-displaceable
volume of distribution (VND) in human subjects. However, more recently, [11C]-PBR28 VND
and BPND (VT/VND-1) were estimated using the TSPO agonist XBD173 (emapunil),
medication for generalized anxiety (Rupprecht et al., 2009) in the healthy human subjects
and reported specific binding accounted for a substantial portion of VT (Owen et al., 2014).
Population VND was estimated to be 1.98 (95% Confidence Interval 1.69, 2.26).
Both microglia and astrocytes are the major cellular sources of TSPO expression in the brain
(Lavisse et al., 2012; Venneti et al., 2009). Unlike acute insults such as encephalitis, stroke,
or traumatic brain injury in which peripheral immune cells enter through damaged BBB,
microglia and astrocytes are mostly likely key immune response cells in neurodegenerative
diseases such as PD (Ouchi, Yagi, Yokokura, & Sakamoto, 2009). [3H]-DAA1106 binding
overlapped mostly with activated microglia compared with activated astrocytes in various
pathological conditions such as cerebral infarcts, amyotrophic lateral sclerosis, AD,
frontotemporal dementia, and MS (Venneti et al., 2008). However, due to lack of
autoradiography and immunohistochemistry studies of [18F]-FEPPA, the cellular sources of
[18F]-FEPPA binding is unknown.
6.6.3 Neuroinflammation in Parkinson’s disease
Our preliminary investigation of neuroinflammation using [18F]-FEPPA showed genetic
133
effects on VT values in both PD patients and HC subjects in the putamen and CN. HABs
showed significantly higher VT values than MABs in these regions except HC group in the
CN. On the other hand, there were no significant disease effects in either region. PD-HABs
showed 16% increase in TSPO binging compared to PD-MABs in these regions.
Some of the previous studies using [11C]-PK11195 found elevated microglial activation in
the putamen (Iannacone, et al., 2012; Ouchi, et al., 2009) and striatum (Gerhard, et al., 2006)
while others did not in the striatum (Bartel, et al., 2010; Edison, et al., 2013; Kobylecki, et
al., 2013). In fact, intervariability in neuroinflammation in PD patients was also reported in
postmortem (Graeber, Bise, & Mehraein, 1994) and neuroimaging studies (Edison et al.,
2013). It can depend on the disease stage, proximity to the lesion (Cosenza-Nashat et al.,
2009), disease severity, the rate of disease progression, and polymorphism in immune-related
genes (International Parkinson Disease Genomics Consortium et al., 2011). The investigation
of neuroinflammation will be extended to include a larger sample size or PD patients with
different clinical characteristics in the whole brain regions. Such study will determine
whether [18F]-FEPPA can be a biomarker for PD.
6.7 Neuroimaging biomarkers for Parkinson’s disease
This thesis utilized different neuroimaging tools to investigate structural, functional and
molecular changes in PD. The three studies included different cohorts of PD patients and HC
subjects. Therefore, the conclusions integrating all the findings should be interpreted with
caution. Our results suggest that (1) frontal motor (i.e., pre-SMA) and cognitive (i.e.,
DLPFC) regions may be early biomarker for motor and cognitive symptoms of PD, (2)
increased FC in the DLPFC may precede more substantial brain changes such as atrophy
associated with cognitive impairment, and (3) WM changes may precede GM changes and
functional changes as it revealed more pathological changes in diffuse area of the brain or
DTI with TBSS be the most promising biomarker detecting pathological changes compared
with other methods. Lastly, there was no evidence of neuroinflammation in the striatum of
PD. TSPO density was not correlated with any clinical measures, either. However, further
studies are needed to conclude whether neuroinflammation can be a biomarker for PD by
134
extending ROIs beyond the striatum, in particular in the frontal cortex and white matter
where we found brain changes using the other methods (Figure 6-1).
To date, there has been no single study using the multimodal neuroimaging tools and
analysis methods employed in this thesis. However, the findings of the two studies that
investigated GM, WM, and FC in the same cohort of PD patients are in line with our
findings. One study used VBM, TBSS and tractography for pedenculopontine tract, as well
as ICA to compare them between PD patients with freezing of gait with HC subjects (Canu et
al., 2015). The other study used CTA, TBSS and seed-based FC analysis with bilateral
posterior parietal cortex and CN as seed regions to compare these changes among PD
patients with normal cognition, PD patients who developed MCI in earlier disease and those
who developed MCI in later disease (Shin et al., 2016). Neither of the studies found
significant group differences in gray matter changes. On the other hand, both studies found
significant group differences in WM and FC changes. Using the whole brain analysis
methods including TBSS and ICA, WM revealed distributed changes in cortical and
subcortical regions including midbrain as well as cellebrum while FC changes were found
primarily in the sensorimotor network in PD patients with freezing of gait in an OFF
medication compared with HC subjects (Canu, et al., 2015). In addition, severe motor
symptoms measured by UPDRS III scores were associated with greater axial diffusivity of
the right pedunculopontine tract in these patients. In the other study, TBSS revealed
extensive changes in the frontal region, which corroborated our findings with the frontal
white matter most affected. Thus, consistent with our findings, these studies suggest that
WM and FC changes may precede GM changes or that DTI with TBSS be more sensitive
neuroimaging method to detect brain abnormalities associated with PD symptoms than
structural MRI with CTA or rsfMRI with ICA in PD.
135
Figure 6-1. Neuroimaging biomarkers for Parkinson’s disease. Our findings suggest that
TBSS be most promising biomarker for PD. However, further investigation on
neuroinflammation in the extra-striatal GM regions and white matter needs to confirm this
conclusion.
6.8 Limitations
Some limitations must be considered when interpreting the results of the studies included in
this thesis. First, the sample sizes were relatively small due to the inherent difficulty in
recruiting patients and “healthy” older adults. The replication of the findings in a larger
sample size is needed. Small sample sizes also limit us from subdividing patients into
subgroups with different disease stages, motor symptoms (e.g., right or left dominant,
tremor-dominant), cognitive phenotypes, and types of dopaminergic medication to further
dismantle the complexity of PD. Second, the study procedures were performed in an ON
medication state for patients except for the [18F]-FEPPA scan. This is due to the concerns
136
about the side effects and worsening of motor symptoms that patients may experience
without medication, which also can compromise the data quality. Thus, the findings in the
studies are associated with medicated PD patients. Third, the exact mechanisms underlying
for the changes in MRI and rsfMRI data are unknown. Similarly, the exact roles of activated
microglia and astrocytes (i.e., neurotoxic or neuroprotective) are unknown. Behavioral
correlates may help to interpret them. Fourth, the [18F]-FEPPA scans were performed using
HRRT, which features high spatial image resolution and high sensitivity (Wienhard et al.,
2002) and thereby allows for the accurate quantitative measurement of the radioligand
concentration even in small regions and regions with low tracer concentrations, such as the
striatal regions. However in aging and disease brains, where brain atrophy and enlarged
ventricles can be a significant issue for the accurate quantification, we analyzed our data both
with and without PVEC using the Mueller-Gartner PVEC algorithm. However, geometric
transfer matrix (GTM) (Rousset, Ma, & Evans, 1998) is often used for subcortical structures
such as BG to address spill-over effects due to the close proximity of structures (e.g. CN and
putamen) (Erlandsson, Buvat, Pretorius, Thomas, & Hutton, 2012). Lastly, there are general
limitations of second-generation TSPO radioligands including the inability to assess LABs
and the requirement of serial arterial blood sampling for quantification of TSPO, which poses
a challenge for the clinical use (Vivash & O'Brien, 2016). However, alternative methods for
the serial arterial blood sampling to derive input function have been proposed (Lavisse et al.,
2015; Mabrouk et al., 2014).
137
7.0 Conclusion
This thesis has shed light on both structural and functional brain changes underlying for the
clinical manifestations in PD patients using multimodal neuroimaging tools and techniques.
We demonstrated that cortical thinning in the left SFG (BA8 & 9) and diffused WM
abnormalities with frontal and temporal WM tracts mostly affected accounted for general
cognition as well as executive function in PD patients. We also demonstrated that the
diminished bridging role of pre-SMA with DMN is associated with more severe
bradykinesia. Taken together, we demonstrated the frontal changes as earlier sign of disease
associated with motor and cognitive dysfunction, and that structural MRI using CTA, DTI
with TBSS, and rsfMRI with graph theoretical analysis are potential in vivo biomarker for
PD. Among these brain changes, white matter changes were observed in diffuse area of the
brain, suggesting DTI with TBSS as the most promising neuroimaging method to detect
pathological changes in PD. Finally, using a second generation TSPO radioligand, [18F]-
FEPPA, the percentage of increase in TSPO binding in PD patients was 16 % in the putamen
as well as in the CN, which did not yield a statistically significant group difference. The
TSPO density was not correlated with any clinical measures, either. However, further studies
are needed to conclude whether neuroinflammation measured by [18F]-FEPPA can be a
biomarker for PD by extending ROIs beyond the striatum and by investigating a larger
sample size of PD patients as well as PD patients with different clinical characteristics.
138
8.0 Future Directions
In addition to addressing the limitations listed above, several possible future directions
arising from the studies in this thesis are as follows:
Proposed study 1: Whole brain investigation of neuroinflammation
We did not find any evidence of neuroinflammation in the striatum in 16 PD patients
compared with 16 HC subjects. However, we demonstrated pathological changes measured
by CTA and graph theoretical analysis in the frontal cortex in PD patients. Postmortem
studies showed more extensive neuroinflammation in the frontal cortex (Mythri et al., 2011)
and cortical regions (Braak et al., 2007) than in the striatal region. Thus, a pertinent
investigation of the extra-striatal GM regions would be needed. Similarly, we demonstrated
diffuse WM pathological changes in the cortical and subcortical regions, and postmortem PD
brains showed extensive axonal neuropathology, suggesting that further investigations
warrant in WM.
Proposed study 2: Neuroinflammation in subtypes of patients with Parkinson’s disease
Investigation in PD patients with different clinical characteristics may reveal more specific
role of neuroinflammation in different symptoms such as cognitive impairment, psychiatric
problems, and fatigue. Inflammatory markers in CSF samples from 87 PD patients including
16 PDD were compared with those form 33 HC subjects and studied for their association
with the severity of depression, anxiety, fatigue, and cognitive impairment (Lindqvist et al.,
2013). Only PDD patients showed significantly elevated levels of some of these markers
compared with non demented PD patients and HC subjects. However, higher levels of the
inflammatory markers were significantly associated with more severe depression, anxiety,
fatigue, and cognitive impairment, which was not observed in HC subjects. In addition, two
studies using [11C]-PK11195 demonstrated significantly elevated level of microglial
activation in distributed cortical regions (Edison et al., 2013; Fan et al., 2015), and the
amygdala and hippocampus (Fan et al., 2015) in PDD compared with HC. Both studies
showed inverse correlations between the level of microglial activation and MMSE scores.
139
Furthermore, neuroinflammation was detected in patients with AD (Suridgan, et al., 2015)
and with MDS (Setiawan, et al., 2015) using [18F]-FEPPA. Therefore, PD patients presenting
different NMS mentioned above would be investigated whether these patients show
neuroinflammation compared to PD patients without these NMS and to HC subjects using
[18F]-FEPPA. Furthermore, as the role of neuroinflammation has been implicated to
contribute to disease progression, PD patients with faster disease progression and those with
slower disease progression would be studied and followed up longitudinally to determine
whether TSPO imaging using [18F]-FEPPA can be a predictive measure for disease
progression.
Proposed study 3: Association between neuroinflammation and atrophy
We demonstrated that PD patients had structural changes in the study 1. The one of the
potential mechanisms underlying for GM and WM atrophy is neuroinflammation. Recently
published study demonstrated that volume loss in the hippocampus was associated with
elevated levels of neuroinflammation in the hippocampus measured by [11C]-PK11195 in AD
and PDD combined (Femminella et al., 2016). It also showed that the volume loss in the
hippocampus was associated with elevated levels of neuroinflammation in the diffuse
cortical and other subcortical regions. The authors suggested that neuronal degeneration
mediated by microglial activation follows the cortical connection from hippocampus because
of anatomical connections between hippocampal neurons project and the whole cortex via
entorhinal cortex. A few studies have investigated WM changes measured by DTI and
neuroinflammation using [11C]-PK11195 (Ramlackhansingh et al., 2011; Scott et al., 2015;
Thiel et al., 2010). However, to date, the relationships between structural and
neuroinflammation using the second generation TSPO radioligands have been little studied.
One study investigated TSPO binding using [11C]-DPA-713 as well as GM changes using the
CTA and subcortical volume analysis in the former NHL football players (Coughlin et al.,
2015). However, the relationships between GM changes and TSPO findings were not
investigated or not reported. Thus, multimodal approach combining structural MRI using
CTA, DTI, and TSPO imaging with [18F]-FEPPA may be able to elucidate such
relationships. Neuroinflammation has been implicated to contribute to the disease
progression of PD and it is also possible that positive findings of TSPO imaging may precede
140
GM or WM atrophy detected by MRI. Thus, a longitudinal study would reveal whether
TSPO density at one point could predict the disease progression that would be manifested on
structural changes at a following up.
Proposed study 4: Modeling Parkinson’s disease using hub changes
The current study identified nodes as hubs using network measures combined degree and
BC. However, as we demonstrated, diminished bridging role of a node between networks
measured by BC had an important implication for one of the cardinal motor symptoms of
PD. Thus, hub roles of nodes can be further investigated using graph measures such as
within-module degree z-score and participation coefficient. A node with higher within-
module degree z-score indicates that the node is a provincial hub while a node with higher
participation coefficient is a connector hub that has higher proportion of edges connecting
with nodes in other modules. To date, only one study investigated participation coefficient in
37 PD patients compared to 20 HC subjects (Gottlich et al., 2013). This study investigated
modular participation coefficient. It identified seven modules in HC subjects, and PD
patients showed significantly reduced modular participation coefficient in the visual module
compared to HC subjects. It further interrogated which nodes in the visual module
significantly reduced participation coefficient in PD patients compared with HC subjects, and
found that the calcarine and cuneus showed significant reduction in connectivity with frontal
regions including the ACC, superior orbital region and rolandic operculum. However, this
study used anatomically parcellated ROIs, which is known to compromise functional
boundaries. In addition, nodal participation coefficient was not investigated in other modules.
Furthermore, which nodes are high in participation coefficient in each module was not
investigated. Therefore, in the proposed project, data-driven nodes and modules would be
used. To this end, a group of older HC subjects would undergo fMRI scans, during which
they perform motor and cognitive tasks such as executive, memory, working memory tasks
to derive functional nodes. Additionally, PD patients and another group of older HC subjects
would undergo rsfMRI scans and these two groups would be interrogated for modular and
hub organizations including modularity, within-module degree z-score and participation
coefficient. Affected hubs in PD would be further investigated for their association with
clinical measures and for their target connections using a seed-based analysis as we
141
performed in study 2. This study would be longitudinally followed up. In the proposed
studies, it is hypothesized that pathology of connector hubs would (1) have greater impact on
symptoms than that of provincial hubs, (2) predict spread of disease depending on where the
hubs have connection to, and (3) predict faster progression of disease affecting distributed
brain regions. Furthermore, individual differences in robustness in such hub regions would
be able to predict the variability of disease progression. Longitudinal studies including early-
diagnosed PD patients may be able to confirm these models.
Proposed study 5: Investigation of increased functional connectivity as a pathological
response using MR Spectroscopy
We found increased FC within and between cognitive subnetworks in bilateral DLPFC in PD
patients compared to HC subjects, which is consistent with some of the PD literature.
Increased cortical FC was also reported in other diseases/disorders such as AD, depressive
and bipolar disorders, and SCZ. This enhanced FC is typically interpreted as a compensatory
response to the disease. However, it may also be pathological. Literature suggests that
increased FC in certain regions or increased FC in wider area of regions in cognitive
subnetworks might be due to altered glutamatergic/GABAergic activity (Anticevic et al.,
2012; Fornito et al., 2015; Krystal et al., 2003). In fact, postmortem PD brains showed
impairment of GABAergic neurotransmission in BA9 (Lanoue et al., 2010). The recent
development of 1HMRS at high magnetic field strengths allows for more reliable estimation
of the amino acids glutamine (Gln), Glu, and GABA (Ciurleo, Di Lorenzo, Bramanti, &
Marino, 2014) and thus may be able to elucidate the mechanisms underlying for cortico-
cortical FC changes in vivo. To this end, resting-state 1HMRS data would be acquired along
with fMRI data in PD patients and HC subjects. During the fMRI scans, the participants
would perform an executive task associated with the activation of DLPFC (BA9). Although
increased FC in bilateral DLPFC was not statistically associated with dopaminergic
medication in study 2, since DA exerts effects on NMDA and GABA activities, PD patients
would undergo fMRI scans in both on- and off-medication states. This study would allow to
determine (1) whether the resting-state Glu and GABA activities in the DLPFC would be
altered in PD patients compared to HC subjects, (2) whether these activities would predict
brain activation in the DLPFC and executive performance in PD patients, and (3) how
142
dopaminergic medication would affect cortical Glu and GABA activities as well as brain
activation in the DLPFC and executive function in PD patients.
Proposed study 6: Investigation of structural and functional brain network changes
using graph theoretical approach
Interpreting both structural and functional MRI data can be challenging even in the same
cohort due to different processing and analysis methods as well as outcome measures (e.g.,
gray matter thickness or volume; white matter microstructure; or connectivity strength or
characteristics). Both structural and function MRI data can be analyzed using the graph
theoretical approach. To date, published PD studies and our study applied such analysis
method solely to functional neuroimaging data such as rsfMRI, EEG, and MEG, structural
data including cortical thickness and subcortical volume (Pereira et al., 2015) or DTI data
with tractrography (Nigro et al., 2016). Therefore, by applying the graph theoretical approach
to structural, DTI, and rsfMRI would improve direct comparisons of the findings and thereby
elucidate their relationships with one another, as well as with the clinical symptoms in PD
patients.
Proposed study 7: Multimodal neuroimaging approach using PET and MRI to study
subtypes of Parkinson’s disease
PET can investigate more specific neuropathological changes than MRI, and therefore, such
findings can lead to the opportunities for pharmacological interventions. On the other hand,
MRI can be more preferred in a clinical setting as it is more widely available, more
accessible, less invasive and requires less scan time to monitor the course of disease
compared with PET. To utilize the advantages of both techniques, more studies using
multimodal neuroimaging approach are needed to investigate the relationships among the
outcome measures, and their associations with clinical symptoms. Such studies also need to
conduct longitudinally. In addition to the proposed study 2 (i.e., neuroinflammation and
structural changes), other PET radioligands such as [11C]-MP4A and [18F]FP-TZTP to
measure cholinergic function, and Pittsburgh compound B (PiB) to measure β amyloid
combined with MRI would also be used to study particular PD subpopulation such as PD-
MCI and PDD. PET with [18F]-Setoperone and [11C]-DASB to measure serotoninergic
143
function combined with MRI would be studied in PD patients with visual hallucinations.
Furthermore, PET findings can be used to group PD patients into subtypes instead of using
clinical/behavioral measures. In such study, PD patients would be first classified into
subgroups, such as high vs. low neuroinflammation, PiB positive vs negative, and PD with
normal vs abnormal cholinergic or serotonergic dysfunction based on their PET findings, and
these two PD patient subgroups would be compared with each other and with HC subjects,
and would be monitored using MRI to further characterize neuropathology in PD subtypes
and NMS. This study would also potentially pave the way for pharmacological interventions
for NMS.
144
References
Aarsland, D., Bronnick, K., & Fladby, T. (2011). Mild cognitive impairment in parkinson's
disease. Current Neurology and Neuroscience Reports, 11(4), 371-378.
Achard, S., & Bullmore, E. (2007). Efficiency and cost of economical brain functional
networks. PLoS Computational Biology, 3(2), e17.
Achard, S., Salvador, R., Whitcher, B., Suckling, J., & Bullmore, E. (2006). A resilient, low-
frequency, small-world human brain functional network with highly connected
association cortical hubs. The Journal of Neuroscience : The Official Journal of the
Society for Neuroscience, 26(1), 63-72.
Acosta-Cabronero, J., Williams, G. B., Pengas, G., & Nestor, P. J. (2010). Absolute
diffusivities define the landscape of white matter degeneration in alzheimer's disease.
Brain : A Journal of Neurology, 133(Pt 2), 529-539.
Agosta, F., Canu, E., Stefanova, E., Sarro, L., Tomic, A., Spica, V., . . . Filippi, M. (2014).
Mild cognitive impairment in parkinson's disease is associated with a distributed pattern
of brain white matter damage. Human Brain Mapping, 35(5), 1921-1929.
Agosta, F., Canu, E., Stojkovic, T., Pievani, M., Tomic, A., Sarro, L., . . . Filippi, M. (2013).
The topography of brain damage at different stages of parkinson's disease. Human Brain
Mapping, 34(11), 2798-2807.
Akaike, H. (1974). A new look at the statistical model identification. IEEE Trans Automat
Contr, 19, 716-723.
Alexander, A. L., Lee, J. E., Lazar, M., & Field, A. S. (2007). Diffusion tensor imaging of
the brain. Neurotherapeutics : The Journal of the American Society for Experimental
NeuroTherapeutics, 4(3), 316-329.
Alexander, G. E., DeLong, M. R., & Strick, P. L. (1986). Parallel organization of
functionally segregated circuits linking basal ganglia and cortex. Annual Review of
Neuroscience, 9, 357-381.
Alvarez, J. I., Katayama, T., & Prat, A. (2013). Glial influence on the blood brain barrier.
Glia, 61(12), 1939-1958.
Amaral, L. A., Scala, A., Barthelemy, M., & Stanley, H. E. (2000). Classes of small-world
networks. Proceedings of the National Academy of Sciences of the United States of
America, 97(21), 11149-11152.
145
Andersson, J. L. R., Jenkinson, M., & Smith, S. (2007a). Non-linear optimisation. FMRIB
technical report TR07JA1 from www.fmrib.ox.ac.uk/analysis/techrep.
Andersson, J. L. R., Jenkinson, M., & Smith, S. (2007b). Non-linear registration, aka spatial
normalisation FMRIB technical report TR07JA2 from
www.fmrib.ox.ac.uk/analysis/techrep.
Anticevic, A., Gancsos, M., Murray, J. D., Repovs, G., Driesen, N. R., Ennis, D. J., . . .
Corlett, P. R. (2012). NMDA receptor function in large-scale anticorrelated neural
systems with implications for cognition and schizophrenia. Proceedings of the National
Academy of Sciences of the United States of America, 109(41), 16720-16725.
Antkiewicz-Michaluk, L., Mukhin, A. G., Guidotti, A., & Krueger, K. E. (1988). Purification
and characterization of a protein associated with peripheral-type benzodiazepine binding
sites. The Journal of Biological Chemistry, 263(33), 17317-17321.
Arlicot, N., Vercouillie, J., Ribeiro, M. J., Tauber, C., Venel, Y., Baulieu, J. L., . . .
Guilloteau, D. (2012). Initial evaluation in healthy humans of [18F]DPA-714, a
potential PET biomarker for neuroinflammation. Nuclear Medicine and Biology, 39(4),
570-578.
Ashburner, J., & Friston, K. J. (2000). Voxel-based morphometry--the methods.
NeuroImage, 11(6 Pt 1), 805-821.
Assaf, Y., & Pasternak, O. (2008). Diffusion tensor imaging (DTI)-based white matter
mapping in brain research: A review. Journal of Molecular Neuroscience : MN, 34(1),
51-61.
Azmitia, E. C., & Nixon, R. (2008). Dystrophic serotonergic axons in neurodegenerative
diseases. Brain Research, 1217, 185-194.
Baggio, H. C., Sala-Llonch, R., Segura, B., Marti, M. J., Valldeoriola, F., Compta, Y., . . .
Junque, C. (2014). Functional brain networks and cognitive deficits in parkinson's
disease. Human Brain Mapping, 35(9), 4620-4634.
Baggio, H. C., Segura, B., Ibarretxe-Bilbao, N., Valldeoriola, F., Marti, M. J., Compta, Y., . .
. Junque, C. (2012). Structural correlates of facial emotion recognition deficits in
parkinson's disease patients. Neuropsychologia, 50(8), 2121-2128.
Bammer, R. (2003). Basic principles of diffusion-weighted imaging. European Journal of
Radiology, 45(3), 169-184.
Banati, R. B., Daniel, S. E., & Blunt, S. B. (1998). Glial pathology but absence of apoptotic
nigral neurons in long-standing parkinson's disease. Movement Disorders : Official
Journal of the Movement Disorder Society, 13(2), 221-227.
146
Banati, R. B., Newcombe, J., Gunn, R. N., Cagnin, A., Turkheimer, F., Heppner, F., . . .
Myers, R. (2000). The peripheral benzodiazepine binding site in the brain in multiple
sclerosis: Quantitative in vivo imaging of microglia as a measure of disease activity.
Brain, 123(Pt 11), 2321-2337.
Barnum, C. J., & Tansey, M. G. (2012). Neuroinflammation and non-motor symptoms: The
dark passenger of parkinson's disease? Current Neurology and Neuroscience Reports,
12(4), 350-358.
Bartels, A. L., Willemsen, A. T., Doorduin, J., de Vries, E. F., Dierckx, R. A., & Leenders,
K. L. (2010). [11C]-PK11195 PET: Quantification of neuroinflammation and a monitor
of anti-inflammatory treatment in parkinson's disease? Parkinsonism & Related
Disorders, 16(1), 57-59.
Basser, P. J. (1995). Inferring microstructural features and the physiological state of tissues
from diffusion-weighted images. NMR in Biomedicine, 8(7-8), 333-344.
Basser, P. J., Mattiello, J., & LeBihan, D. (1994). MR diffusion tensor spectroscopy and
imaging. Biophysical Journal, 66(1), 259-267.
Beaulieu, C. (2002). The basis of anisotropic water diffusion in the nervous system - a
technical review. NMR in Biomedicine, 15(7-8), 435-455.
Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for
measuring depression. Archives of General Psychiatry, 4, 561-571.
Behrens, T. E., Berg, H. J., Jbabdi, S., Rushworth, M. F., & Woolrich, M. W. (2007).
Probabilistic diffusion tractography with multiple fibre orientations: What can we gain?
NeuroImage, 34(1), 144-155.
Behrens, T. E., Woolrich, M. W., Jenkinson, M., Johansen-Berg, H., Nunes, R. G., Clare, S.,
. . . Smith, S. M. (2003). Characterization and propagation of uncertainty in diffusion-
weighted MR imaging. Magnetic Resonance in Medicine, 50(5), 1077-1088.
Behzadi, Y., Restom, K., Liau, J., & Liu, T. T. (2007). A component based noise correction
method (CompCor) for BOLD and perfusion based fMRI. NeuroImage, 37(1), 90-101.
Bellucci, A., Mercuri, N. B., Venneri, A., Faustini, G., Longhena, F., Pizzi, M., . . . Spano, P.
(2016). Review: Parkinson's disease: From synaptic loss to connectome dysfunction.
Neuropathology and Applied Neurobiology, 42(1), 77-94.
Benavides, J., Quarteronet, D., Imbault, F., Malgouris, C., Uzan, A., Renault, C., . . . Le Fur,
G. (1983). Labelling of "peripheral-type" benzodiazepine binding sites in the rat brain
by using [3H]PK 11195, an isoquinoline carboxamide derivative: Kinetic studies and
autoradiographic localization. Journal of Neurochemistry, 41(6), 1744-1750.
147
Bencherif, B., Stumpf, M. J., Links, J. M., & Frost, J. J. (2004). Application of MRI-based
partial-volume correction to the analysis of PET images of mu-opioid receptors using
statistical parametric mapping. Journal of Nuclear Medicine : Official Publication,
Society of Nuclear Medicine, 45(3), 402-408.
Bennacef, I. S. C., Horvath, G., Gunn, R., Bonasera, T., Wilson, A., Gee. A., & Laruelle, M.
(2008). Comparison of [11C]PBR28 and [18F]-FEPPA as CNS peripheral
benzodiazepine receptor PET ligands in the pig. J Nucl Med Meeting, 49, 81P-b.
Benskey, M. J., Perez, R. G., & Manfredsson, F. P. (2016). The contribution of alpha
synuclein to neuronal survival and function - implications for parkinson's disease.
Journal of Neurochemistry, 137(3), 331-359.
Benton, A. B., Sivan, K. D., Hamsher, N. R., Varney, N. R., & Spreen, O. (1994).
Contributions to neuropsychological assessment-2nd edition. Orland, FL: Psychological
Assessment Resources.
Berardelli, A., Rothwell, J. C., Thompson, P. D., & Hallett, M. (2001). Pathophysiology of
bradykinesia in parkinson's disease. Brain : A Journal of Neurology, 124(Pt 11), 2131-
2146.
Berg, D., Postuma, R. B., Bloem, B., Chan, P., Dubois, B., Gasser, T., . . . Deuschl, G.
(2014). Time to redefine PD? introductory statement of the MDS task force on the
definition of parkinson's disease. Movement Disorders : Official Journal of the
Movement Disorder Society, 29(4), 454-462.
Beyer, M. K., Janvin, C. C., Larsen, J. P., & Aarsland, D. (2007). A magnetic resonance
imaging study of patients with parkinson's disease with mild cognitive impairment and
dementia using voxel-based morphometry. Journal of Neurology, Neurosurgery, and
Psychiatry, 78(3), 254-259.
Biswal, B., Yetkin, F. Z., Haughton, V. M., & Hyde, J. S. (1995). Functional connectivity in
the motor cortex of resting human brain using echo-planar MRI. Magnetic Resonance in
Medicine, 34(4), 537-541.
Biundo, R., Calabrese, M., Weis, L., Facchini, S., Ricchieri, G., Gallo, P., & Antonini, A.
(2013). Anatomical correlates of cognitive functions in early parkinson's disease
patients. PloS One, 8(5), e64222.
Biundo, R., Formento-Dojot, P., Facchini, S., Vallelunga, A., Ghezzo, L., Foscolo, L., . . .
Antonini, A. (2011). Brain volume changes in parkinson's disease and their relationship
with cognitive and behavioural abnormalities. Journal of the Neurological Sciences,
310(1-2), 64-69.
Bloomfield, P. S., Selvaraj, S., Veronese, M., Rizzo, G., Bertoldo, A., Owen, D. R., . . .
Howes, O. D. (2016). Microglial activity in people at ultra high risk of psychosis and in
148
schizophrenia: An [(11)C]PBR28 PET brain imaging study. The American Journal of
Psychiatry, 173(1), 44-52.
Blum-Degen, D., Muller, T., Kuhn, W., Gerlach, M., Przuntek, H., & Riederer, P. (1995).
Interleukin-1 beta and interleukin-6 are elevated in the cerebrospinal fluid of alzheimer's
and de novo parkinson's disease patients. Neuroscience Letters, 202(1-2), 17-20.
Bonnelle, V., Ham, T. E., Leech, R., Kinnunen, K. M., Mehta, M. A., Greenwood, R. J., &
Sharp, D. J. (2012). Salience network integrity predicts default mode network function
after traumatic brain injury. Proceedings of the National Academy of Sciences of the
United States of America, 109(12), 4690-4695.
Bonnet, A. M., Jutras, M. F., Czernecki, V., Corvol, J. C., & Vidailhet, M. (2012). Nonmotor
symptoms in parkinson's disease in 2012: Relevant clinical aspects. Parkinson's
Disease, 2012, 198316.
Boutin, H., Chauveau, F., Thominiaux, C., Gregoire, M. C., James, M. L., Trebossen, R., . . .
Kassiou, M. (2007a). 11C-DPA-713: A novel peripheral benzodiazepine receptor PET
ligand for in vivo imaging of neuroinflammation. Journal of Nuclear Medicine : Official
Publication, Society of Nuclear Medicine, 48(4), 573-581.
Boutin, H., Chauveau, F., Thominiaux, C., Kuhnast, B., Gregoire, M. C., Jan, S., . . . Katsifis,
A. (2007b). In vivo imaging of brain lesions with [(11)C]CLINME, a new PET
radioligand of peripheral benzodiazepine receptors. Glia, 55(14), 1459-1468.
Boutin, H., Murray, K., Pradillo, J., Maroy, R., Smigova, A., Gerhard, A., . . . Trigg, W.
(2015). 18F-GE-180: A novel TSPO radiotracer compared to 11C-R-PK11195 in a
preclinical model of stroke. European Journal of Nuclear Medicine and Molecular
Imaging, 42(3), 503-511.
Braak, H., Bohl, J. R., Muller, C. M., Rub, U., de Vos, R. A., & Del Tredici, K. (2006).
Stanley fahn lecture 2005: The staging procedure for the inclusion body pathology
associated with sporadic parkinson's disease reconsidered. Movement Disorders :
Official Journal of the Movement Disorder Society, 21(12), 2042-2051.
Braak, H., & Braak, E. (2000). Pathoanatomy of parkinson's disease. Journal of Neurology,
247 Suppl 2, II3-10.
Braak, H., Braak, E., Yilmazer, D., de Vos, R. A., Jansen, E. N., Bohl, J., & Jellinger, K.
(1994). Amygdala pathology in parkinson's disease. Acta Neuropathologica, 88(6), 493-
500.
Braak, H., Braak, E., Yilmazer, D., Schultz, C., de Vos, R. A., & Jansen, E. N. (1995). Nigral
and extranigral pathology in parkinson's disease. Journal of Neural
Transmission.Supplementum, 46, 15-31.
149
Braak, H., de Vos, R. A., Jansen, E. N., Bratzke, H., & Braak, E. (1998). Neuropathological
hallmarks of alzheimer's and parkinson's diseases. Progress in Brain Research, 117,
267-285.
Braak, H., & Del Tredici, K. (2009). Neuroanatomy and pathology of sporadic parkinson's
disease. Advances in Anatomy, Embryology, and Cell Biology, 201, 1-119.
Braak, H., Del Tredici, K., Rub, U., de Vos, R. A., Jansen Steur, E. N., & Braak, E. (2003).
Staging of brain pathology related to sporadic parkinson's disease. Neurobiology of
Aging, 24(2), 197-211.
Braak, H., Ghebremedhin, E., Rub, U., Bratzke, H., & Del Tredici, K. (2004). Stages in the
development of parkinson's disease-related pathology. Cell and Tissue Research, 318(1),
121-134.
Braak, H., Rub, U., Jansen Steur, E. N., Del Tredici, K., & de Vos, R. A. (2005). Cognitive
status correlates with neuropathologic stage in parkinson disease. Neurology, 64(8),
1404-1410.
Braak, H., Sastre, M., & Del Tredici, K. (2007). Development of alpha-synuclein
immunoreactive astrocytes in the forebrain parallels stages of intraneuronal pathology in
sporadic parkinson's disease. Acta Neuropathologica, 114(3), 231-241.
Braestrup, C., & Squires, R. F. (1977). Specific benzodiazepine receptors in rat brain
characterized by high-affinity (3H)diazepam binding. Proceedings of the National
Academy of Sciences of the United States of America, 74(9), 3805-3809.
Brenneis, C., Seppi, K., Schocke, M. F., Muller, J., Luginger, E., Bosch, S., . . . Wenning, G.
K. (2003). Voxel-based morphometry detects cortical atrophy in the parkinson variant of
multiple system atrophy. Movement Disorders : Official Journal of the Movement
Disorder Society, 18(10), 1132-1138.
Bribes, E., Carriere, D., Goubet, C., Galiegue, S., Casellas, P., & Simony-Lafontaine, J.
(2004). Immunohistochemical assessment of the peripheral benzodiazepine receptor in
human tissues. The Journal of Histochemistry and Cytochemistry : Official Journal of
the Histochemistry Society, 52(1), 19-28.
Brundin, P., Li, J. Y., Holton, J. L., Lindvall, O., & Revesz, T. (2008). Research in motion:
The enigma of parkinson's disease pathology spread. Nature Reviews.Neuroscience,
9(10), 741-745.
Brundin, P., Melki, R., & Kopito, R. (2010). Prion-like transmission of protein aggregates in
neurodegenerative diseases. Nature Reviews.Molecular Cell Biology, 11(4), 301-307.
150
Bruno, J., Hosseini, S. M., & Kesler, S. (2012). Altered resting state functional brain network
topology in chemotherapy-treated breast cancer survivors. Neurobiology of Disease,
48(3), 329-338.
Buckner, R. L., Sepulcre, J., Talukdar, T., Krienen, F. M., Liu, H., Hedden, T., . . . Johnson,
K. A. (2009). Cortical hubs revealed by intrinsic functional connectivity: Mapping,
assessment of stability, and relation to alzheimer's disease. The Journal of Neuroscience
: The Official Journal of the Society for Neuroscience, 29(6), 1860-1873.
Buckner, R. L., & Vincent, J. L. (2007). Unrest at rest: Default activity and spontaneous
network correlations. NeuroImage, 37(4), 1091-6; discussion 1097-9.
Bullmore, E., & Sporns, O. (2009). Complex brain networks: Graph theoretical analysis of
structural and functional systems. Nature Reviews.Neuroscience, 10(3), 186-198.
Bullmore, E., & Sporns, O. (2012). The economy of brain network organization. Nature
Reviews.Neuroscience, 13(5), 336-349.
Burton, E. J., McKeith, I. G., Burn, D. J., Williams, E. D., & O'Brien, J. T. (2004). Cerebral
atrophy in parkinson's disease with and without dementia: A comparison with
alzheimer's disease, dementia with lewy bodies and controls. Brain : A Journal of
Neurology, 127(Pt 4), 791-800.
Cabeza, R., Anderson, N. D., Locantore, J. K., & McIntosh, A. R. (2002). Aging gracefully:
Compensatory brain activity in high-performing older adults. NeuroImage, 17(3), 1394-
1402.
Cagnin, A., Brooks, D. J., Kennedy, A. M., Gunn, R. N., Myers, R., Turkheimer, F. E., . . .
Banati, R. B. (2001). In-vivo measurement of activated microglia in dementia. Lancet,
358(9280), 461-467.
Cagnin, A., Rossor, M., Sampson, E. L., Mackinnon, T., & Banati, R. B. (2004). In vivo
detection of microglial activation in frontotemporal dementia. Annals of Neurology,
56(6), 894-897.
Camicioli, R., Gee, M., Bouchard, T. P., Fisher, N. J., Hanstock, C. C., Emery, D. J., &
Martin, W. R. (2009). Voxel-based morphometry reveals extra-nigral atrophy patterns
associated with dopamine refractory cognitive and motor impairment in parkinsonism.
Parkinsonism & Related Disorders, 15(3), 187-195.
Campbell, M. C., Koller, J. M., Snyder, A. Z., Buddhala, C., Kotzbauer, P. T., & Perlmutter,
J. S. (2015). CSF proteins and resting-state functional connectivity in parkinson disease.
Neurology, 84(24), 2413-2421.
151
Canat, X., Carayon, P., Bouaboula, M., Cahard, D., Shire, D., Roque, C., . . . Casellas, P.
(1993). Distribution profile and properties of peripheral-type benzodiazepine receptors
on human hemopoietic cells. Life Sciences, 52(1), 107-118.
Canu, E., Agosta, F., Sarasso, E., Volontè, M.A., Basaia, S., Stojkovic, T., … Filippi, M.
(2015). Brain structural and functional connectivity in Parkinson's disease with freezing
of gait. Human Brain Mapping, 36(12):5064-78.
Cao, H., Xu, X., Zhao, Y., Long, D., & Zhang, M. (2011). Altered brain activation and
connectivity in early parkinson disease tactile perception. AJNR.American Journal of
Neuroradiology, 32(10), 1969-1974.
Casellas, P., Galiegue, S., & Basile, A. S. (2002). Peripheral benzodiazepine receptors and
mitochondrial function. Neurochemistry International, 40(6), 475-486.
Caviness, J. N., Driver-Dunckley, E., Connor, D. J., Sabbagh, M. N., Hentz, J. G., Noble, B.,
. . . Adler, C. H. (2007). Defining mild cognitive impairment in parkinson's disease.
Movement Disorders : Official Journal of the Movement Disorder Society, 22(9), 1272-
1277.
Cerasa, A., Hagberg, G. E., Peppe, A., Bianciardi, M., Gioia, M. C., Costa, A., . . . Sabatini,
U. (2006). Functional changes in the activity of cerebellum and frontostriatal regions
during externally and internally timed movement in parkinson's disease. Brain Research
Bulletin, 71(1-3), 259-269.
Cerasa, A., Morelli, M., Augimeri, A., Salsone, M., Novellino, F., Gioia, M. C., . . .
Quattrone, A. (2013). Prefrontal thickening in PD with levodopa-induced dyskinesias:
New evidence from cortical thickness measurement. Parkinsonism & Related Disorders,
19(1), 123-125.
Chaki, S., Funakoshi, T., Yoshikawa, R., Okuyama, S., Okubo, T., Nakazato, A., . . .
Tomisawa, K. (1999). Binding characteristics of [3H]DAA1106, a novel and selective
ligand for peripheral benzodiazepine receptors. European Journal of Pharmacology,
371(2-3), 197-204.
Chang, L. J., Yarkoni, T., Khaw, M. W., & Sanfey, A. G. (2013). Decoding the role of the
insula in human cognition: Functional parcellation and large-scale reverse inference.
Cerebral Cortex, 23(3), 739-749.
Chaudhuri, K. R., Odin, P., Antonini, A., & Martinez-Martin, P. (2011). Parkinson's disease:
The non-motor issues. Parkinsonism & Related Disorders, 17(10), 717-723.
Chaudhuri, K. R., & Schapira, A. H. (2009). Non-motor symptoms of parkinson's disease:
Dopaminergic pathophysiology and treatment. The Lancet.Neurology, 8(5), 464-474.
152
Chauveau, F., Boutin, H., Van Camp, N., Dolle, F., & Tavitian, B. (2008). Nuclear imaging
of neuroinflammation: A comprehensive review of [11C]PK11195 challengers.
European Journal of Nuclear Medicine and Molecular Imaging, 35(12), 2304-2319.
Chen, M. K., Baidoo, K., Verina, T., & Guilarte, T. R. (2004). Peripheral benzodiazepine
receptor imaging in CNS demyelination: Functional implications of anatomical and
cellular localization. Brain : A Journal of Neurology, 127(Pt 6), 1379-1392.
Chen, M. K., & Guilarte, T. R. (2008). Translocator protein 18 kDa (TSPO): Molecular
sensor of brain injury and repair. Pharmacology & Therapeutics, 118(1), 1-17.
Chikama, M., McFarland, N. R., Amaral, D. G., & Haber, S. N. (1997). Insular cortical
projections to functional regions of the striatum correlate with cortical cytoarchitectonic
organization in the primate. The Journal of Neuroscience : The Official Journal of the
Society for Neuroscience, 17(24), 9686-9705.
Chiong, W., Wilson, S. M., D'Esposito, M., Kayser, A. S., Grossman, S. N., Poorzand, P., . .
. Rankin, K. P. (2013). The salience network causally influences default mode network
activity during moral reasoning. Brain : A Journal of Neurology, 136(Pt 6), 1929-1941.
Choi, H. B., Khoo, C., Ryu, J. K., van Breemen, E., Kim, S. U., & McLarnon, J. G. (2002).
Inhibition of lipopolysaccharide-induced cyclooxygenase-2, tumor necrosis factor-alpha
and [Ca2+]i responses in human microglia by the peripheral benzodiazepine receptor
ligand PK11195. Journal of Neurochemistry, 83(3), 546-555.
Christopher, L., Duff-Canning, S., Koshimori, Y., Segura, B., Boileau, I., Chen, R., . . .
Strafella, A. P. (2015). Salience network and parahippocampal dopamine dysfunction in
memory-impaired parkinson disease. Annals of Neurology, 77(2), 269-280.
Christopher, L., Koshimori, Y., Lang, A. E., Criaud, M., & Strafella, A. P. (2014).
Uncovering the role of the insula in non-motor symptoms of parkinson's disease. Brain :
A Journal of Neurology, 137(Pt 8), 2143-2154.
Christopher, L., Marras, C., Duff-Canning, S., Koshimori, Y., Chen, R., Boileau, I., . . .
Strafella, A. P. (2014). Combined insular and striatal dopamine dysfunction are
associated with executive deficits in parkinson's disease with mild cognitive impairment.
Brain : A Journal of Neurology, 137(Pt 2), 565-575.
Christopher, L., & Strafella, A. P. (2013). Neuroimaging of brain changes associated with
cognitive impairment in parkinson's disease. Journal of Neuropsychology, 7(2), 225-
240.
Chung, G. H., Han, Y. M., Jeong, S. H., & Jack, C. R.,Jr. (2005). Functional heterogeneity of
the supplementary motor area. AJNR.American Journal of Neuroradiology, 26(7), 1819-
1823.
153
Ciurleo, R., Di Lorenzo, G., Bramanti, P., & Marino, S. (2014). Magnetic resonance
spectroscopy: An in vivo molecular imaging biomarker for parkinson's disease? BioMed
Research International, 2014, 519816.
Clement, F., Gauthier, S., & Belleville, S. (2013). Executive functions in mild cognitive
impairment: Emergence and breakdown of neural plasticity. Cortex; a Journal Devoted
to the Study of the Nervous System and Behavior, 49(5), 1268-1279.
Cochrane, C. J., & Ebmeier, K. P. (2013). Diffusion tensor imaging in parkinsonian
syndromes: A systematic review and meta-analysis. Neurology, 80(9), 857-864.
Colasanti, A., Guo, Q., Giannetti, P., Wall, M.B., Newbould, R.D., Bishop, C., … Rabiner,
E.A. (2016). Hippocampal Neuroinflammation, Functional Connectivity, and
Depressive Symptoms in Multiple Sclerosis. Biological Psychiatry, 80(1), 62-72.
Cole, D.M., Beckmann, C.F., Oei, N.Y., Both, S., van Gerven, J.M., Rombouts. S.A. (2013).
Differential and distributed effects of dopamine neuromodulations on resting-state
network connectivity. Neuroimage, 78, 59-67.
Cole, D.M., Oei, N.Y., Soeter, R.P., Both, S., van Gerven, J.M., Rombouts, S.A., Beckmann,
C.F. (2013). Dopamine-dependent architecture of cortico-subcortical network
connectivity. Cereb Cortex, 23(7), 1509-16.
Compta, Y., Ibarretxe-Bilbao, N., Pereira, J. B., Junque, C., Bargallo, N., Tolosa, E., . . .
Marti, M. J. (2012). Grey matter volume correlates of cerebrospinal markers of
alzheimer-pathology in parkinson's disease and related dementia. Parkinsonism &
Related Disorders, 18(8), 941-947.
Compta, Y., Pereira, J. B., Rios, J., Ibarretxe-Bilbao, N., Junque, C., Bargallo, N., . . . Marti,
M. J. (2013). Combined dementia-risk biomarkers in parkinson's disease: A prospective
longitudinal study. Parkinsonism & Related Disorders, 19(8), 717-724.
Corbetta, M., Patel, G., & Shulman, G. L. (2008). The reorienting system of the human
brain: From environment to theory of mind. Neuron, 58(3), 306-324.
Corcia, P., Tauber, C., Vercoullie, J., Arlicot, N., Prunier, C., Praline, J., . . . Ribeiro, M. J.
(2012). Molecular imaging of microglial activation in amyotrophic lateral sclerosis.
PloS One, 7(12), e52941.
Cosenza-Nashat, M., Zhao, M. L., Suh, H. S., Morgan, J., Natividad, R., Morgello, S., &
Lee, S. C. (2009). Expression of the translocator protein of 18 kDa by microglia,
macrophages and astrocytes based on immunohistochemical localization in abnormal
human brain. Neuropathology and Applied Neurobiology, 35(3), 306-328.
Costa, B., Pini, S., Gabelloni, P., Da Pozzo, E., Abelli, M., Lari, L., . . . Martini, C. (2009).
The spontaneous Ala147Thr amino acid substitution within the translocator protein
154
influences pregnenolone production in lymphomonocytes of healthy individuals.
Endocrinology, 150(12), 5438-5445.
Coughlin, J. M., Wang, Y., Ma, S., Yue, C., Kim, P. K., Adams, A. V., . . . Pomper, M. G.
(2014). Regional brain distribution of translocator protein using [(11)C]DPA-713 PET
in individuals infected with HIV. Journal of Neurovirology, 20(3), 219-232.
Coughlin, J. M., Wang, Y., Munro, C. A., Ma, S., Yue, C., Chen, S., . . . Pomper, M. G.
(2015). Neuroinflammation and brain atrophy in former NFL players: An in vivo
multimodal imaging pilot study. Neurobiology of Disease, 74, 58-65.
Courtney, S. M., Petit, L., Maisog, J. M., Ungerleider, L. G., & Haxby, J. V. (1998). An area
specialized for spatial working memory in human frontal cortex. Science, 279(5355),
1347-1351.
Criaud, M., Christopher, L., Boulinguez, P., Ballanger, B., Lang, A. E., Cho, S. S., . . .
Strafella, A. P. (2016). Contribution of insula in parkinson's disease: A quantitative
meta-analysis study. Human Brain Mapping, 37(4), 1375-1392.
Croisier, E., Moran, L. B., Dexter, D. T., Pearce, R. K., & Graeber, M. B. (2005). Microglial
inflammation in the parkinsonian substantia nigra: Relationship to alpha-synuclein
deposition. Journal of Neuroinflammation, 2, 14.
Crossley, N. A., Mechelli, A., Scott, J., Carletti, F., Fox, P. T., McGuire, P., & Bullmore, E.
T. (2014). The hubs of the human connectome are generally implicated in the anatomy
of brain disorders. Brain : A Journal of Neurology, 137(Pt 8), 2382-2395.
Crossley, N. A., Mechelli, A., Vertes, P. E., Winton-Brown, T. T., Patel, A. X., Ginestet, C.
E., . . . Bullmore, E. T. (2013). Cognitive relevance of the community structure of the
human brain functional coactivation network. Proceedings of the National Academy of
Sciences of the United States of America, 110(28), 11583-11588.
Cubon, V. A., Putukian, M., Boyer, C., & Dettwiler, A. (2011). A diffusion tensor imaging
study on the white matter skeleton in individuals with sports-related concussion. Journal
of Neurotrauma, 28(2), 189-201.
Czeh, M., Gressens, P., & Kaindl, A. M. (2011). The yin and yang of microglia.
Developmental Neuroscience, 33(3-4), 199-209.
Dale, A. M., Fischl, B., & Sereno, M. I. (1999). Cortical surface-based analysis. I.
segmentation and surface reconstruction. NeuroImage, 9(2), 179-194.
Dale, A. M., & Sereno, M. I. (1993). Improved localizadon of cortical activity by combining
EEG and MEG with MRI cortical surface reconstruction: A linear approach. Journal of
Cognitive Neuroscience, 5(2), 162-176.
155
Damani, M. R., Zhao, L., Fontainhas, A. M., Amaral, J., Fariss, R. N., & Wong, W. T.
(2011). Age-related alterations in the dynamic behavior of microglia. Aging Cell, 10(2),
263-276.
Davis, S. W., Dennis, N. A., Daselaar, S. M., Fleck, M. S., & Cabeza, R. (2008). Que PASA?
the posterior-anterior shift in aging. Cerebral Cortex, 18(5), 1201-1209.
de Pasquale, F., Della Penna, S., Sporns, O., Romani, G. L., & Corbetta, M. (2015). A
dynamic core network and global efficiency in the resting human brain. Cerebral
Cortex, Sep 6.
de Pasquale, F., Sabatini, U., Della Penna, S., Sestieri, C., Caravasso, C. F., Formisano, R., &
Peran, P. (2013). The connectivity of functional cores reveals different degrees of
segregation and integration in the brain at rest. NeuroImage, 69, 51-61.
de Vries, F. E., de Wit, S. J., Cath, D. C., van der Werf, Y. D., van der Borden, V., van
Rossum, T. B., . . . van den Heuvel, O. A. (2014). Compensatory frontoparietal activity
during working memory: An endophenotype of obsessive-compulsive disorder.
Biological Psychiatry, 76(11), 878-887.
Defer, G. L., Widner, H., Marie, R. M., Remy, P., & Levivier, M. (1999). Core assessment
program for surgical interventional therapies in parkinson's disease (CAPSIT-PD).
Movement Disorders : Official Journal of the Movement Disorder Society, 14(4), 572-
584.
Delis, D. C., Kaplan, E., & Kramer, J. H. (2001). Examiner's manual for the delis-kaplan
executive function system. San Antonio, TX: The Psychological Corporation.
Delis, D. C., Kramer, J. H., Kaplan, B. A., & Ober, B. A. (2002). Manual for the california
verbal learning test-second edition. San Antonio TX: The psychological Corportaion.
Della-Maggiore, V., Sekuler, A. B., Grady, C. L., Bennett, P. J., Sekuler, R., & McIntosh, A.
R. (2000). Corticolimbic interactions associated with performance on a short-term
memory task are modified by age. The Journal of Neuroscience : The Official Journal of
the Society for Neuroscience, 20(22), 8410-8416.
DeLong, M. R., & Wichmann, T. (2015). Basal ganglia circuits as targets for
neuromodulation in parkinson disease. JAMA Neurology, 72(11), 1354-1360.
Deng, B., Zhang, Y., Wang, L., Peng, K., Han, L., Nie, K., . . . Wang, J. (2013). Diffusion
tensor imaging reveals white matter changes associated with cognitive status in patients
with parkinson's disease. American Journal of Alzheimer's Disease and Other
Dementias, 28(2), 154-164.
Desai Bradaric, B., Patel, A., Schneider, J. A., Carvey, P. M., & Hendey, B. (2012).
Evidence for angiogenesis in parkinson's disease, incidental lewy body disease, and
156
progressive supranuclear palsy. Journal of Neural Transmission (Vienna, Austria :
1996), 119(1), 59-71.
Desgranges, B., Baron, J. C., & Eustache, F. (1998). The functional neuroanatomy of
episodic memory: The role of the frontal lobes, the hippocampal formation, and other
areas. NeuroImage, 8(2), 198-213.
Deshpande, M., Zheng, J., Borgmann, K., Persidsky, R., Wu, L., Schellpeper, C., &
Ghorpade, A. (2005). Role of activated astrocytes in neuronal damage: Potential links to
HIV-1-associated dementia. Neurotoxicity Research, 7(3), 183-192.
Desplats, P., Lee, H. J., Bae, E. J., Patrick, C., Rockenstein, E., Crews, L., . . . Lee, S. J.
(2009). Inclusion formation and neuronal cell death through neuron-to-neuron
transmission of alpha-synuclein. Proceedings of the National Academy of Sciences of
the United States of America, 106(31), 13010-13015.
Dexter, D. T., & Jenner, P. (2013). Parkinson disease: From pathology to molecular disease
mechanisms. Free Radical Biology & Medicine, 62, 132-144.
Dilger, R. N., & Johnson, R. W. (2008). Aging, microglial cell priming, and the discordant
central inflammatory response to signals from the peripheral immune system. Journal of
Leukocyte Biology, 84(4), 932-939.
Dobbs, R. J., Charlett, A., Purkiss, A. G., Dobbs, S. M., Weller, C., & Peterson, D. W.
(1999). Association of circulating TNF-alpha and IL-6 with ageing and parkinsonism.
Acta Neurologica Scandinavica, 100(1), 34-41.
Doble, A., Malgouris, C., Daniel, M., Daniel, N., Imbault, F., Basbaum, A., . . . Le Fur, G.
(1987). Labelling of peripheral-type benzodiazepine binding sites in human brain with
[3H]PK 11195: Anatomical and subcellular distribution. Brain Research Bulletin, 18(1),
49-61.
Doorduin, J., Klein, H. C., Dierckx, R. A., James, M., Kassiou, M., & de Vries, E. F. (2009).
11C]-DPA-713 and [18F]-DPA-714 as new PET tracers for TSPO: A comparison with
[11C]-(R)-PK11195 in a rat model of herpes encephalitis. Molecular Imaging and
Biology : MIB : The Official Publication of the Academy of Molecular Imaging, 11(6),
386-398.
Doorn, K. J., Goudriaan, A., Blits-Huizinga, C., Bol, J. G., Rozemuller, A. J., Hoogland, P.
V., . . . van Dam, A. M. (2014a). Increased amoeboid microglial density in the olfactory
bulb of parkinson's and alzheimer's patients. Brain Pathology, 24(2), 152-165.
Doorn, K. J., Moors, T., Drukarch, B., van de Berg, W. D., Lucassen, P. J., & van Dam, A.
M. (2014b). Microglial phenotypes and toll-like receptor 2 in the substantia nigra and
hippocampus of incidental lewy body disease cases and parkinson's disease patients.
Acta Neuropathologica Communications, 2, 90-014-0090-1.
157
Dosenbach, N. U., Fair, D. A., Cohen, A. L., Schlaggar, B. L., & Petersen, S.E. (2008). A
dual-networks architecture of top-down control. Trendsin Cognitive Sciences, 12(3), 99-
105.
Dosenbach, N. U., Fair, D. A., Miezin, F. M., Cohen, A. L., Wenger, K. K., Dosenbach, R.
A., . . . Petersen, S. E. (2007). Distinct brain networks for adaptive and stable task
control in humans. Proceedings of the National Academy of Sciences of the United
States of America, 104(26), 11073-11078.
Dosenbach, N. U., Nardos, B., Cohen, A. L., Fair, D. A., Power, J. D., Church, J. A., . . .
Schlaggar, B. L. (2010). Prediction of individual brain maturity using fMRI. Science,
329(5997), 1358-1361.
du Boisgueheneuc, F., Levy, R., Volle, E., Seassau, M., Duffau, H., Kinkingnehun, S., . . .
Dubois, B. (2006). Functions of the left superior frontal gyrus in humans: A lesion
study. Brain : A Journal of Neurology, 129(Pt 12), 3315-3328.
Duncan, G. W., Firbank, M. J., O'Brien, J. T., & Burn, D. J. (2013). Magnetic resonance
imaging: A biomarker for cognitive impairment in parkinson's disease? Movement
Disorders : Official Journal of the Movement Disorder Society, 28(4), 425-438.
Edison, P., Ahmed, I., Fan, Z., Hinz, R., Gelosa, G., Ray Chaudhuri, K., . . . Brooks, D. J.
(2013). Microglia, amyloid, and glucose metabolism in parkinson's disease with and
without dementia. Neuropsychopharmacology : Official Publication of the American
College of Neuropsychopharmacology, 38(6), 938-949.
Endres, C. J., Pomper, M. G., James, M., Uzuner, O., Hammoud, D. A., Watkins, C. C., . . .
Kassiou, M. (2009). Initial evaluation of 11C-DPA-713, a novel TSPO PET ligand, in
humans. Journal of Nuclear Medicine : Official Publication, Society of Nuclear
Medicine, 50(8), 1276-1282.
Eng, L. F., Ghirnikar, R. S., & Lee, Y. L. (2000). Glial fibrillary acidic protein: GFAP-thirty-
one years (1969-2000). Neurochemical Research, 25(9-10), 1439-1451.
Engel, A. K., Gerloff, C., Hilgetag, C. C., & Nolte, G. (2013). Intrinsic coupling modes:
Multiscale interactions in ongoing brain activity. Neuron, 80(4), 867-886.
Erlandsson, K., Buvat, I., Pretorius, P. H., Thomas, B. A., & Hutton, B. F. (2012). A review
of partial volume correction techniques for emission tomography and their applications
in neurology, cardiology and oncology. Physics in Medicine and Biology, 57(21), R119-
59.
Esposito, F., Tessitore, A., Giordano, A., De Micco, R., Paccone, A., Conforti, R., …
Tedeschi, G. (2013). Rhythm-specific modulation of the sensorimotor network in
drug-naive patients with Parkinson's disease by levodopa. Brain,136(Pt 3):710-725.
Evans, A. H., Katzenschlager, R., Paviour, D., O'Sullivan, J. D., Appel, S., Lawrence, A. D.,
158
& Lees, A. J. (2004). Punding in parkinson's disease: Its relation to the dopamine
dysregulation syndrome. Movement Disorders : Official Journal of the Movement
Disorder Society, 19(4), 397-405.
Fan, J., Lindemann, P., Feuilloley, M. G., & Papadopoulos, V. (2012). Structural and
functional evolution of the translocator protein (18 kDa). Current Molecular Medicine,
12(4), 369-386.
Farina, E., Gattellaro, G., Pomati, S., Magni, E., Perretti, A., Cannata, A. P., . . . Mariani, C.
(2000). Researching a differential impairment of frontal functions and explicit memory
in early parkinson's disease. European Journal of Neurology, 7(3), 259-267.
Feldmann, A., Illes, Z., Kosztolanyi, P., Illes, E., Mike, A., Kover, F., . . . Nagy, F. (2008).
Morphometric changes of gray matter in parkinson's disease with depression: A voxel-
based morphometry study. Movement Disorders : Official Journal of the Movement
Disorder Society, 23(1), 42-46.
Femminella, G. D., Ninan, S., Atkinson, R., Fan, Z., Brooks, D. J., & Edison, P. (2016).
Does microglial activation influence hippocampal volume and neuronal function in
alzheimer's disease and parkinson's disease dementia? Journal of Alzheimer's Disease :
JAD, 51(4), 1275-1289.
Ferrer, I., Martinez, A., Blanco, R., Dalfo, E., & Carmona, M. (2011). Neuropathology of
sporadic parkinson disease before the appearance of parkinsonism: Preclinical parkinson
disease. Journal of Neural Transmission, 118(5), 821-839.
Finch, C. E. (2003). Neurons, glia, and plasticity in normal brain aging. Neurobiology of
Aging, 24 Suppl 1, S123-7; discussion S131.
Fischl, B., & Dale, A. M. (2000). Measuring the thickness of the human cerebral cortex from
magnetic resonance images. Proceedings of the National Academy of Sciences of the
United States of America, 97(20), 11050-11055.
Fischl, B., Liu, A., & Dale, A. M. (2001). Automated manifold surgery: Constructing
geometrically accurate and topologically correct models of the human cerebral cortex.
IEEE Transactions on Medical Imaging, 20(1), 70-80.
Floresco, S. B., & Magyar, O. (2006). Mesocortical dopamine modulation of executive
functions: Beyond working memory. Psychopharmacology, 188(4), 567-585.
Fogelson, N., Li, L., Li, Y., Fernandez-Del-Olmo, M., Santos-Garcia, D., & Peled, A.
(2013). Functional connectivity abnormalities during contextual processing in
schizophrenia and in parkinson's disease. Brain and Cognition, 82(3), 243-253.
Foley, P., & Riederer, P. (1999). Pathogenesis and preclinical course of parkinson's disease.
Journal of Neural Transmission.Supplementum, 56, 31-74.
159
Fornito, A., Zalesky, A., & Breakspear, M. (2015). The connectomics of brain disorders.
Nature Reviews.Neuroscience, 16(3), 159-172.
Freeman, M. R. (2010). Specification and morphogenesis of astrocytes. Science, 330(6005),
774-778.
Friston, K. J., Frith, C. D., Liddle, P. F., & Frackowiak, R. S. (1993). Functional
connectivity: The principal-component analysis of large (PET) data sets. Journal of
Cerebral Blood Flow and Metabolism : Official Journal of the International Society of
Cerebral Blood Flow and Metabolism, 13(1), 5-14.
Friston, K. J., Harrison, L., & Penny, W. (2003). Dynamic causal modelling. NeuroImage,
19(4), 1273-1302.
Fujimura, Y., Ikoma, Y., Yasuno, F., Suhara, T., Ota, M., Matsumoto, R., . . . Ito, H. (2006).
Quantitative analyses of 18F-FEDAA1106 binding to peripheral benzodiazepine
receptors in living human brain. Journal of Nuclear Medicine : Official Publication,
Society of Nuclear Medicine, 47(1), 43-50.
Fujimura, Y., Zoghbi, S. S., Simeon, F. G., Taku, A., Pike, V. W., Innis, R. B., & Fujita, M.
(2009). Quantification of translocator protein (18 kDa) in the human brain with PET and
a novel radioligand, (18)F-PBR06. Journal of Nuclear Medicine : Official Publication,
Society of Nuclear Medicine, 50(7), 1047-1053.
Fujita, M., Imaizumi, M., Zoghbi, S. S., Fujimura, Y., Farris, A. G., Suhara, T., . . . Innis, R.
B. (2008). Kinetic analysis in healthy humans of a novel positron emission tomography
radioligand to image the peripheral benzodiazepine receptor, a potential biomarker for
inflammation. NeuroImage, 40(1), 43-52.
Gallagher, C., Bell, B., Bendlin, B., Palotti, M., Okonkwo, O., Sodhi, A., . . . Alexander, A.
(2013). White matter microstructural integrity and executive function in parkinson's
disease. Journal of the International Neuropsychological Society : JINS, 19(3), 349-354.
Galvan, A., Devergnas, A., & Wichmann, T. (2015). Alterations in neuronal activity in basal
ganglia-thalamocortical circuits in the parkinsonian state. Frontiers in Neuroanatomy, 9,
5.
Gao, H. M., & Hong, J. S. (2008). Why neurodegenerative diseases are progressive:
Uncontrolled inflammation drives disease progression. Trends in Immunology, 29(8),
357-365.
Garnier, M., Dimchev, A. B., Boujrad, N., Price, J. M., Musto, N. A., & Papadopoulos, V.
(1994). In vitro reconstitution of a functional peripheral-type benzodiazepine receptor
from mouse leydig tumor cells. Molecular Pharmacology, 45(2), 201-211.
160
Gatliff, J., & Campanella, M. (2012). The 18 kDa translocator protein (TSPO): A new
perspective in mitochondrial biology. Current Molecular Medicine, 12(4), 356-368.
Gattellaro, G., Minati, L., Grisoli, M., Mariani, C., Carella, F., Osio, M., . . . Bruzzone, M. G.
(2009). White matter involvement in idiopathic parkinson disease: A diffusion tensor
imaging study. AJNR.American Journal of Neuroradiology, 30(6), 1222-1226.
Gerhard, A., Neumaier, B., Elitok, E., Glatting, G., Ries, V., Tomczak, R., . . . Reske, S. N.
(2000). In vivo imaging of activated microglia using [11C]PK11195 and positron
emission tomography in patients after ischemic stroke. Neuroreport, 11(13), 2957-2960.
Gerhard, A., Pavese, N., Hotton, G., Turkheimer, F., Es, M., Hammers, A., . . . Brooks, D. J.
(2006). In vivo imaging of microglial activation with [11C](R)-PK11195 PET in
idiopathic parkinson's disease. Neurobiology of Disease, 21(2), 404-412.
Gerrits, N. J., van der Werf, Y. D., Verhoef, K. M., Veltman, D. J., Groenewegen, H. J.,
Berendse, H. W., & van den Heuvel, O. A. (2015). Compensatory fronto-parietal
hyperactivation during set-shifting in unmedicated patients with parkinson's disease.
Neuropsychologia, 68, 107-116.
Gerrits, N. J., van Loenhoud A.C., van den Berg, S. F., Berendse, H. W., Foncke, E. M.,
Klein, M., … van den Heuvel, O. A. (2016). Cortical thickness. Surface are and
subcoritical volume differntially contribute to cognitive heterogeneity in Parkinson's
disease. PLoS One, 11(2), e0148852.
Glass, C. K., Saijo, K., Winner, B., Marchetto, M. C., & Gage, F. H. (2010). Mechanisms
underlying inflammation in neurodegeneration. Cell, 140(6), 918-934.
Goedert, M., Spillantini, M. G., Del Tredici, K., & Braak, H. (2013). 100 years of lewy
pathology. Nature Reviews.Neurology, 9(1), 13-24.
Goldman, J. G., Stebbins, G. T., Bernard, B., Stoub, T. R., Goetz, C. G., & deToledo-
Morrell, L. (2012). Entorhinal cortex atrophy differentiates parkinson's disease patients
with and without dementia. Movement Disorders : Official Journal of the Movement
Disorder Society, 27(6), 727-734.
Golla, S. S., Boellaard, R., Oikonen, V., Hoffmann, A., van Berckel, B. N., Windhorst, A.
D., . . . Rinne, J. O. (2015). Quantification of [18F]DPA-714 binding in the human
brain: Initial studies in healthy controls and alzheimer's disease patients. Journal of
Cerebral Blood Flow and Metabolism : Official Journal of the International Society of
Cerebral Blood Flow and Metabolism, 35(5), 766-772.
Gong, G., Rosa-Neto, P., Carbonell, F., Chen, Z. J., He, Y., & Evans, A. C. (2009). Age- and
gender-related differences in the cortical anatomical network. The Journal of
Neuroscience : The Official Journal of the Society for Neuroscience, 29(50), 15684-
15693.
161
Gottlich, M., Munte, T. F., Heldmann, M., Kasten, M., Hagenah, J., & Kramer, U. M.
(2013). Altered resting state brain networks in parkinson's disease. PloS One, 8(10),
e77336.
Grady, C. (2012). The cognitive neuroscience of ageing. Nature Reviews.Neuroscience,
13(7), 491-505.
Grady, C. L. (2008). Cognitive neuroscience of aging. Annals of the New York Academy of
Sciences, 1124, 127-144.
Graeber, M. B., Bise, K., & Mehraein, P. (1994). CR3/43, a marker for activated human
microglia: Application to diagnostic neuropathology. Neuropathology and Applied
Neurobiology, 20(4), 406-408.
Graeber, M. B., & Streit, W. J. (2010). Microglia: Biology and pathology. Acta
Neuropathologica, 119(1), 89-105.
Gratwicke, J., Jahanshahi, M., & Foltynie, T. (2015). Parkinson's disease dementia: A neural
networks perspective. Brain : A Journal of Neurology, 138(Pt 6), 1454-1476.
Greicius, M. (2008). Resting-state functional connectivity in neuropsychiatric disorders.
Current Opinion in Neurology, 21(4), 424-430.
Griffith, H. R., den Hollander, J. A., Okonkwo, O. C., O'Brien, T., Watts, R. L., & Marson,
D. C. (2008). Brain metabolism differs in alzheimer's disease and parkinson's disease
dementia. Alzheimer's & Dementia : The Journal of the Alzheimer's Association, 4(6),
421-427.
Griffith, H. R., Okonkwo, O. C., O'Brien, T., & Hollander, J. A. (2008). Reduced brain
glutamate in patients with parkinson's disease. NMR in Biomedicine, 21(4), 381-387.
Guidotti, A., Forchetti, C. M., Corda, M. G., Konkel, D., Bennett, C. D., & Costa, E. (1983).
Isolation, characterization, and purification to homogeneity of an endogenous
polypeptide with agonistic action on benzodiazepine receptors. Proceedings of the
National Academy of Sciences of the United States of America, 80(11), 3531-3535.
Guimera, R., & Amaral, L. A. (2005). Cartography of complex networks: Modules and
universal roles. Journal of Statistical Mechanics, 2005(P02001), nihpa35573.
Gulani, V., & Sundgren, P. C. (2006). Diffusion tensor magnetic resonance imaging. Journal
of Neuro-Ophthalmology : The Official Journal of the North American Neuro-
Ophthalmology Society, 26(1), 51-60.
Gulyas, B., Vas, A., Toth, M., Takano, A., Varrone, A., Cselenyi, Z., . . . Halldin, C. (2011).
Age and disease related changes in the translocator protein (TSPO) system in the human
162
brain: Positron emission tomography measurements with [11C]vinpocetine.
NeuroImage, 56(3), 1111-1121.
Guo, Q., Colasanti, A., Owen, D. R., Onega, M., Kamalakaran, A., Bennacef, I., . . . Gunn,
R. N. (2013). Quantification of the specific translocator protein signal of 18F-PBR111
in healthy humans: A genetic polymorphism effect on in vivo binding. Journal of
Nuclear Medicine : Official Publication, Society of Nuclear Medicine, 54(11), 1915-
1923.
Halliday, G. M., & Stevens, C. H. (2011). Glia: Initiators and progressors of pathology in
parkinson's disease. Movement Disorders : Official Journal of the Movement Disorder
Society, 26(1), 6-17.
Hamby, M. E., & Sofroniew, M. V. (2010). Reactive astrocytes as therapeutic targets for
CNS disorders. Neurotherapeutics : The Journal of the American Society for
Experimental NeuroTherapeutics, 7(4), 494-506.
Hammoud, D. A., Endres, C. J., Chander, A. R., Guilarte, T. R., Wong, D. F., Sacktor, N. C.,
. . . Pomper, M. G. (2005). Imaging glial cell activation with [11C]-R-PK11195 in
patients with AIDS. Journal of Neurovirology, 11(4), 346-355.
Hanganu, A., Bedetti, C., Jubault, T., Gagnon, J. F., Mejia-Constain, B., Degroot, C., . . .
Monchi, O. (2013). Mild cognitive impairment in patients with parkinson's disease is
associated with increased cortical degeneration. Movement Disorders : Official Journal
of the Movement Disorder Society, 28(10), 1360-1369.
Hansen, C., Angot, E., Bergstrom, A. L., Steiner, J. A., Pieri, L., Paul, G., . . . Brundin, P.
(2011). Alpha-synuclein propagates from mouse brain to grafted dopaminergic neurons
and seeds aggregation in cultured human cells. The Journal of Clinical Investigation,
121(2), 715-725.
Haslinger, B., Erhard, P., Kampfe, N., Boecker, H., Rummeny, E., Schwaiger, M., . . .
Ceballos-Baumann, A. O. (2001). Event-related functional magnetic resonance imaging
in parkinson's disease before and after levodopa. Brain : A Journal of Neurology, 124(Pt
3), 558-570.
Hatano, K., Yamada, T., Toyama, H., Kudo, G., Nomura, M., Suzuki, H., . . . Ito, K. (2010).
Correlation between FEPPA uptake and microglia activation in 6-OHDA injured rat
brain. NeuroImage, 52, S138-S.
Hattori, T., Orimo, S., Aoki, S., Ito, K., Abe, O., Amano, A., . . . Mizusawa, H. (2012).
Cognitive status correlates with white matter alteration in parkinson's disease. Human
Brain Mapping, 33(3), 727-739.
Hawkes, C. H., Del Tredici, K., & Braak, H. (2009). Parkinson's disease: The dual hit theory
revisited. Annals of the New York Academy of Sciences, 1170, 615-622.
163
Heeger, D. J., & Ress, D. (2002). What does fMRI tell us about neuronal activity? Nature
Reviews.Neuroscience, 3(2), 142-151.
Helmich, R. C., Derikx, L. C., Bakker, M., Scheeringa, R., Bloem, B. R., & Toni, I. (2010).
Spatial remapping of cortico-striatal connectivity in parkinson's disease. Cerebral
Cortex, 20(5), 1175-1186.
Hickman, S. E., Allison, E. K., & El Khoury, J. (2008). Microglial dysfunction and defective
beta-amyloid clearance pathways in aging alzheimer's disease mice. The Journal of
Neuroscience : The Official Journal of the Society for Neuroscience, 28(33), 8354-8360.
Hirsch, E. C., & Hunot, S. (2009). Neuroinflammation in parkinson's disease: A target for
neuroprotection? Lancet Neurology, 8(4), 382-397.
Honey, C. J., Sporns, O., Cammoun, L., Gigandet, X., Thiran, J. P., Meuli, R., & Hagmann,
P. (2009). Predicting human resting-state functional connectivity from structural
connectivity. Proceedings of the National Academy of Sciences of the United States of
America, 106(6), 2035-2040.
Hong, J. Y., Lee, J. E., Sohn, Y. H., & Lee, P. H. (2012). Neurocognitive and atrophic
patterns in parkinson's disease based on subjective memory complaints. Journal of
Neurology, 259(8), 1706-1712.
Horsfield, M. A., Larsson, H. B., Jones, D. K., & Gass, A. (1998). Diffusion magnetic
resonance imaging in multiple sclerosis. Journal of Neurology, Neurosurgery, and
Psychiatry, 64 Suppl 1, S80-4.
Hosseini, S. M., Hoeft, F., & Kesler, S. R. (2012). GAT: A graph-theoretical analysis
toolbox for analyzing between-group differences in large-scale structural and functional
brain networks. PloS One, 7(7), e40709.
Houlden, H., & Singleton, A. B. (2012). The genetics and neuropathology of parkinson's
disease. Acta Neuropathologica, 124(3), 325-338.
Howlett, D. R., Whitfield, D., Johnson, M., Attems, J., O'Brien, J. T., Aarsland, D., . . .
Francis, P. T. (2015). Regional multiple pathology scores are associated with cognitive
decline in lewy body dementias. Brain Pathology (Zurich, Switzerland), 25(4), 401-408.
Hu, M. T., White, S. J., Chaudhuri, K. R., Morris, R. G., Bydder, G. M., & Brooks, D. J.
(2001). Correlating rates of cerebral atrophy in parkinson's disease with measures of
cognitive decline. Journal of Neural Transmission, 108(5), 571-580.
Hua, K., Zhang, J., Wakana, S., Jiang, H., Li, X., Reich, D. S., . . . Mori, S. (2008). Tract
probability maps in stereotaxic spaces: Analyses of white matter anatomy and tract-
specific quantification. NeuroImage, 39(1), 336-347.
164
Huang, P., Lou, Y., Xuan, M., Gu, Q., Guan, X., Xu, X., et al., … Zhang, M. (2016). Cortical
abnormalities in Parkinson's disease patients and relationship to depression: A surface-
based morphometry study. Psychiatry Reserch: Neuroimaging, 250, 24-28.
Huang, X., Alonso, A., Guo, X., Umbach, D. M., Lichtenstein, M. L., Ballantyne, C. M., . . .
Chen, H. (2015). Statins, plasma cholesterol, and risk of parkinson's disease: A
prospective study. Movement Disorders : Official Journal of the Movement Disorder
Society, 30(4), 552-559.
Huppi, P. S., Maier, S. E., Peled, S., Zientara, G. P., Barnes, P. D., Jolesz, F. A., & Volpe, J.
J. (1998). Microstructural development of human newborn cerebral white matter
assessed in vivo by diffusion tensor magnetic resonance imaging. Pediatric Research,
44(4), 584-590.
Iannaccone, S., Cerami, C., Alessio, M., Garibotto, V., Panzacchi, A., Olivieri, S., . . . Perani,
D. (2013). In vivo microglia activation in very early dementia with lewy bodies,
comparison with parkinson's disease. Parkinsonism & Related Disorders, 19(1), 47-52.
Ibarretxe-Bilbao, N., Junque, C., Marti, M. J., Valldeoriola, F., Vendrell, P., Bargallo, N., . . .
Tolosa, E. (2010). Olfactory impairment in parkinson's disease and white matter
abnormalities in central olfactory areas: A voxel-based diffusion tensor imaging study.
Movement Disorders : Official Journal of the Movement Disorder Society, 25(12),
1888-1894.
Ibarretxe-Bilbao, N., Junque, C., Segura, B., Baggio, H. C., Marti, M. J., Valldeoriola, F., . . .
Tolosa, E. (2012). Progression of cortical thinning in early parkinson's disease.
Movement Disorders : Official Journal of the Movement Disorder Society, 27(14),
1746-1753.
Ibarretxe-Bilbao, N., Junque, C., Tolosa, E., Marti, M. J., Valldeoriola, F., Bargallo, N., &
Zarei, M. (2009). Neuroanatomical correlates of impaired decision-making and facial
emotion recognition in early parkinson's disease. The European Journal of
Neuroscience, 30(6), 1162-1171.
Ibarretxe-Bilbao, N., Ramirez-Ruiz, B., Junque, C., Marti, M. J., Valldeoriola, F., Bargallo,
N., . . . Tolosa, E. (2010). Differential progression of brain atrophy in parkinson's
disease with and without visual hallucinations. Journal of Neurology, Neurosurgery,
and Psychiatry, 81(6), 650-657.
Ikeda, A., Yazawa, S., Kunieda, T., Ohara, S., Terada, K., Mikuni, N., . . . Shibasaki, H.
(1999). Cognitive motor control in human pre-supplementary motor area studied by
subdural recording of discrimination/selection-related potentials. Brain : A Journal of
Neurology, 122(Pt5), 915-931.
165
Imaizumi, M., Briard, E., Zoghbi, S. S., Gourley, J. P., Hong, J., Musachio, J. L., . . . Fujita,
M. (2007a). Kinetic evaluation in nonhuman primates of two new PET ligands for
peripheral benzodiazepine receptors in brain. Synapse, 61(8), 595-605.
Imaizumi, M., Kim, H. J., Zoghbi, S. S., Briard, E., Hong, J., Musachio, J. L., . . . Fujita, M.
(2007b). PET imaging with [11C]PBR28 can localize and quantify upregulated
peripheral benzodiazepine receptors associated with cerebral ischemia in rat.
Neuroscience Letters, 411(3), 200-205.
Imamura, K., Hishikawa, N., Sawada, M., Nagatsu, T., Yoshida, M., & Hashizume, Y.
(2003). Distribution of major histocompatibility complex class II-positive microglia and
cytokine profile of parkinson's disease brains. Acta Neuropathologica, 106(6), 518-526.
Lindqvist, D., Hall, S., Surova, Y., Henrietta, M., Nielsen, M., Janelidze, S…. Hansson.
(2013). Cerebrospinal fluid inflammatory markers in Parkinson's disease - associations
with depression, fatigue, and cognitive impairment. Brain, Behavior, and Immunity, 33,
183-189.
Innis, R. B., Cunningham, V. J., Delforge, J., Fujita, M., Gjedde, A., Gunn, R. N., . . .
Carson, R. E. (2007). Consensus nomenclature for in vivo imaging of reversibly binding
radioligands. Journal of Cerebral Blood Flow and Metabolism : Official Journal of the
International Society of Cerebral Blood Flow and Metabolism, 27(9), 1533-1539.
International Parkinson Disease Genomics Consortium, Nalls, M. A., Plagnol, V.,
Hernandez, D. G., Sharma, M., Sheerin, U. M., . . . Wood, N. W. (2011). Imputation of
sequence variants for identification of genetic risks for parkinson's disease: A meta-
analysis of genome-wide association studies. Lancet, 377(9766), 641-649.
Iwai, A., Masliah, E., Yoshimoto, M., Ge, N., Flanagan, L., de Silva, H. A., . . . Saitoh, T.
(1995). The precursor protein of non-A beta component of alzheimer's disease amyloid
is a presynaptic protein of the central nervous system. Neuron, 14(2), 467-475.
Iwata, K., & Harano, H. (1986). Regional cerebral blood flow changes in aging. Acta
Radiologica.Supplementum, 369, 440-443.
James, M. L., Fulton, R. R., Vercoullie, J., Henderson, D. J., Garreau, L., Chalon, S., . . .
Kassiou, M. (2008). DPA-714, a new translocator protein-specific ligand: Synthesis,
radiofluorination, and pharmacologic characterization. Journal of Nuclear Medicine :
Official Publication, Society of Nuclear Medicine, 49(5), 814-822.
James, M. L., Selleri, S., & Kassiou, M. (2006). Development of ligands for the peripheral
benzodiazepine receptor. Current Medicinal Chemistry, 13(17), 1991-2001.
Jankovic, J. (2008). Parkinson's disease: Clinical features and diagnosis. Journal of
Neurology, Neurosurgery, and Psychiatry, 79(4), 368-376.
166
Janvin, C. C., Larsen, J. P., Aarsland, D., & Hugdahl, K. (2006). Subtypes of mild cognitive
impairment in parkinson's disease: Progression to dementia. Movement Disorders :
Official Journal of the Movement Disorder Society, 21(9), 1343-1349.
Jellinger, K. A. (1991). Pathology of parkinson's disease. changes other than the nigrostriatal
pathway. Molecular and Chemical Neuropathology / Sponsored by the International
Society for Neurochemistry and the World Federation of Neurology and Research
Groups on Neurochemistry and Cerebrospinal Fluid, 14(3), 153-197.
Jellinger, K. A. (2012). Neurobiology of cognitive impairment in parkinson's disease. Expert
Review of Neurotherapeutics, 12(12), 1451-1466.
Jenkinson, M., Bannister, P., Brady, M., & Smith, S. (2002). Improved optimization for the
robust and accurate linear registration and motion correction of brain images.
NeuroImage, 17(2), 825-841.
Jenkinson, M., & Smith, S. (2001). A global optimisation method for robust affine
registration of brain images. Medical Image Analysis, 5(2), 143-156.
Johansen-Berg, H. & Rushworth, M.F. (2009). Using diffusion imaging to study human
connectional anatomy. Annual Review Neuroscience, 32, 75-94.
Jones, D. K., Knosche, T. R., & Turner, R. (2013). White matter integrity, fiber count, and
other fallacies: The do's and don'ts of diffusion MRI. NeuroImage, 73, 239-254.
Jubault, T., Brambati, S. M., Degroot, C., Kullmann, B., Strafella, A. P., Lafontaine, A. L., . .
. Monchi, O. (2009). Regional brain stem atrophy in idiopathic parkinson's disease
detected by anatomical MRI. PloS One, 4(12), e8247.
Jubault, T., Gagnon, J. F., Karama, S., Ptito, A., Lafontaine, A. L., Evans, A. C., & Monchi,
O. (2011). Patterns of cortical thickness and surface area in early parkinson's disease.
NeuroImage, 55(2), 462-467.
Kaiser, M., & Hilgetag, C. C. (2006). Nonoptimal component placement, but short
processing paths, due to long-distance projections in neural systems. PLoS
Computational Biology, 2(7), e95.
Kashani, A., Betancur, C., Giros, B., Hirsch, E., & El Mestikawy, S. (2007). Altered
expression of vesicular glutamate transporters VGLUT1 and VGLUT2 in parkinson
disease. Neurobiology of Aging, 28(4), 568-578.
Katsuse, O., Iseki, E., Marui, W., & Kosaka, K. (2003). Developmental stages of cortical
lewy bodies and their relation to axonal transport blockage in brains of patients with
dementia with lewy bodies. Journal of the Neurological Sciences, 211(1-2), 29-35.
167
Kehagia, A. A., Barker, R. A., & Robbins, T. W. (2010). Neuropsychological and clinical
heterogeneity of cognitive impairment and dementia in patients with parkinson's disease.
The Lancet.Neurology, 9(12), 1200-1213.
Kenk, M., Selvanathan, T., Rao, N., Suridjan, I., Rusjan, P., Remington, G., . . . Mizrahi, R.
(2015). Imaging neuroinflammation in gray and white matter in schizophrenia: An in-
vivo PET study with [18F]-FEPPA. Schizophrenia Bulletin, 41(1), 85-93.
Khandelwal, P. J., Herman, A. M., & Moussa, C. E. (2011). Inflammation in the early stages
of neurodegenerative pathology. Journal of Neuroimmunology, 238(1-2), 1-11.
Kim, H. J., Kim, S. J., Kim, H. S., Choi, C. G., Kim, N., Han, S., . . . Lee, C. S. (2013).
Alterations of mean diffusivity in brain white matter and deep gray matter in parkinson's
disease. Neuroscience Letters, 550, 64-68.
Kim, J. S., Yang, J. J., Lee, J. M., Youn, J., Kim, J. M., & Cho, J. W. (2014). Topographic
pattern of cortical thinning with consideration of motor laterality in parkinson disease.
Parkinsonism & Related Disorders, 20(11), 1186-1190.
Ko, J. H., Koshimori, Y., Mizrahi, R., Rusjan, P., Wilson, A. A., Lang, A. E., . . . Strafella,
A. P. (2013). Voxel-based imaging of translocator protein 18 kDa (TSPO) in high-
resolution PET. Journal of Cerebral Blood Flow and Metabolism : Official Journal of
the International Society of Cerebral Blood Flow and Metabolism, 33(3), 348-350.
Kobylecki, C., Counsell, S. J., Cabanel, N., Wachter, T., Turkheimer, F. E., Eggert, K., . . .
Gerhard, A. (2013). Diffusion-weighted imaging and its relationship to microglial
activation in parkinsonian syndromes. Parkinsonism & Related Disorders, 19(5), 527-
532.
Kofler, J., & Wiley, C. A. (2011). Microglia: Key innate immune cells of the brain.
Toxicologic Pathology, 39(1), 103-114.
Kordower, J. H., Chu, Y., Hauser, R. A., Freeman, T. B., & Olanow, C. W. (2008). Lewy
body-like pathology in long-term embryonic nigral transplants in parkinson's disease.
Nature Medicine, 14(5), 504-506.
Kordower, J. H., Chu, Y., Hauser, R. A., Olanow, C. W., & Freeman, T. B. (2008).
Transplanted dopaminergic neurons develop PD pathologic changes: A second case
report. Movement Disorders : Official Journal of the Movement Disorder Society,
23(16), 2303-2306.
Koshimori, Y., Ko, J. H., Mizrahi, R., Rusjan, P., Mabrouk, R., Jacobs, M. F., . . . Strafella,
A. P. (2015). Imaging striatal microglial activation in patients with parkinson's disease.
PloS One, 10(9), e0138721.
168
Koshimori, Y., Segura, B., Christopher, L., Lobaugh, N., Duff-Canning, S., Mizrahi, R., . . .
Strafella, A. P. (2015). Imaging changes associated with cognitive abnormalities in
parkinson's disease. Brain Structure & Function, 220(4), 2249-2261.
Kostic, V. S., Agosta, F., Petrovic, I., Galantucci, S., Spica, V., Jecmenica-Lukic, M., &
Filippi, M. (2010). Regional patterns of brain tissue loss associated with depression in
parkinson disease. Neurology, 75(10), 857-863.
Krack, P., Hariz, M. I., Baunez, C., Guridi, J., & Obeso, J. A. (2010). Deep brain stimulation:
From neurology to psychiatry? Trends in Neurosciences, 33(10), 474-484.
Kreisl, W. C., Fujita, M., Fujimura, Y., Kimura, N., Jenko, K. J., Kannan, P., . . . Innis, R. B.
(2010). Comparison of [(11)C]-(R)-PK 11195 and [(11)C]PBR28, two radioligands for
translocator protein (18 kDa) in human and monkey: Implications for positron emission
tomographic imaging of this inflammation biomarker. NeuroImage, 49(4), 2924-2932.
Kreisl, W. C., Jenko, K. J., Hines, C. S., Lyoo, C. H., Corona, W., Morse, C. L., . . .
Biomarkers Consortium PET Radioligand Project Team. (2013). A genetic
polymorphism for translocator protein 18 kDa affects both in vitro and in vivo
radioligand binding in human brain to this putative biomarker of neuroinflammation.
Journal of Cerebral Blood Flow and Metabolism : Official Journal of the International
Society of Cerebral Blood Flow and Metabolism, 33(1), 53-58.
Kropholler, M. A., Boellaard, R., Schuitemaker, A., Folkersma, H., van Berckel, B. N., &
Lammertsma, A. A. (2006). Evaluation of reference tissue models for the analysis of
[11C](R)-PK11195 studies. Journal of Cerebral Blood Flow and Metabolism : Official
Journal of the International Society of Cerebral Blood Flow and Metabolism, 26(11),
1431-1441.
Krystal, J. H., D'Souza, D. C., Mathalon, D., Perry, E., Belger, A., & Hoffman, R. (2003).
NMDA receptor antagonist effects, cortical glutamatergic function, and schizophrenia:
Toward a paradigm shift in medication development. Psychopharmacology, 169(3-4),
215-233.
Kuhlmann, A. C., & Guilarte, T. R. (2000). Cellular and subcellular localization of
peripheral benzodiazepine receptors after trimethyltin neurotoxicity. Journal of
Neurochemistry, 74(4), 1694-1704.
Kulisevsky, J., García-Sánchez, C., Berthier, M.L., Barbanoj, M., Pascual-Sedano, B.,
Gironell, A., & Estévez-González, A. (2000). Chronic effects of dopaminergic
replacement on cognitive function in Parkinson's disease: a two-year follow-up study
of previously untreated patients. Movement Disorders : Official Journal of the
Movement Disorder Society, 15(4), 613-626.
169
Kurth, F., Zilles, K., Fox, P. T., Laird, A. R., & Eickhoff, S. B. (2010). A link between the
systems: Functional differentiation and integration within the human insula revealed by
meta-analysis. Brain Structure & Function, 214(5-6), 519-534.
Lahiri, D. K., & Nurnberger, J. I.,Jr. (1991). A rapid non-enzymatic method for the
preparation of HMW DNA from blood for RFLP studies. Nucleic Acids Research,
19(19), 5444.
Lammertsma, A. A., & Hume, S. P. (1996). Simplified reference tissue model for PET
receptor studies. NeuroImage, 4(3 Pt 1), 153-158.
Lang, A. E., & Lozano, A. M. (1998). Parkinson's disease. first of two parts. The New
England Journal of Medicine, 339(15), 1044-1053.
Langston, J. W. (2006). The parkinson's complex: Parkinsonism is just the tip of the iceberg.
Annals of Neurology, 59(4), 591-596.
Langston, J. W., Widner, H., Goetz, C. G., Brooks, D., Fahn, S., Freeman, T., & Watts, R.
(1992). Core assessment program for intracerebral transplantations (CAPIT). Movement
Disorders : Official Journal of the Movement Disorder Society, 7(1), 2-13.
Lanoue, A. C., Dumitriu, A., Myers, R. H., & Soghomonian, J. J. (2010). Decreased glutamic
acid decarboxylase mRNA expression in prefrontal cortex in parkinson's disease.
Experimental Neurology, 226(1), 207-217.
Latora, V., & Marchiori, M. (2001). Efficient behavior of small-world networks. Physical
Review Letters, 87(19), 198701.
Lavisse, S., Garcia-Lorenzo, D., Peyronneau, M. A., Bodini, B., Thiriez, C., Kuhnast, B., . . .
Bottlaender, M. (2015). Optimized quantification of translocator protein radioligand
18F-DPA-714 uptake in the brain of genotyped healthy volunteers. Journal of Nuclear
Medicine : Official Publication, Society of Nuclear Medicine, 56(7), 1048-1054.
Lavisse, S., Guillermier, M., Herard, A. S., Petit, F., Delahaye, M., Van Camp, N., . . .
Escartin, C. (2012). Reactive astrocytes overexpress TSPO and are detected by TSPO
positron emission tomography imaging. The Journal of Neuroscience : The Official
Journal of the Society for Neuroscience, 32(32), 10809-10818.
Le Bihan, D. (2003). Looking into the functional architecture of the brain with diffusion
MRI. Nature Reviews.Neuroscience, 4(6), 469-480.
Le Bihan, D., Mangin, J. F., Poupon, C., Clark, C. A., Pappata, S., Molko, N., & Chabriat, H.
(2001). Diffusion tensor imaging: Concepts and applications. Journal of Magnetic
Resonance Imaging : JMRI, 13(4), 534-546.
170
Le Bihan, D., Poupon, C., Amadon, A., & Lethimonnier, F. (2006). Artifacts and pitfalls in
diffusion MRI. Journal of Magnetic Resonance Imaging : JMRI, 24(3), 478-488.
Lebedev, A. V., Westman, E., Simmons, A., Lebedeva, A., Siepel, F. J., Pereira, J. B., &
Aarsland, D. (2014). Large-scale resting state network correlates of cognitive
impairment in parkinson's disease and related dopaminergic deficits. Frontiers in
Systems Neuroscience, 8, 45.
Lee, H. J., Bae, E. J., & Lee, S. J. (2014). Extracellular alpha--synuclein-a novel and crucial
factor in lewy body diseases. Nature Reviews.Neurology, 10(2), 92-98.
Lee, H. J., Kim, C., & Lee, S. J. (2010). Alpha-synuclein stimulation of astrocytes: Potential
role for neuroinflammation and neuroprotection. Oxidative Medicine and Cellular
Longevity, 3(4), 283-287.
Lee, H. J., Suk, J. E., Bae, E. J., & Lee, S. J. (2008). Clearance and deposition of
extracellular alpha-synuclein aggregates in microglia. Biochemical and Biophysical
Research Communications, 372(3), 423-428.
Lee, J. Y., & Jeon, B. S. (2014). Maladaptive reward-learning and impulse control disorders
in patients with parkinson's disease: A clinical overview and pathophysiology update.
Journal of Movement Disorders, 7(2), 67-76.
Lee, S., Varvel, N. H., Konerth, M. E., Xu, G., Cardona, A. E., Ransohoff, R. M., & Lamb,
B. T. (2010). CX3CR1 deficiency alters microglial activation and reduces beta-amyloid
deposition in two alzheimer's disease mouse models. The American Journal of
Pathology, 177(5), 2549-2562.
Levy, R., & Goldman-Rakic, P. S. (2000). Segregation of working memory functions within
the dorsolateral prefrontal cortex. Experimental Brain Research.Experimentelle
Hirnforschung.Experimentation Cerebrale, 133(1), 23-32.
Lewis, D. A., Pierri, J. N., Volk, D. W., Melchitzky, D. S., & Woo, T. U. (1999). Altered
GABA neurotransmission and prefrontal cortical dysfunction in schizophrenia.
Biological Psychiatry, 46(5), 616-626.
Li, J. Y., Englund, E., Holton, J. L., Soulet, D., Hagell, P., Lees, A. J., . . . Brundin, P.
(2008). Lewy bodies in grafted neurons in subjects with parkinson's disease suggest
host-to-graft disease propagation. Nature Medicine, 14(5), 501-503.
Li, J. Y., Englund, E., Widner, H., Rehncrona, S., Bjorklund, A., Lindvall, O., & Brundin, P.
(2010). Characterization of lewy body pathology in 12- and 16-year-old intrastriatal
mesencephalic grafts surviving in a patient with parkinson's disease. Movement
Disorders : Official Journal of the Movement Disorder Society, 25(8), 1091-1096.
171
Li, W., Qin, W., Liu, H., Fan, L., Wang, J., Jiang, T., & Yu, C. (2013). Subregions of the
human superior frontal gyrus and their connections. NeuroImage, 78, 46-58.
Li, Y., Liang, P., Jia, X., & Li, K. (2016). Abnormal regional homogeneity in parkinson's
disease: A resting state fMRI study. Clinical Radiology, 71(1), e28-34.
Liang, X., Zou, Q., He, Y., & Yang, Y. (2013). Coupling of functional connectivity and
regional cerebral blood flow reveals a physiological basis for network hubs of the
human brain. Proceedings of the National Academy of Sciences of the United States of
America, 110(5), 1929-1934.
Litvan, I., Goldman, J. G., Troster, A. I., Schmand, B. A., Weintraub, D., Petersen, R. C., . . .
Emre, M. (2012). Diagnostic criteria for mild cognitive impairment in parkinson's
disease: Movement disorder society task force guidelines. Movement Disorders :
Official Journal of the Movement Disorder Society, 27(3), 349-356.
Liu, B., Le, K. X., Park, M. A., Wang, S., Belanger, A. P., Dubey, S., . . . Lemere, C. A.
(2015). In vivo detection of age- and disease-related increases in neuroinflammation by
18F-GE180 TSPO MicroPET imaging in wild-type and alzheimer's transgenic mice. The
Journal of Neuroscience : The Official Journal of the Society for Neuroscience, 35(47),
15716-15730.
Liu, W., Tang, Y., & Feng, J. (2011). Cross talk between activation of microglia and
astrocytes in pathological conditions in the central nervous system. Life Sciences, 89(5-
6), 141-146.
Logothetis, N. K., Pauls, J., Augath, M., Trinath, T., & Oeltermann, A. (2001).
Neurophysiological investigation of the basis of the fMRI signal. Nature, 412(6843),
150-157.
Lull, M. E., & Block, M. L. (2010). Microglial activation and chronic neurodegeneration.
Neurotherapeutics : The Journal of the American Society for Experimental
NeuroTherapeutics, 7(4), 354-365.
Luo, C. Y., Guo, X. Y., Song, W., Chen, Q., Cao, B., Yang, J., . . . Shang, H. F. (2015).
Functional connectome assessed using graph theory in drug-naive parkinson's disease.
Journal of Neurology, 262(6), 1557-1567.
Lyoo, C. H., Ryu, Y. H., & Lee, M. S. (2010). Topographical distribution of cerebral cortical
thinning in patients with mild parkinson's disease without dementia. Movement
Disorders : Official Journal of the Movement Disorder Society, 25(4), 496-499.
Lyoo, C. H., Ryu, Y. H., & Lee, M. S. (2011). Cerebral cortical areas in which thickness
correlates with severity of motor deficits of parkinson's disease. Journal of Neurology,
258(10), 1871-1876.
172
Ma, Q., Huang, B., Wang, J., Seger, C., Yang, W., Li, C., . . . Huang, R. (2016). Altered
modular organization of intrinsic brain functional networks in patients with parkinson's
disease. Brain Imaging and Behavior, Feb 10.
Mabrouk, R., Rusjan, P. M., Mizrahi, R., Jacobs, M. F., Koshimori, Y., Houle, S., . . .
Strafella, A. P. (2014). Image derived input function for [18F]-FEPPA: Application to
quantify translocator protein (18 kDa) in the human brain. PloS One, 9(12), e115768.
MacDonald, A. A., Monchi, O., Seergobin, K. N., Ganjavi, H., Tamjeedi, R., & MacDonald,
P. A. (2013). Parkinson's disease duration determines effect of dopaminergic therapy on
ventral striatum function. Movement Disorders : Official Journal of the Movement
Disorder Society, 28(2), 153-160.
Macuga, K. L., & Frey, S. H. (2011). Selective responses in right inferior frontal and
supramarginal gyri differentiate between observed movements of oneself vs. another.
Neuropsychologia, 49(5), 1202-1207.
Madden, D. J., Turkington, T. G., Provenzale, J. M., Denny, L. L., Hawk, T. C., Gottlob, L.
R., & Coleman, R. E. (1999). Adult age differences in the functional neuroanatomy of
verbal recognition memory. Human Brain Mapping, 7(2), 115-135.
Madhyastha, T. M., Askren, M. K., Zhang, J., Leverenz, J. B., Montine, T. J., & Grabowski,
T. J. (2015). Group comparison of spatiotemporal dynamics of intrinsic networks in
parkinson's disease. Brain : A Journal of Neurology, 138(Pt 9), 2672-2686.
Maeda, J., Suhara, T., Zhang, M. R., Okauchi, T., Yasuno, F., Ikoma, Y., . . . Suzuki, K.
(2004). Novel peripheral benzodiazepine receptor ligand [11C]DAA1106 for PET: An
imaging tool for glial cells in the brain. Synapse, 52(4), 283-291.
Malpass, K. (2011). MRI may predict the onset of alzheimer disease. Nature
Reviews.Neurology, 7(6), 302.
Matsui, H., Nishinaka, K., Oda, M., Niikawa, H., Komatsu, K., Kubori, T., & Udaka, F.
(2007). Depression in parkinson's disease. diffusion tensor imaging study. Journal of
Neurology, 254(9), 1170-1173.
Matsui, H., Nishinaka, K., Oda, M., Niikawa, H., Kubori, T., & Udaka, F. (2007). Dementia
in parkinson's disease: Diffusion tensor imaging. Acta Neurologica Scandinavica,
116(3), 177-181.
McColgan, P., Seunarine, K. K., Razi, A., Cole, J. H., Gregory, S., Durr, A., . . . Track-HD
Investigators. (2015). Selective vulnerability of rich club brain regions is an
organizational principle of structural connectivity loss in huntington's disease. Brain : A
Journal of Neurology, 138(Pt 11), 3327-3344.
173
McGeer, P. L., Itagaki, M. D., Boyes, B. E., & McGeer, E. G. (1988). Reactive microglia are
positiver for HLA-DR in the substantia nigra of parkinson's and alzheimer's disease
brains. Neurology, 38, 1285-1291.
Medran-Navarrete, V., Bernards, N., Kuhnast, B., Damont, A., Pottier, G., Peyronneau, M.
A., . . . Dolle, F. (2014). 18F]DPA-C5yne, a novel fluorine-18-labelled analogue of
DPA-714: Radiosynthesis and preliminary evaluation as a radiotracer for imaging
neuroinflammation with PET. Journal of Labelled Compounds &
Radiopharmaceuticals, 57(6), 410-418.
Melzer, T. R., Watts, R., MacAskill, M. R., Pitcher, T. L., Livingston, L., Keenan, R. J., . . .
Anderson, T. J. (2012). Grey matter atrophy in cognitively impaired parkinson's disease.
Journal of Neurology, Neurosurgery, and Psychiatry, 83(2), 188-194.
Melzer, T. R., Watts, R., Macaskill, M. R., Pitcher, T. L., Livingston, L., Keenan, R. J., . . .
Anderson, T. J. (2013). White matter microstructure deteriorates across cognitive stages
in parkinson disease. Neurology, 80(20), 1841-1849.
Menon, V., & Uddin, L. Q. (2010). Saliency, switching, attention and control: A network
model of insula function. Brain Structure & Function, 214(5-6), 655-667.
Messina, D., Cerasa, A., Condino, F., Arabia, G., Novellino, F., Nicoletti, G., . . . Quattrone,
A. (2011). Patterns of brain atrophy in parkinson's disease, progressive supranuclear
palsy and multiple system atrophy. Parkinsonism & Related Disorders, 17(3), 172-176.
Mesulam, M. M. (1998). From sensation to cognition. Brain : A Journal of Neurology, 121
(Pt 6), 1013-1052.
Meunier, D., Lambiotte, R., & Bullmore, E. T. (2010). Modular and hierarchically modular
organization of brain networks. Frontiers in Neuroscience, 4, 200.
Mirza, B., Hadberg, H., Thomsen, P., & Moos, T. (2000). The absence of reactive
astrocytosis is indicative of a unique inflammatory process in parkinson's disease.
Neuroscience, 95(2), 425-432.
Mittelbronn, M., Dietz, K., Schluesener, H. J., & Meyermann, R. (2001). Local distribution
of microglia in the normal adult human central nervous system differs by up to one order
of magnitude. Acta Neuropathologica, 101(3), 249-255.
Mizrahi, R., Rusjan, P. M., Kennedy, J., Pollock, B., Mulsant, B., Suridjan, I., . . . Houle, S.
(2012). Translocator protein (18 kDa) polymorphism (rs6971) explains in-vivo brain
binding affinity of the PET radioligand [(18)F]-FEPPA. Journal of Cerebral Blood Flow
and Metabolism : Official Journal of the International Society of Cerebral Blood Flow
and Metabolism, 32(6), 968-972.
174
Mogi, M., Harada, M., Kondo, T., Riederer, P., Inagaki, H., Minami, M., & Nagatsu, T.
(1994). Interleukin-1 beta, interleukin-6, epidermal growth factor and transforming
growth factor-alpha are elevated in the brain from parkinsonian patients. Neuroscience
Letters, 180(2), 147-150.
Mori, S., Wakana, S., Nagae-Poetscher, L. M., & van Zijl, P. C. M. (2005). MRI atlas of
human white matter. Amsterdam, The Netherlands: Elsevier.
Mori, S., Crain, B. J., Chacko, V. P., & van Zijl, P. C. (1999). Three-dimensional tracking of
axonal projections in the brain by magnetic resonance imaging. Annals of Neurology,
45(2), 265-269.
Morrison, J. H., & Hof, P. R. (1997). Life and death of neurons in the aging brain. Science,
278(5337), 412-419.
Mowinckel, A. M., Espeseth, T., & Westlye, L. T. (2012). Network-specific effects of age
and in-scanner subject motion: A resting-state fMRI study of 238 healthy adults.
NeuroImage, 63(3), 1364-1373.
Mukherjee, P., Berman, J. I., Chung, S. W., Hess, C. P., & Henry, R. G. (2008). Diffusion
tensor MR imaging and fiber tractography: Theoretic underpinnings. AJNR.American
Journal of Neuroradiology, 29(4), 632-641.
Mukherjee, P., Miller, J. H., Shimony, J. S., Conturo, T. E., Lee, B. C., Almli, C. R., &
McKinstry, R. C. (2001). Normal brain maturation during childhood: Developmental
trends characterized with diffusion-tensor MR imaging. Radiology, 221(2), 349-358.
Muller, T., Blum-Degen, D., Przuntek, H., & Kuhn, W. (1998). Interleukin-6 levels in
cerebrospinal fluid inversely correlate to severity of parkinson's disease. Acta
Neurologica Scandinavica, 98(2), 142-144.
Muzerengi, S., Contrafatto, D., & Chaudhuri, K. R. (2007). Non-motor symptoms:
Identification and management. Parkinsonism & Related Disorders, 13 Suppl 3, S450-6.
Mythri, R. B., Venkateshappa, C., Harish, G., Mahadevan, A., Muthane, U. B., Yasha, T. C.,
. . . Shankar, S. K. (2011). Evaluation of markers of oxidative stress, antioxidant
function and astrocytic proliferation in the striatum and frontal cortex of parkinson's
disease brains. Neurochemical Research, 36(8), 1452-1463.
Nagano-Saito, A., Habak, C., Mejia-Constain, B., Degroot, C., Monetta, L., Jubault, T., . . .
Monchi, O. (2013). Effect of mild cognitive impairment on the patterns of neural
activity in early parkinson's disease. Neurobiology of Aging, 35(1), 223-231.
Nagano-Saito, A., Martinu, K., & Monchi, O. (2014). Function of basal ganglia in bridging
cognitive and motor modules to perform an action. Frontiers in Neuroscience, 8, 187.
175
Nagatsu, T., & Sawada, M. (2005). Inflammatory process in parkinson's disease: Role for
cytokines. Current Pharmaceutical Design, 11(8), 999-1016.
Narendran, R., Lopresti, B. J., Mason, N. S., Deuitch, L., Paris, J., Himes, M. L., . . .
Nimgaonkar, V. L. (2014). Cocaine abuse in humans is not associated with increased
microglial activation: An 18-kDa translocator protein positron emission tomography
imaging study with [11C]PBR28. The Journal of Neuroscience : The Official Journal of
the Society for Neuroscience, 34(30), 9945-9950.
Nasreddine, Z. S., Phillips, N. A., Bedirian, V., Charbonneau, S., Whitehead, V., Collin, I., . .
. Chertkow, H. (2005). The montreal cognitive assessment, MoCA: A brief screening
tool for mild cognitive impairment. Journal of the American Geriatrics Society, 53(4),
695-699.
Neil, J. J. (2008). Diffusion imaging concepts for clinicians. Journal of Magnetic Resonance
Imaging : JMRI, 27(1), 1-7.
Newman, M. E. (2006). Modularity and community structure in networks. Proceedings of
the National Academy of Sciences of the United States of America, 103(23), 8577-8582.
Nichols, N. R., Day, J. R., Laping, N. J., Johnson, S. A., & Finch, C. E. (1993). GFAP
mRNA increases with age in rat and human brain. Neurobiology of Aging, 14(5), 421-
429.
Nigro, S., Riccelli, R., Passamonti, L., Arabia, G., Morelli, M., Nistico, R., …Quattrone, A.
(2016). Characterizing structural neural networks in de novo Parkinson Disease patients
using diffusion tensor imaging. Human Brain Mapping, in press.
Norden, D. M., & Godbout, J. P. (2013). Review: Microglia of the aged brain: Primed to be
activated and resistant to regulation. Neuropathology and Applied Neurobiology, 39(1),
19-34.
Nucifora, P. G., Verma, R., Lee, S. K., & Melhem, E. R. (2007). Diffusion-tensor MR
imaging and tractography: Exploring brain microstructure and connectivity. Radiology,
245(2), 367-384.
Oades, R. D., & Halliday, G. M. (1987). Ventral tegmental (A10) system: Neurobiology. 1.
anatomy and connectivity. Brain Research, 434(2), 117-165.
Obeso, J. A., & Lanciego, J. L. (2011). Past, present, and future of the pathophysiological
model of the basal ganglia. Frontiers in Neuroanatomy, 5, 39.
Ogawa, S., & Lee, T. M. (1990). Magnetic resonance imaging of blood vessels at high fields:
In vivo and in vitro measurements and image simulation. Magnetic Resonance in
Medicine, 16(1), 9-18.
176
Okubo, T., Yoshikawa, R., Chaki, S., Okuyama, S., & Nakazato, A. (2004). Design,
synthesis, and structure-activity relationships of novel tetracyclic compounds as
peripheral benzodiazepine receptor ligands. Bioorganic & Medicinal Chemistry, 12(13),
3569-3580.
Olde Dubbelink, K. T., Hillebrand, A., Stoffers, D., Deijen, J. B., Twisk, J. W., Stam, C. J.,
& Berendse, H. W. (2014). Disrupted brain network topology in parkinson's disease: A
longitudinal magnetoencephalography study. Brain : A Journal of Neurology, 137(Pt 1),
197-207.
Ouchi, Y., Yagi, S., Yokokura, M., & Sakamoto, M. (2009). Neuroinflammation in the living
brain of parkinson's disease. Parkinsonism & Related Disorders, 15 Suppl 3, S200-4.
Ouchi, Y., Yoshikawa, E., Sekine, Y., Futatsubashi, M., Kanno, T., Ogusu, T., & Torizuka,
T. (2005). Microglial activation and dopamine terminal loss in early parkinson's disease.
Annals of Neurology, 57(2), 168-175.
Owen, A. M. (2004). Cognitive dysfunction in parkinson's disease: The role of frontostriatal
circuitry. The Neuroscientist : A Review Journal Bringing Neurobiology, Neurology and
Psychiatry, 10(6), 525-537.
Owen, A. M., Stern, C. E., Look, R. B., Tracey, I., Rosen, B. R., & Petrides, M. (1998).
Functional organization of spatial and nonspatial working memory processing within the
human lateral frontal cortex. Proceedings of the National Academy of Sciences of the
United States of America, 95(13), 7721-7726.
Owen, D. R., Gunn, R. N., Rabiner, E. A., Bennacef, I., Fujita, M., Kreisl, W. C., . . . Parker,
C. A. (2011). Mixed-affinity binding in humans with 18-kDa translocator protein
ligands. Journal of Nuclear Medicine : Official Publication, Society of Nuclear
Medicine, 52(1), 24-32.
Owen, D. R., Guo, Q., Kalk, N. J., Colasanti, A., Kalogiannopoulou, D., Dimber, R., . . .
Rabiner, E. A. (2014). Determination of [(11)C]PBR28 binding potential in vivo: A first
human TSPO blocking study. Journal of Cerebral Blood Flow and Metabolism :
Official Journal of the International Society of Cerebral Blood Flow and Metabolism,
34(6), 989-994.
Owen, D. R., Howell, O. W., Tang, S. P., Wells, L. A., Bennacef, I., Bergstrom, M., . . .
Parker, C. A. (2010). Two binding sites for [3H]PBR28 in human brain: Implications for
TSPO PET imaging of neuroinflammation. Journal of Cerebral Blood Flow and
Metabolism : Official Journal of the International Society of Cerebral Blood Flow and
Metabolism, 30(9), 1608-1618.
Owen, D. R., Yeo, A. J., Gunn, R. N., Song, K., Wadsworth, G., Lewis, A., . . . Rubio, J. P.
(2012). An 18-kDa translocator protein (TSPO) polymorphism explains differences in
binding affinity of the PET radioligand PBR28. Journal of Cerebral Blood Flow and
177
Metabolism : Official Journal of the International Society of Cerebral Blood Flow and
Metabolism, 32(1), 1-5.
Pagonabarraga, J., & Kulisevsky, J. (2012). Cognitive impairment and dementia in
parkinson's disease. Neurobiology of Disease, 46(3), 590-596.
Pagonabarraga, J., Kulisevsky, J., Strafella, A. P., & Krack, P. (2015). Apathy in parkinson's
disease: Clinical features, neural substrates, diagnosis, and treatment. The
Lancet.Neurology, 14(5), 518-531.
Pan, P. L., Song, W., & Shang, H. F. (2012). Voxel-wise meta-analysis of gray matter
abnormalities in idiopathic parkinson's disease. European Journal of Neurology : The
Official Journal of the European Federation of Neurological Societies, 19(2), 199-206.
Paolicelli, R. C., Bolasco, G., Pagani, F., Maggi, L., Scianni, M., Panzanelli, P., . . . Gross, C.
T. (2011). Synaptic pruning by microglia is necessary for normal brain development.
Science (New York, N.Y.), 333(6048), 1456-1458.
Papadopoulos, V., Amri, H., Boujrad, N., Cascio, C., Culty, M., Garnier, M., . . . Drieu, K.
(1997). Peripheral benzodiazepine receptor in cholesterol transport and steroidogenesis.
Steroids, 62(1), 21-28.
Papadopoulos, V., Baraldi, M., Guilarte, T. R., Knudsen, T. B., Lacapere, J. J., Lindemann,
P., . . . Gavish, M. (2006). Translocator protein (18kDa): New nomenclature for the
peripheral-type benzodiazepine receptor based on its structure and molecular function.
Trends in Pharmacological Sciences, 27(8), 402-409.
Pappata, S., Cornu, P., Samson, Y., Prenant, C., Benavides, J., Scatton, B., . . . Syrota, A.
(1991). PET study of carbon-11-PK 11195 binding to peripheral type benzodiazepine
sites in glioblastoma: A case report. Journal of Nuclear Medicine : Official Publication,
Society of Nuclear Medicine, 32(8), 1608-1610.
Parente, A., Feltes, P. K., Vallez Garcia, D., Sijbesma, J. W., Moriguchi Jeckel, C. M.,
Dierckx, R. A., . . . Doorduin, J. (2016). Pharmacokinetic analysis of 11C-PBR28 in the
rat model of herpes encephalitis: Comparison with (R)-11C-PK11195. Journal of
Nuclear Medicine : Official Publication, Society of Nuclear Medicine, 57(5), 785-791.
Parker, G. J., & Alexander, D. C. (2005). Probabilistic anatomical connectivity derived from
the microscopic persistent angular structure of cerebral tissue. Philosophical
Transactions of the Royal Society of London.Series B, Biological Sciences, 360(1457),
893-902.
Passingham, R. E., Bengtsson, S. L., & Lau, H. C. (2010). Medial frontal cortex: From self-
generated action to reflection on one's own performance. Trends in Cognitive Sciences,
14(1), 16-21.
178
Pellicano, C., Assogna, F., Piras, F., Caltagirone, C., Pontieri, F. E., & Spalletta, G. (2012).
Regional cortical thickness and cognitive functions in non-demented parkinson's disease
patients: A pilot study. European Journal of Neurology : The Official Journal of the
European Federation of Neurological Societies, 19(1), 172-175.
Pereira, J. B., Aarsland, D., Ginestet, C. E., Lebedev, A.V., Wahlund L., Simmons, A., …
Westman, E. (2015). Aberrant cerebral network topology and mild cognitive impairment
in early Parkinson's disease. Human Brain Mapping, 36(8), 2980-2995.
Pereira, J. B., Ibarretxe-Bilbao, N., Marti, M. J., Compta, Y., Junque, C., Bargallo, N., &
Tolosa, E. (2012). Assessment of cortical degeneration in patients with parkinson's
disease by voxel-based morphometry, cortical folding, and cortical thickness. Human
Brain Mapping, 33(11), 2521-2534.
Petit-Taboue, M. C., Baron, J. C., Barre, L., Travere, J. M., Speckel, D., Camsonne, R., &
MacKenzie, E. T. (1991). Brain kinetics and specific binding of [11C]PK 11195 to
omega 3 sites in baboons: Positron emission tomography study. European Journal of
Pharmacology, 200(2-3), 347-351.
Picard, N., & Strick, P. L. (2001). Imaging the premotor areas. Current Opinion in
Neurobiology, 11(6), 663-672.
Pierpaoli, C., & Basser, P. J. (1996). Toward a quantitative assessment of diffusion
anisotropy. Magnetic Resonance in Medicine, 36(6), 893-906.
Pinto, S., Mancini, L., Jahanshahi, M., Thornton, J.S., Tripoliti, E., Yousry, T.A., Limousin,
P. (2010). Functional magnetic resonance imaging exploration of combined hand and
speech movements in Parkinson's disease. Movement Disorders : Official Journal of the
Movement Disorder Society, 26(12), 2212-2219.
Poletti, M., & Bonuccelli, U. (2013). Acute and chronic cognitive effects of levodopa and
dopamine agonists on patients with parkinson's disease: A review. Therapeutic
Advances in Psychopharmacology, 3(2), 101-113.
Polymeropoulos, M. H., Lavedan, C., Leroy, E., Ide, S. E., Dehejia, A., Dutra, A., . . .
Nussbaum, R. L. (1997). Mutation in the alpha-synuclein gene identified in families
with parkinson's disease. Science, 276(5321), 2045-2047.
Porchet, R., Probst, A., Bouras, C., Draberova, E., Draber, P., & Riederer, B. M. (2003).
Analysis of glial acidic fibrillary protein in the human entorhinal cortex during aging
and in alzheimer's disease. Proteomics, 3(8), 1476-1485.
Postle, B. R., Stern, C. E., Rosen, B. R., & Corkin, S. (2000). An fMRI investigation of
cortical contributions to spatial and nonspatial visual working memory. NeuroImage,
11(5 Pt 1), 409-423.
179
Power, J. D., Cohen, A. L., Nelson, S. M., Wig, G. S., Barnes, K. A., Church, J. A., . . .
Petersen, S. E. (2011). Functional network organization of the human brain. Neuron,
72(4), 665-678.
Power, J. D., & Petersen, S. E. (2013). Control-related systems in the human brain. Current
Opinion in Neurobiology, 23(2), 223-228.
Qian, L., & Flood, P. M. (2008). Microglial cells and parkinson's disease. Immunologic
Research, 41(3), 155-164.
Rae, C. L., Correia, M. M., Altena, E., Hughes, L. E., Barker, R. A., & Rowe, J. B. (2012).
White matter pathology in parkinson's disease: The effect of imaging protocol
differences and relevance to executive function. NeuroImage, 62(3), 1675-1684.
Raichle, M. E., & Mintun, M. A. (2006). Brain work and brain imaging. Annual Review of
Neuroscience, 29, 449-476.
Ramirez-Ruiz, B., Marti, M. J., Tolosa, E., Gimenez, M., Bargallo, N., Valldeoriola, F., &
Junque, C. (2007). Cerebral atrophy in parkinson's disease patients with visual
hallucinations. European Journal of Neurology : The Official Journal of the European
Federation of Neurological Societies, 14(7), 750-756.
Ramlackhansingh, A. F., Brooks, D. J., Greenwood, R. J., Bose, S. K., Turkheimer, F. E.,
Kinnunen, K. M., . . . Sharp, D. J. (2011). Inflammation after trauma: Microglial
activation and traumatic brain injury. Annals of Neurology, 70(3), 374-383.
Ransohoff, R. M., & Stevens, B. (2011). Neuroscience. how many cell types does it take to
wire a brain? Science, 333(6048), 1391-1392.
Rektorová, I., Rektor, I., Bares, M., Dostál, V., Ehler, E., Fanfrdlová, Z., … Velísková, J.
(2005). Cognitive performance in people with Parkinson's disease and mild or
moderate depression: effects of dopamine agonists in an add-on to L-dopa therapy.
European Journal of Neurology, 12(1):9-15.
Relja, M., & Klepac, N. (2006). A dopamine agonist, pramipexole, and cognitive functions
in Parkinson's disease. Neurological Sciences, 248(1-2), 251-254.
Riond, J., Mattei, M. G., Kaghad, M., Dumont, X., Guillemot, J. C., Le Fur, G., . . . Ferrara,
P. (1991). Molecular cloning and chromosomal localization of a human peripheral-type
benzodiazepine receptor. European Journal of Biochemistry / FEBS, 195(2), 305-311.
Rone, M. B., Midzak, A. S., Issop, L., Rammouz, G., Jagannathan, S., Fan, J., . . .
Papadopoulos, V. (2012). Identification of a dynamic mitochondrial protein complex
driving cholesterol import, trafficking, and metabolism to steroid hormones. Molecular
Endocrinology, 26(11), 1868-1882.
180
Rosenberg-Katz, K., Herman, T., Jacob, Y., Giladi, N., Hendler, T., & Hausdorff, J. M.
(2013). Gray matter atrophy distinguishes between parkinson disease motor subtypes.
Neurology, 80(16), 1476-1484.
Rousset, O. G., Ma, Y., & Evans, A. C. (1998). Correction for partial volume effects in PET:
Principle and validation. Journal of Nuclear Medicine : Official Publication, Society of
Nuclear Medicine, 39(5), 904-911.
Rubinov, M., & Sporns, O. (2010). Complex network measures of brain connectivity: Uses
and interpretations. NeuroImage, 52(3), 1059-1069.
Rupprecht, R., Papadopoulos, V., Rammes, G., Baghai, T. C., Fan, J., Akula, N., . . .
Schumacher, M. (2010). Translocator protein (18 kDa) (TSPO) as a therapeutic target
for neurological and psychiatric disorders. Nature Reviews.Drug Discovery, 9(12), 971-
988.
Rupprecht, R., Rammes, G., Eser, D., Baghai, T. C., Schule, C., Nothdurfter, C., . . . Kucher,
K. (2009). Translocator protein (18 kD) as target for anxiolytics without
benzodiazepine-like side effects. Science, 325(5939), 490-493.
Rusjan, P., Mamo, D., Ginovart, N., Hussey, D., Vitcu, I., Yasuno, F., . . . Kapur, S. (2006).
An automated method for the extraction of regional data from PET images. Psychiatry
Research, 147(1), 79-89.
Rusjan, P. M., Wilson, A. A., Bloomfield, P. M., Vitcu, I., Meyer, J. H., Houle, S., &
Mizrahi, R. (2011). Quantitation of translocator protein binding in human brain with the
novel radioligand [18F]-FEPPA and positron emission tomography. Journal of Cerebral
Blood Flow and Metabolism : Official Journal of the International Society of Cerebral
Blood Flow and Metabolism, 31(8), 1807-1816.
Rusjan, P. M., Wilson, A. A., Miler, L., Fan, I., Mizrahi, R., Houle, S., . . . Meyer, J. H.
(2014). Kinetic modeling of the monoamine oxidase B radioligand [11C]SL25.1188 in
human brain with high-resolution positron emission tomography. Journal of Cerebral
Blood Flow and Metabolism : Official Journal of the International Society of Cerebral
Blood Flow and Metabolism, 34(5), 883-889.
Ryterska, A., Jahanshahi, M., & Osman, M. (2013). What are people with parkinson's
disease really impaired on when it comes to making decisions? A meta-analysis of the
evidence. Neuroscience and Biobehavioral Reviews, 37(10 Pt 2), 2836-2846.
Sabatini, U., Boulanouar, K., Fabre, N., Martin, F., Carel, C., Colonnese, C., . . . Rascol, O.
(2000). Cortical motor reorganization in akinetic patients with parkinson's disease: A
functional MRI study. Brain : A Journal of Neurology, 123(Pt 2), 394-403.
Saijo, K., & Glass, C. K. (2011). Microglial cell origin and phenotypes in health and disease.
Nature Reviews.Immunology, 11(11), 775-787.
181
Salinas, C. A., Searle, G. E., & Gunn, R. N. (2015). The simplified reference tissue model:
Model assumption violations and their impact on binding potential. Journal of Cerebral
Blood Flow and Metabolism : Official Journal of the International Society of Cerebral
Blood Flow and Metabolism, 35(2), 304-311.
Sang, L., Zhang, J., Wang, L., Zhang, J., Zhang, Y., Li, P., . . . Qiu, M. (2015). Alteration of
brain functional networks in early-stage parkinson's disease: A resting-state fMRI study.
PloS One, 10(10), e0141815.
Saxena, S., & Caroni, P. (2011). Selective neuronal vulnerability in neurodegenerative
diseases: From stressor thresholds to degeneration. Neuron, 71(1), 35-48.
Schapira, A. H., & Tolosa, E. (2010). Molecular and clinical prodrome of parkinson disease:
Implications for treatment. Nature Reviews.Neurology, 6(6), 309-317.
Schuitemaker, A., Van Berckel, B. N., Kropholler, M., Boellaard, R., Jonker, C., Scheltens,
P., & Lammertsma, A. A. (2004). Microglia activation in mild cognitive impairment.
Neurobiol. Aging, 25(Suppl. 2), S286.
Schuitemaker, A., VanBerckel, B. N., Boellaard, R., Kropholler, M., Boellaard, R., Jonker,
C., . . . Lammertsma, A. A. (2006). Assessment of microglial activation in mild
cognitive impairment using [11C](R)-PK11195 and PET. NeuroImage, 31, T159.
Schumacher, M., Akwa, Y., Guennoun, R., Robert, F., Labombarda, F., Desarnaud, F., . . .
Baulieu, E. E. (2000). Steroid synthesis and metabolism in the nervous system: Trophic
and protective effects. Journal of Neurocytology, 29(5-6), 307-326.
Schwarz, G. E. (1978). Estimating the dimension of a model. Annals of Statistics, 6(2),
10789-10795.
Scott, G., Hellyer, P. J., Ramlackhansingh, A. F., Brooks, D. J., Matthews, P. M., & Sharp,
D. J. (2015). Thalamic inflammation after brain trauma is associated with thalamo-
cortical white matter damage. Journal of Neuroinflammation, 12, 224.
Seeley, W. W., Crawford, R. K., Zhou, J., Miller, B. L., & Greicius, M. D. (2009).
Neurodegenerative diseases target large-scale human brain networks. Neuron, 62(1), 42-
52.
Seeley, W. W., Menon, V., Schatzberg, A. F., Keller, J., Glover, G. H., Kenna, H., . . .
Greicius, M. D. (2007). Dissociable intrinsic connectivity networks for salience
processing and executive control. The Journal of Neuroscience : The Official Journal of
the Society for Neuroscience, 27(9), 2349-2356.
Segonne, F., Pacheco, J., & Fischl, B. (2007). Geometrically accurate topology-correction of
cortical surfaces using nonseparating loops. IEEE Transactions on Medical Imaging,
26(4), 518-529.
182
Segura, B., Baggio, H. C., Marti, M. J., Valldeoriola, F., Compta, Y., Garcia-Diaz, A. I., . . .
Junque, C. (2014). Cortical thinning associated with mild cognitive impairment in
parkinson's disease. Movement Disorders : Official Journal of the Movement Disorder
Society, 29(12), 1495-1503.
Selden, N. R., Gitelman, D. R., Salamon-Murayama, N., Parrish, T. B., & Mesulam, M. M.
(1998). Trajectories of cholinergic pathways within the cerebral hemispheres of the
human brain. Brain : A Journal of Neurology, 12 (Pt 12), 2249-2257.
Setiawan, E., Wilson, A. A., Mizrahi, R., Rusjan, P. M., Miler, L., Rajkowska, G., . . .
Meyer, J. H. (2015). Role of translocator protein density, a marker of
neuroinflammation, in the brain during major depressive episodes. JAMA Psychiatry,
72(3), 268-275.
Shenton, M. E., Hamoda, H. M., Schneiderman, J. S., Bouix, S., Pasternak, O., Rathi, Y., . . .
Zafonte, R. (2012). A review of magnetic resonance imaging and diffusion tensor
imaging findings in mild traumatic brain injury. Brain Imaging and Behavior, 6(2), 137-
192.
Shin, N.Y., Shin, Y.S., Lee, P.H., Yoon, U., Han, S., Kim, D.J., & Lee, S.K. (2016).
Different Functional and Microstructural Changes Depending on Duration of Mild
Cognitive Impairment in Parkinson Disease. AJNR Ameican Journal of
Neuroradiology, 37(5), 897-903.
Skidmore, F., Korenkevych, D., Liu, Y., He, G., Bullmore, E., & Pardalos, P. M. (2011).
Connectivity brain networks based on wavelet correlation analysis in parkinson fMRI
data. Neuroscience Letters, 499(1), 47-51.
Sled, J. G., Zijdenbos, A. P., & Evans, A. C. (1998). A nonparametric method for automatic
correction of intensity nonuniformity in MRI data. IEEE Transactions on Medical
Imaging, 17(1), 87-97.
Smith, J. A., Das, A., Ray, S. K., & Banik, N. L. (2012). Role of pro-inflammatory cytokines
released from microglia in neurodegenerative diseases. Brain Research Bulletin, 87(1),
10-20.
Smith, S. M. (2002). Fast robust automated brain extraction. Human Brain Mapping, 17(3),
143-155.
Smith, S. M., Jenkinson, M., Johansen-Berg, H., Rueckert, D., Nichols, T. E., Mackay, C. E.,
. . . Behrens, T. E. (2006). Tract-based spatial statistics: Voxelwise analysis of multi-
subject diffusion data. NeuroImage, 31(4), 1487-1505.
Smith, S. M., Jenkinson, M., Woolrich, M. W., Beckmann, C. F., Behrens, T. E., Johansen-
Berg, H., . . . Matthews, P. M. (2004). Advances in functional and structural MR image
analysis and implementation as FSL. NeuroImage, 23 Suppl 1, S208-19.
183
Smith, S. M., & Nichols, T. E. (2009). Threshold-free cluster enhancement: Addressing
problems of smoothing, threshold dependence and localisation in cluster inference.
NeuroImage, 44(1), 83-98.
Sofroniew, M. V., & Vinters, H. V. (2010). Astrocytes: Biology and pathology. Acta
Neuropathologica, 119(1), 7-35.
Song, S. K., Lee, J. E., Park, H. J., Sohn, Y. H., Lee, J. D., & Lee, P. H. (2011). The pattern
of cortical atrophy in patients with parkinson's disease according to cognitive status.
Movement Disorders : Official Journal of the Movement Disorder Society, 26(2), 289-
296.
Spillantini, M. G., Crowther, R. A., Jakes, R., Hasegawa, M., & Goedert, M. (1998). Alpha-
synuclein in filamentous inclusions of lewy bodies from parkinson's disease and
dementia with lewy bodies. Proceedings of the National Academy of Sciences of the
United States of America, 95(11), 6469-6473.
Spillantini, M. G., Schmidt, M. L., Lee, V. M., Trojanowski, J. Q., Jakes, R., & Goedert, M.
(1997). Alpha-synuclein in lewy bodies. Nature, 388(6645), 839-840.
Sporns, O. (2013). Network attributes for segregation and integration in the human brain.
Current Opinion in Neurobiology, 23(2), 162-171.
Sporns, O. (2014). Contributions and challenges for network models in cognitive
neuroscience. Nature Neuroscience, 17(5), 652-660.
Sporns, O., Honey, C. J., & Kotter, R. (2007). Identification and classification of hubs in
brain networks. PloS One, 2(10), e1049.
Spreng, R. N., Sepulcre, J., Turner, G. R., Stevens, W. D., & Schacter, D. L. (2013). Intrinsic
architecture underlying the relations among the default, dorsal attention, and
frontoparietal control networks of the human brain. Journal of Cognitive Neuroscience,
25(1), 74-86.
Stam, C. J. (2014). Modern network science of neurological disorders. Nature
Reviews.Neuroscience, 15(10), 683-695.
Stam, C. J., de Haan, W., Daffertshofer, A., Jones, B. F., Manshanden, I., van Cappellen van
Walsum, A. M., . . . Scheltens, P. (2009). Graph theoretical analysis of
magnetoencephalographic functional connectivity in alzheimer's disease. Brain : A
Journal of Neurology, 132(Pt 1), 213-224.
Stoessl, A. J., Lehericy, S., & Strafella, A. P. (2014). Imaging insights into basal ganglia
function, parkinson's disease, and dystonia. Lancet, 384(9942), 532-544.
184
Stone, D. K., Reynolds, A. D., Mosley, R. L., & Gendelman, H. E. (2009). Innate and
adaptive immunity for the pathobiology of parkinson's disease. Antioxidants & Redox
Signaling, 11(9), 2151-2166.
Streit, W. J. (2006). Microglial senescence: Does the brain's immune system have an
expiration date? Trends in Neurosciences, 29(9), 506-510.
Streit, W. J., Sammons, N. W., Kuhns, A. J., & Sparks, D. L. (2004). Dystrophic microglia in
the aging human brain. Glia, 45(2), 208-212.
Streit, W. J., & Xue, Q. S. (2010). The brain's aging immune system. Aging and Disease,
1(3), 254-261.
Sullivan, E. V., & Pfefferbaum, A. (2006). Diffusion tensor imaging and aging.
Neuroscience and Biobehavioral Reviews, 30(6), 749-761.
Suridjan, I., Pollock, B. G., Verhoeff, N. P., Voineskos, A. N., Chow, T., Rusjan, P. M., . . .
Mizrahi, R. (2015). In-vivo imaging of grey and white matter neuroinflammation in
alzheimer's disease: A positron emission tomography study with a novel radioligand,
[18F]-FEPPA. Molecular Psychiatry, 20(12), 1579-1587.
Suridjan, I., Rusjan, P. M., Kenk, M., Verhoeff, N. P., Voineskos, A. N., Rotenberg, D., . . .
Mizrahi, R. (2014). Quantitative imaging of neuroinflammation in human white matter:
A positron emission tomography study with translocator protein 18 kDa radioligand,
[18F]-FEPPA. Synapse (New York, N.Y.), 68(11), 536-547.
Suridjan, I., Rusjan, P. M., Voineskos, A. N., Selvanathan, T., Setiawan, E., Strafella, A. P., .
. . Mizrahi, R. (2014). Neuroinflammation in healthy aging: A PET study using a novel
translocator protein 18kDa (TSPO) radioligand, [(18)F]-FEPPA. NeuroImage, 84, 868-
875.
Tahmasian, M., Bettray, L.M., van Eimeren, T., Drzezga, A., Timmermann, L., Eickhoff, C.
R., … Eggers, C. (2015). A systematic review on the applications of resting-state fMRI
in Parkinson's disease: Does dopamine replacement therapy play a role? Cortex, 73, 80-
105.
Takahashi, K., Yamaguchi, S., Kobayashi, S., & Yamamoto, Y. (2005). Effects of aging on
regional cerebral blood flow assessed by using technetium tc 99m
hexamethylpropyleneamine oxime single-photon emission tomography with 3D
stereotactic surface projection analysis. AJNR.American Journal of Neuroradiology,
26(8), 2005-2009.
Takano, A., Piehl, F., Hillert, J., Varrone, A., Nag, S., Gulyas, B., . . . Halldin, C. (2013). In
vivo TSPO imaging in patients with multiple sclerosis: A brain PET study with
[18F]FEDAA1106. EJNMMI Research, 3(1), 30-219X-3-30.
185
Tang, D., Nickels, M. L., Tantawy, M. N., Buck, J. R., & Manning, H. C. (2014). Preclinical
imaging evaluation of novel TSPO-PET ligand 2-(5,7-diethyl-2-(4-(2-
[(18)F]fluoroethoxy)phenyl)pyrazolo[1,5-a]pyrimidin-3-yl)- N,N-diethylacetamide ([
(18)F]VUIIS1008) in glioma. Molecular Imaging and Biology : MIB : The Official
Publication of the Academy of Molecular Imaging, 16(6), 813-820.
Tessitore, A., Amboni, M., Cirillo, G., Corbo, D., Picillo, M., Russo, A., . . . Tedeschi, G.
(2012). Regional gray matter atrophy in patients with parkinson disease and freezing of
gait. AJNR.American Journal of Neuroradiology, 33(9), 1804-1809.
Tessitore, A., Santangelo, G., De Micco, R., Vitale, C., Giordano, A., Raimo, S., …
Tedeschi, G. (2016). Cortical thickness changes in patients with Parkinson's disease and
impulse control disorders. Parkinsonism & Related Disorders, 24, 119-126.
Theilmann, R. J., Reed, J. D., Song, D. D., Huang, M. X., Lee, R. R., Litvan, I., &
Harrington, D. L. (2013). White-matter changes correlate with cognitive functioning in
parkinson's disease. Frontiers in Neurology, 4, 37.
Thiel, A., Radlinska, B. A., Paquette, C., Sidel, M., Soucy, J. P., Schirrmacher, R., & Minuk,
J. (2010). The temporal dynamics of poststroke neuroinflammation: A longitudinal
diffusion tensor imaging-guided PET study with 11C-PK11195 in acute subcortical
stroke. Journal of Nuclear Medicine : Official Publication, Society of Nuclear Medicine,
51(9), 1404-1412.
Tinaz, S., Courtney, M. G., & Stern, C. E. (2011). Focal cortical and subcortical atrophy in
early parkinson's disease. Movement Disorders : Official Journal of the Movement
Disorder Society, 26(3), 436-441.
Tinaz, S., Lauro, P., Hallett, M., & Horovitz, S. G. (2016). Deficits in task-set maintenance
and execution networks in parkinson's disease. Brain Structure & Function, 221(3),
1413-1425.
Tison, F., Dartigues, J. F., Auriacombe, S., Letenneur, L., Boller, F., & Alperovitch, A.
(1995). Dementia in parkinson's disease: A population-based study in ambulatory and
institutionalized individuals. Neurology, 45(4), 705-708.
Tomasi, D., Wang, G. J., & Volkow, N. D. (2013). Energetic cost of brain functional
connectivity. Proceedings of the National Academy of Sciences of the United States of
America, 110(33), 13642-13647.
Tomlinson, C. L., Stowe, R., Patel, S., Rick, C., Gray, R., & Clarke, C. E. (2010). Systematic
review of levodopa dose equivalency reporting in parkinson's disease. Movement
Disorders : Official Journal of the Movement Disorder Society, 25(15), 2649-2653.
186
Tononi, G., Sporns, O., & Edelman, G. M. (1994). A measure for brain complexity: Relating
functional segregation and integration in the nervous system. Proceedings of the
National Academy of Sciences of the United States of America, 91(11), 5033-5037.
Trujillo, J. P., Gerrits, N. J., Vriend, C., Berendse, H. W., van den Heuvel, O. A., & van der
Werf, Y. D. (2015). Impaired planning in parkinson's disease is reflected by reduced
brain activation and connectivity. Human Brain Mapping, 36(9), 3703-3715.
Tufekci, K. U., Meuwissen, R., Genc, S., & Genc, K. (2012). Inflammation in parkinson's
disease. Advances in Protein Chemistry and Structural Biology, 88, 69-132.
Turner, M. R., Cagnin, A., Turkheimer, F. E., Miller, C. C., Shaw, C. E., Brooks, D. J., . . .
Banati, R. B. (2004). Evidence of widespread cerebral microglial activation in
amyotrophic lateral sclerosis: An [11C](R)-PK11195 positron emission tomography
study. Neurobiology of Disease, 15(3), 601-609.
Uribe, C., Segura, B., Baggio, H. C., Abos, A., Marti, M. J., Valldeoriola, F., . . . Junque, C.
(2016). Patterns of cortical thinning in nondemented parkinson's disease patients.
Movement Disorders : Official Journal of the Movement Disorder Society, 31(5), 699-
708.
Uversky, V. N. (2007). Neuropathology, biochemistry, and biophysics of alpha-synuclein
aggregation. Journal of Neurochemistry, 103(1), 17-37.
Vaishnavi, S. N., Vlassenko, A. G., Rundle, M. M., Snyder, A. Z., Mintun, M. A., & Raichle,
M. E. (2010). Regional aerobic glycolysis in the human brain. Proceedings of the
National Academy of Sciences of the United States of America, 107(41), 17757-17762.
Van Camp, N., Boisgard, R., Kuhnast, B., Theze, B., Viel, T., Gregoire, M. C., . . . Tavitian,
B. (2010). In vivo imaging of neuroinflammation: A comparative study between
[(18)F]PBR111, [ (11)C]CLINME and [ (11)C]PK11195 in an acute rodent model.
European Journal of Nuclear Medicine and Molecular Imaging, 37(5), 962-972.
van den Heuvel, M., Mandl, R., Luigjes, J., & Hulshoff Pol, H. (2008). Microstructural
organization of the cingulum tract and the level of default mode functional connectivity.
The Journal of Neuroscience : The Official Journal of the Society for Neuroscience,
28(43), 10844-10851.
van den Heuvel, M. P., & Hulshoff Pol, H. E. (2010). Exploring the brain network: A review
on resting-state fMRI functional connectivity. European Neuropsychopharmacology :
The Journal of the European College of Neuropsychopharmacology, 20(8), 519-534.
van den Heuvel, M. P., & Sporns, O. (2013). Network hubs in the human brain. Trends in
Cognitive Sciences, 17(12), 683-696.
187
Varrone, A., Mattsson, P., Forsberg, A., Takano, A., Nag, S., Gulyas, B., . . . Halldin, C.
(2013). In vivo imaging of the 18-kDa translocator protein (TSPO) with
[18F]FEDAA1106 and PET does not show increased binding in alzheimer's disease
patients. European Journal of Nuclear Medicine and Molecular Imaging, 40(6), 921-
931.
Varrone, A., Oikonen, V., Forsberg, A., Joutsa, J., Takano, A., Solin, O., . . . Rinne, J. O.
(2015). Positron emission tomography imaging of the 18-kDa translocator protein
(TSPO) with [18F]FEMPA in alzheimer's disease patients and control subjects.
European Journal of Nuclear Medicine and Molecular Imaging, 42(3), 438-446.
Venneti, S., Lopresti, B. J., Wang, G., Slagel, S. L., Mason, N. S., Mathis, C. A., . . . Wiley,
C. A. (2007). A comparison of the high-affinity peripheral benzodiazepine receptor
ligands DAA1106 and (R)-PK11195 in rat models of neuroinflammation: Implications
for PET imaging of microglial activation. Journal of Neurochemistry, 102(6), 2118-
2131.
Venneti, S., Lopresti, B. J., & Wiley, C. A. (2013). Molecular imaging of
microglia/macrophages in the brain. Glia, 61(1), 10-23.
Venneti, S., Wagner, A. K., Wang, G., Slagel, S. L., Chen, X., Lopresti, B. J., . . . Wiley, C.
A. (2007). The high affinity peripheral benzodiazepine receptor ligand DAA1106 binds
specifically to microglia in a rat model of traumatic brain injury: Implications for PET
imaging. Experimental Neurology, 207(1), 118-127.
Venneti, S., Wang, G., Nguyen, J., & Wiley, C. A. (2008). The positron emission
tomography ligand DAA1106 binds with high affinity to activated microglia in human
neurological disorders. Journal of Neuropathology and Experimental Neurology,
67(10), 1001-1010.
Venneti, S., Wang, G., & Wiley, C. A. (2008). The high affinity peripheral benzodiazepine
receptor ligand DAA1106 binds to activated and infected brain macrophages in areas of
synaptic degeneration: Implications for PET imaging of neuroinflammation in lentiviral
encephalitis. Neurobiology of Disease, 29(2), 232-241.
Venneti, S., Wiley, C. A., & Kofler, J. (2009). Imaging microglial activation during
neuroinflammation and alzheimer's disease. Journal of Neuroimmune Pharmacology :
The Official Journal of the Society on NeuroImmune Pharmacology, 4(2), 227-243.
Vivash, L., & O'Brien, T. J. (2016). Imaging microglial activation with TSPO PET: Lighting
up neurologic diseases? Journal of Nuclear Medicine : Official Publication, Society of
Nuclear Medicine, 57(2), 165-168.
Volpicelli-Daley, L. A., Luk, K. C., Patel, T. P., Tanik, S. A., Riddle, D. M., Stieber, A., . . .
Lee, V. M. (2011). Exogenous alpha-synuclein fibrils induce lewy body pathology
leading to synaptic dysfunction and neuron death. Neuron, 72(1), 57-71.
188
von Bernhardi, R., Tichauer, J. E., & Eugenin, J. (2010). Aging-dependent changes of
microglial cells and their relevance for neurodegenerative disorders. Journal of
Neurochemistry, 112(5), 1099-1114.
Vroon, A., Drukarch, B., Bol, J. G., Cras, P., Breve, J. J., Allan, S. M., . . . Van Dam, A. M.
(2007). Neuroinflammation in parkinson's patients and MPTP-treated mice is not
restricted to the nigrostriatal system: Microgliosis and differential expression of
interleukin-1 receptors in the olfactory bulb. Experimental Gerontology, 42(8), 762-771.
Wadsworth, H., Jones, P. A., Chau, W. F., Durrant, C., Fouladi, N., Passmore, J., . . . Trigg,
W. (2012). (1)(8)F]GE-180: A novel fluorine-18 labelled PET tracer for imaging
translocator protein 18 kDa (TSPO). Bioorganic & Medicinal Chemistry Letters, 22(3),
1308-1313.
Wakana, S., Caprihan, A., Panzenboeck, M. M., Fallon, J. H., Perry, M., Gollub, R. L., . . .
Mori, S. (2007). Reproducibility of quantitative tractography methods applied to
cerebral white matter. NeuroImage, 36(3), 630-644.
Wang, Y., Shima, K., Sawamura, H., & Tanji, J. (2001). Spatial distribution of cingulate
cells projecting to the primary, supplementary, and pre-supplementary motor areas: A
retrograde multiple labeling study in the macaque monkey. Neuroscience Research,
39(1), 39-49.
Wattendorf, E., Welge-Lussen, A., Fiedler, K., Bilecen, D., Wolfensberger, M., Fuhr, P., . . .
Westermann, B. (2009). Olfactory impairment predicts brain atrophy in parkinson's
disease. The Journal of Neuroscience : The Official Journal of the Society for
Neuroscience, 29(49), 15410-15413.
Watts, D. J., & Strogatz, S. H. (1998). Collective dynamics of 'small-world' networks.
Nature, 393(6684), 440-442.
Watts, R., Liston, C., Niogi, S., & Ulug, A. M. (2003). Fiber tracking using magnetic
resonance diffusion tensor imaging and its applications to human brain development.
Mental Retardation and Developmental Disabilities Research Reviews, 9(3), 168-177.
Wechsler, D. (1997). Manual for the wechsler memeory scale-3rd edition. New York: The
Psychological Corporation.
Wei, L., Zhang, J., Long, Z., Wu, G. R., Hu, X., Zhang, Y., & Wang, J. (2014). Reduced
topological efficiency in cortical-basal ganglia motor network of parkinson's disease: A
resting state fMRI study. PloS One, 9(10), e108124.
Weingarten, C. P., Sundman, M. H., Hickey, P., & Chen, N. K. (2015). Neuroimaging of
parkinson's disease: Expanding views. Neuroscience and Biobehavioral Reviews, 59,
16-52.
189
Weintraub, D., Doshi, J., Koka, D., Davatzikos, C., Siderowf, A. D., Duda, J. E., . . . Clark,
C. M. (2011a). Neurodegeneration across stages of cognitive decline in parkinson
disease. Archives of Neurology, 68(12), 1562-1568.
Werring, D. J., Clark, C. A., Barker, G. J., Thompson, A. J., & Miller, D. H. (1999).
Diffusion tensor imaging of lesions and normal-appearing white matter in multiple
sclerosis. Neurology, 52(8), 1626-1632.
Whitfield-Gabrieli, S., & Nieto-Castanon, A. (2012). Conn: A functional connectivity
toolbox for correlated and anticorrelated brain networks. Brain Connectivity, 2(3), 125-
141.
Wicklund, A. H., Johnson, N., & Weintraub, S. (2004). Preservation of reasoning in primary
progressive aphasia: Further differentiation from alzheimer's disease and the behavioral
presentation of frontotemporal dementia. Journal of Clinical and Experimental
Neuropsychology, 26(3), 347-355.
Wienhard, K., Schmand, M., Casey, M. E., Baker, K., Bao, J., Eriksson, L., . . . Nutt, R.
(2002). The ECAT HRRT: Performance and first clinical application of the new high
resolution research tomograph. IEEE Trans Nucl Sci, 49, 104-110.
Williams-Gray, C. H., Foltynie, T., Brayne, C. E., Robbins, T. W., & Barker, R. A. (2007).
Evolution of cognitive dysfunction in an incident parkinson's disease cohort. Brain : A
Journal of Neurology, 130(Pt 7), 1787-1798.
Wilson, A. A., Garcia, A., Parkes, J., McCormick, P., Stephenson, K. A., Houle, S., &
Vasdev, N. (2008). Radiosynthesis and initial evaluation of [18F]-FEPPA for PET
imaging of peripheral benzodiazepine receptors. Nuclear Medicine and Biology, 35(3),
305-314.
Wiltshire, K., Concha, L., Gee, M., Bouchard, T., Beaulieu, C., & Camicioli, R. (2010).
Corpus callosum and cingulum tractography in parkinson's disease. The Canadian
Journal of Neurological Sciences.Le Journal Canadien Des Sciences Neurologiques,
37(5), 595-600.
Winkeler, A., Boisgard, R., Martin, A., & Tavitian, B. (2010). Radioisotopic imaging of
neuroinflammation. Journal of Nuclear Medicine : Official Publication, Society of
Nuclear Medicine, 51(1), 1-4.
Woolrich, M. W., Jbabdi, S., Patenaude, B., Chappell, M., Makni, S., Behrens, T., . . . Smith,
S. M. (2009). Bayesian analysis of neuroimaging data in FSL. NeuroImage, 45(1 Suppl),
S173-86.
Wu, T., Liu, J., Zhang, H., Hallett, M., Zheng, Z., & Chan, P. (2015). Attention to automatic
movements in parkinson's disease: Modified automatic mode in the striatum. Cerebral
Cortex, 25(10), 3330-3342.
190
Wu, T., Long, X., Zang, Y., Wang, L., Hallett, M., Li, K., & Chan, P. (2009). Regional
homogeneity changes in patients with parkinson's disease. Human Brain Mapping,
30(5), 1502-1510.
Wu, T., Wang, L., Chen, Y., Zhao, C., Li, K., & Chan, P. (2009). Changes of functional
connectivity of the motor network in the resting state in parkinson's disease.
Neuroscience Letters, 460(1), 6-10.
Yasuno, F., Ota, M., Kosaka, J., Ito, H., Higuchi, M., Doronbekov, T. K., . . . Suhara, T.
(2008). Increased binding of peripheral benzodiazepine receptor in alzheimer's disease
measured by positron emission tomography with [11C]DAA1106. Biological
Psychiatry, 64(10), 835-841.
Yeo, B. T., Krienen, F. M., Sepulcre, J., Sabuncu, M. R., Lashkari, D., Hollinshead, M., . . .
Buckner, R. L. (2011). The organization of the human cerebral cortex estimated by
intrinsic functional connectivity. Journal of Neurophysiology, 106(3), 1125-1165.
Yokokura, M., Terada, T., Bunai, T., Nakaizumi, K., Takebayashi, K., Iwata, Y., . . . Ouchi,
Y. (2016). Depiction of microglial activation in aging and dementia: Positron emission
tomography with [11C]DPA713 versus [11C](R)PK11195. Journal of Cerebral Blood
Flow and Metabolism : Official Journal of the International Society of Cerebral Blood
Flow and Metabolism, April 26.
Yu, H., Sternad, D., Corcos, D. M., & Vaillancourt, D. E. (2007). Role of hyperactive
cerebellum and motor cortex in parkinson's disease. NeuroImage, 35(1), 222-233.
Zalesky, A., Fornito, A., Cocchi, L., Gollo, L. L., & Breakspear, M. (2014). Time-resolved
resting-state brain networks. Proceedings of the National Academy of Sciences of the
United States of America, 111(28), 10341-10346.
Zarei, M., Ibarretxe-Bilbao, N., Compta, Y., Hough, M., Junque, C., Bargallo, N., . . . Marti,
M. J. (2013). Cortical thinning is associated with disease stages and dementia in
parkinson's disease. Journal of Neurology, Neurosurgery, and Psychiatry,
Zgaljardic, D. J., Borod, J. C., Foldi, N. S., Mattis, P. J., Gordon, M. F., Feigin, A., &
Eidelberg, D. (2006). An examination of executive dysfunction associated with
frontostriatal circuitry in parkinson's disease. Journal of Clinical and Experimental
Neuropsychology, 28(7), 1127-1144.
Zhan, W., Kang, G. A., Glass, G. A., Zhang, Y., Shirley, C., Millin, R., . . . Schuff, N.
(2012). Regional alterations of brain microstructure in parkinson's disease using
diffusion tensor imaging. Movement Disorders : Official Journal of the Movement
Disorder Society, 27(1), 90-97.
191
Zhang, K., Yu, C., Zhang, Y., Wu, X., Zhu, C., Chan, P., & Li, K. (2011). Voxel-based
analysis of diffusion tensor indices in the brain in patients with parkinson's disease.
European Journal of Radiology, 77(2), 269-273.
Zhang, M. R., Kida, T., Noguchi, J., Furutsuka, K., Maeda, J., Suhara, T., & Suzuki, K.
(2003). [(11)C]DAA1106: Radiosynthesis and in vivo binding to peripheral
benzodiazepine receptors in mouse brain. Nuclear Medicine and Biology, 30(5), 513-
519.
Zhang, M. R., Maeda, J., Furutsuka, K., Yoshida, Y., Ogawa, M., Suhara, T., & Suzuki, K.
(2003). 18F]FMDAA1106 and [18F]FEDAA1106: Two positron-emitter labeled ligands
for peripheral benzodiazepine receptor (PBR). Bioorganic & Medicinal Chemistry
Letters, 13(2), 201-204.
Zhang, X., Haaf, M., Todorich, B., Grosstephan, E., Schieremberg, H., Surguladze, N., &
Connor, J. R. (2005). Cytokine toxicity to oligodendrocyte precursors is mediated by
iron. Glia, 52(3), 199-208.
Zhao, C., Li, W. W., & Franklin, R. J. (2006). Differences in the early inflammatory
responses to toxin-induced demyelination are associated with the age-related decline in
CNS remyelination. Neurobiology of Aging, 27(9), 1298-1307.
Zheng, Z., Shemmassian, S., Wijekoon, C., Kim, W., Bookheimer, S. Y., & Pouratian, N.
(2014). DTI correlates of distinct cognitive impairments in parkinson's disease. Human
Brain Mapping, 35(4), 1325-1333.