Elemental and Neurochemical Based Analysis of the Pathophysiological Mechanisms of Gilles De La Tourette Syndrome

Home , David Sulzer, Peter Dayan, Premonitory urge, Rolf Gruetter

MEDIZINISCHE HOCHSCHULE HANNOVER

Klinik für Psychiatrie, Sozialpsychiatrie und Psychotherapie

Elemental and Neurochemical Based Analysis of the Pathophysiological Mechanisms of Gilles de la Tourette syndrome

INAUGURALDISSERTATION zur Erlangung des Grades einer Doktorin oder eines Doktors der Naturwissenschaften Doctor rerum naturalium - Dr. rer. nat.

vorgelegt von

Ahmad Seif Kanaan B.Sc., M.Sc

geb. am 17/05/1986, Amman, Jordanien

Hannover, Deutschland (2017)

Wissenschaftliche Zweitbetreuung: Prof. Dr. rer. nat. Claudia Grothe

1. Erst-Gutachterin/Gutachter: Prof. Dr. med. Kirsten Müller-Vahl

2. Gutachterliche Stellungnahme durch: Prof. Dr. rer. nat. Claudia Grothe

3. Gutachterin/Gutachter: Prof Dr. Phil. Florian Beißner

Tag der mündlichen Prüfung: 12.07.2017

Promotionsordnung der Medizinischen Hochschule Hannover für die Erlangung des Grades einer Doktorin/eines Doktors der Naturwissenschaften (Doctor rerum naturalium) Letzte Änderung in der 523. Sitzung Senatssitzung vom 15.07.2015 To Seif, Shireen, Farah & Mira . . . “How can a three-pound mass of jelly that you can hold in your palm imagine angels, contemplate the meaning of inﬁnity, and even question its own place in the cosmos? Especially awe inspiring is the fact that any single brain, including yours, is made up of atoms that were forged in the hearts of countless, far-ﬂung stars billions of years ago. These particles drifted for eons and light-years until gravity and change brought them together here, now. These atoms now form a conglomerate- your brain- that can not only ponder the very stars that gave it birth but can also think about its own ability to think and wonder about its own ability to wonder. With the arrival of humans, it has been said, the universe has suddenly become conscious of itself. This, truly, it the greatest mystery of all.”

V.S. Ramachandran Acknowledgements

I warmly thank that patients and their families for their selfless contribution of time and effort to further the understanding of their affliction — a disease that makes them lead difficult, stigmatized and action packed lives. I particularly admired the determination of many of the patients who had to travel vast distances across Germany to undergo multiple investigations.

My most enduring acknowledgment goes to my supervisors Kirsten Müller-Vahl and Harald E. Möller who were the true pillars of this work. I sincerely thank them both for their excellent mentorship and guidance, and for giving me the opportunity to tackle absorbing ideas while traveling between Leipzig and Hannover.

My deep gratitude also goes to Alfred Anwander for teaching me how to think like an image scientist and for always keeping the door open for discussion. I sincerely thank Sarah Gerasch for her signiﬁcant contribution in recruiting the patients and acquiring the clinical data. I am also grateful to Daniel Margulies for giving me the opportunity to discover image analysis in his laboratory in the initial stages of my PhD.

My deep admiration and gratitude go to Isabel Garcia-Garcia who provided ample ideas and reviewed multiple drafts of initial works. Thank you for crossing my path, for teaching me how to conduct statistical analyses with a smile and for being there when it counted.

For their tremendous support in image and clinical data acquisition, I sincerely thank Saskia Czerwonatis, Claudia Pelke, Leonie Lampe, Tomas Goucha, Sieglinde Remane, Nicole Pampus, Christiane, Driedger-Garbe and Cornelia Gerbothe. I especially thank Andre Pampel and Torsten Schlumm for their support in developing the imaging sequences.

To Jamie Near, Berkin Bilgic and Andreas Schäfer, thank you for your openness to collaborate and for providing technical support in ﬁnding solutions to challenging imaging problems.

I would also like to thank my colleagues in the NMR group: Ricardo Metere, Tobias Lenich, Miguel Martinez Maestro, Henrick Marschner, Maria Guidi, Jakob Georgi and all other members for sharing this experience with me.

To the TS-EUROTRAIN community: Nacho Gonzalez, Francesca Rizzo, Ester Nespoli, Muhammad Sulaman Nawaz, Sam Padmanabhuni, John Alexander, Nuno Nogueira, Natalie Forde, Luca Pagliaroli, Sarah Fan, Joanna Widomska, Rayan Houssari, Peristera Paschou, Andrea Ludolph, Danielle Cathe, Pieter Hoekstra, Zeynep Tümer, Csaba Barta, Jeﬀrey Glennon, Bastian Hengerer, Hrienn Steﬀanson, Jeremiah Scharf, I thank you all for making this journey so cheerful and all the more worthwhile.

I also warmly thank Professor Mary Robertson for our illuminating discussions and for inspiring me to keep going forward. To Jim Leckman, I thank you for appreciating my work and bestowing me with the honor of the 2016 Professor Mary Robertson Award.

To my family Seif, Shireen, Mira, Farah and Hussam, I deeply thank you for your love, understanding and everlasting encouragement. Abstract

Elemental and neurochemical based analysis of the pathophysiological mechanisms of Gilles de le Tourette syndrome. Ahmad Seif Kanaan

Gilles de la Tourette syndrome is a developmental neuropsychiatric movement disorder characterized by the presence of tics and associated comorbid conditions. As current treatment strategies are often unsatisfactory and associated with significant adverse effects, there is an urgent need in further elucidating the nature of GTS pathophysiology to accelerate the drug discovery and development process. Neurochemically, previous work has indicated that the clinical manifestations of GTS are primarily driven by putative abnormalities in dopamine and γ-Aminobutyric-acid (GABA). Given the spatio-temporal and metabolic interdependence exhibit by the neurotransmitter glutamate with dopamine and GABA, respectively, we hypothesized the glutamatergic signalling is related to the pathophysiology of GTS. On a finer scale, considering the critical role exhibited by the element iron in varied biochemical processes sustaining typical neurochemical synthesis and trafficking throughout the lifespan, we additionally postulated that GTS patients exhibit abnormalities in iron metabolism. Utilizing a multi-parametric, quantitative Magnetic Resonance Imaging approach in vivo, we investigated the role of glutamate and iron using 1H-Magnetic Resonance Spectroscopy and Quantitative Susceptibility Mapping, respectively, for the first time. To achieve these aims, two methodological investigations were initially conducted to obtain quantitative neurochemical and magnetic susceptibility measurements of sufficient precision to identify rather subtle changes. Imaging, spectroscopic and clinical data were acquired from a relatively large and well-characterized sample of adult patients with GTS and age/gender matched healthy controls. To interrogate the influence of treatment on neurochemical and clinical characteristics of the study sample, we employed a longitudinal study design in which the patients were invited to undergo treatment with the commonly used antipsychotic aripiprazole. At the neurochemical level, we report significant reductions in the concentrations of spectroscopic glutamatergic signalling markers in the striatum and the thalamus in GTS. These reductions correlated with tic severity and were normalized with aripiprazole treatment. At the elemental level, we report significant reductions in subcortical magnetic susceptibility which is regarded as surrogate index for iron content. Reductions were specific to subcortical nuclei key in coordinating mechanisms of motor and non-motor habit formation, and were mirrored by decreases in serum ferritin levels. Importantly, significant associations were observed between striatal susceptibility and glutamatergic neurotransmission as indexed by the glutamine:glutamate ratio. Clinically, treatment with aripiprazole led to significant reductions of tic severity in the patient sample, and additionally led to an approximate 50% reduction in OCD diagnosis. Our results indicate that patients with GTS exhibit an abnormality in the flux of metabolites in the GABA-glutamate-glutamine cycle, thus implying perturbations in astrocytic-neuronal coupling systems that maintain the subtle balance between excitatory and inhibitory neurotransmission within subcortical nuclei. These abnormalities may be driven or further compounded by the observed abnormalities in iron metabolism. Chronic perturbations in the subcortical GABA-glutamate-glutamine cycle flux could lead to spatially focalized alterations in excitatory, inhibitory and modulatory subcortical neurochemical ratios that would have a profound influence on the neuroplastic mechanisms involved

ii in reinforcement learning and habit formation systems, which are governed by striatal neurons that code the serial order of syntactic natural behaviour. This work sheds a new light on the neurobiological basis of GTS and provides novel clues that may prove critical in the future development of functionally selective pharmacological modulators that target multiple neurochemical systems. Zusammenfassung

Elemental and neurochemical based analysis of the pathophysiological mechanisms of Gilles de le Tourette syndrome. Ahmad Seif Kanaan

Das Gilles de la Tourette-Syndrom (GTS) ist eine neuropsychiatrische Entwicklungsstör- ung, die durch das Bestehen von Tics und psychiatrischen Komorbiditäten gekennze- ichnet ist. Die heute zur Verfügung stehenden Behandlungsmöglichkeiten sind häufig unzureichend wirksam und führen oft zu klinisch relevanten Nebenwirkungen. Die Er- forschung der Ursachen des GTS könnte daher auch die Entwicklung neuer Behand- lungsmöglichkeiten mit sich bringen. Die Mehrzahl der vorliegenden Studienergebnisse weist auf eine dem GTS zugrunde liegende Störung im dopaminergen und GABAer- gen Neurotransmitter-System hin. Wegen der engen Wechselwirkungen zwischen dem dopaminergen und GABAergen System einerseits und dem glutamatergen System an- dererseits, haben wir die Vermutung aufgestellt, dass möglicherweise auch das gluta- materge System in die Pathophysiologie des GTS involviert sein könnte. Vor dem Hin- tergrund, dass das Spurenelement Eisen an zahlreichen biologischen Prozessen beteiligt ist inklusive der Synthese und Funktion zahlreicher Neurotransmitter, haben wir darüber hinaus die Hypothese aufgestellt, dass bei Patienten mit GTS eine Störung im Eisen- stoffwechsel vorliegen könnte. Mit Hilfe des Einsatzes multiparametrischer quantitativer Magnet-Resonanz-Tomographie-Techniken haben wir erstmals regionale Konzentratio- nen von Glutamat (mittels 1H-Kernspinresonanzspektroskopie, MRS) und Eisen (mittels quantitativer Suszeptibilitätskartierung, QSM) bei Patienten mit GTS gemessen. Im Rahmen unserer Studie wurden bei einer großen Gruppe erwachsener Patienten mit GTS sowie einer alters- und geschlechts-gematchten Gruppe gesunder Kontrollpersonen um- fangreiche klinische und bildgebende Daten erhoben. Um zusätzlich den Einfluss einer medikamentösen Behandlung sowohl auf neurochemische als auch auf klinische Parameter untersuchen zu können, haben wir eine Langzeitstudie durchgeführt und eine Subgruppe der Patienten vor und nach einer Behandlung mit dem etablierten Antipsychotikum Arip- iprazol untersucht. Die wesentlichen Ergebnisse diese Studie sind: (a) bei Patienten mit GTS findet sich im Vergleich zu gesunden Kontrollen eine mittels MRS nachgewiesene signifikante Reduktion glutamaterger Marker im Striatum und Thalamus. Diese korreliert mit der Schwere der Tics und normalisiert sich durch eine Behandlung mit Aripipra- zol; (b) mittels QSM ist eine signifikante Verminderung der subkortikalen Suszeptibilität nachweisbar, welche als Index für den Eisengehalt gewertet wird. Interessanterweise fan- den sich diese Verminderungen besonders in jenen Hirnarealen, die als Schlüsselgebiete für die motorische Koordination und Bildung bewegungsunabhängiger Gewohnheiten angesehen werden. Parallel konnte im Serum eine Verminderung der Ferritin-Spiegel nachgewiesen werden; (c) schließlich konnten wir einen Zusammenhang zwischen der im Striatum gemessenen Suszeptibilität und der glutamatergen Neurotransmission fest- stellen; (d) die Behandlung mit Aripiprazol führte zu einer signifikanten Verminderung der Tics und Zwänge. Unsere Ergebnisse belegen, dass bei Patienten mit GTS Verän- derungen im Zusammenspiel der Transmitter GABA, Glutamin und Glutamat bestehen. Dies wiederum deutet auf eine Störung im Zusammenspiel zwischen Astrozyten und Neu- ronen hin, welches für die feine Balance im Zusammenwirken zwischen inhibitorischen und exzitatorischen Neurotransmittern in subkortikalen Kerngebieten entscheidend ist. Es kann spekuliert werden, dass diese Veränderungen auf eine Störung im Eisenstoffwech- sel in der frühen Hirnentwicklung zurückzuführen sind. Die gefundenen Veränderungen iv sowohl im glutamatergen Transmitter-System als auch im Eisenstoffwechsel erlauben neue Einblicke in die Pathogenese des GTS und eröffnen möglicherweise zukünftig neue Behandlungsmöglichkeiten dieser komplexen Erkrankung. Declaration of Authorship

I, Ahmad Seif Kanaan, declare that this thesis titled: "Elemental and neurochemical based analysis of the pathophysiological mechanisms of Gilles de la Tourette syndrome" and the work presented in it, unless otherwise stated, are my own. I conﬁrm that:

This work was done wholly or mainly while in candidature for a research degree at this University.

Where any part of this thesis has previously been submitted for a degree or any other qualiﬁcation at this University or any other institution, this has been clearly stated.

Where I have consulted the published work of others, this is always clearly at- tributed.

Where I have quoted from the work of others, the source is always given. With the exception of such quotations, this thesis is entirely my own work.

I have acknowledged all main sources of help.

Where the thesis is based on work done by myself jointly with others, I have made clear exactly what was done by others and what I have contributed myself.

Signed: Ahmad Seif Kanaan

Date: 06.06.2017

vi Contents

Acknowledgements i Abstract ii Zusammenfassung iv Declaration of Authorship vi List of Figures xi List of Tables xiii Abbreviations xiv Symbols xvii

I INTRODUCTION 2

1 Overview 3

2 Gilles de la Tourette Syndrome 6 2.1 A brief history ...... 6 2.2 Clinical Phenomenology ...... 9 2.3 Etiological Basis ...... 11 2.4 Pathophysiology: functional anatomy ...... 12 2.4.1 Structural neuroimaging studies ...... 12 2.4.2 Functional neuroimaging studies ...... 12 2.4.3 Key points gleaned from neuroimaging studies ...... 13 2.5 Pathophysiology: neurochemical aspects ...... 15 2.5.1 Mico-Circuitry Of the Basal Ganglia ...... 15 2.5.2 Dopamine ...... 17 2.5.3 GABA ...... 20 2.5.4 Glutamate ...... 21 2.5.5 Other neurotranmistters ...... 22 2.6 Pathophysiology: elemental aspects ...... 23 2.7 Treatment ...... 24

3 Objectives 26 3.1 Primary objectives ...... 26 3.2 Study Design ...... 27 3.3 Data acquisition ...... 28 3.4 A note on treatment with aripiprazole ...... 29 3.5 Ethics ...... 30

vii Table of Contents viii

II METHODS 31

4 Magnetic Resonance Imaging and Spectroscopy 32 4.1 Nuclear Magnetic Resonance Imaging ...... 32 4.1.1 Principles of Nuclear Magnetic Resonance Imaging ...... 32 4.1.2 NMR properties of biological tissue environments ...... 34 4.1.3 NMR signal localization ...... 34 4.1.4 Pulse sequence design ...... 36 4.1.5 Image reconstruction ...... 36 4.1.6 Encoding physio-chemical properties in images ...... 38 4.2 Magnetic Resonance Spectroscopy ...... 40 4.2.1 Principles of in-vivo NMR Spectroscopy ...... 40 4.2.2 Acquisition of 1H-MRS spectra ...... 41 4.2.3 Time domain signal processing ...... 43 4.2.4 Spectral quantitation ...... 45 4.2.5 1H-MRS metabolites of the human brain ...... 48 4.3 Quantitative Susceptibility Mapping ...... 50 4.3.1 Magnetic Susceptibility ...... 50 4.3.2 Phase Imaging ...... 51 4.3.3 Susceptibility Weighted Imaging ...... 52 4.3.4 Quantitative Susceptibility Mapping ...... 53 4.3.5 Sources of magnetic susceptibility contrast ...... 56

5 Clinical Assessment 58 5.1 Tics, premonitory urges and quality of life ...... 58 5.2 Obsessive Compulsive Disorder ...... 59 5.3 Attention defecit-hyperactivity disorder ...... 59 5.4 Depression ...... 59 5.5 Anxiety ...... 59 5.6 GTS subgroup classiﬁcation ...... 60

III Methodological Investigations 61

6 Absolute metabolite quantitation 62 6.1 Introduction ...... 62 6.2 Methods ...... 63 6.3 Results ...... 64 6.4 Discussion ...... 65 6.5 Conclusions ...... 67

7 Coil Combination 68 7.1 Introduction ...... 68 7.2 Methods ...... 69 7.3 Results ...... 70 7.4 Discussion ...... 71 7.5 Conclusions ...... 73 Table of Contents ix

IV Pathophysiological Investigations 74

8 Neurochemical Investigation of pathophysiology 75 8.1 Abstract...... 75 8.2 Introduction ...... 76 8.3 Materials and Methods ...... 79 8.3.1 Population Sampling ...... 79 8.3.2 Magnetic Resonance Data Acquisition ...... 81 8.3.3 Absolute Metabolite Quantitation ...... 82 8.3.4 Statistical Analysis ...... 85 8.4 Results ...... 86 8.4.1 Demographic and clinical characteristics ...... 86 8.4.2 Test-Retest Reliability ...... 86 8.4.3 Degree of tic-urges and tic-suppression during MR data acquisition 88 8.4.4 Group Differences in Metabolite Concentrations ...... 89 8.4.5 Correlation of Metabolite Concentrations with Clinical Variables . 96 8.4.6 Comparison of Glutamate/Glutamine separation at 3T and 7T .. 97 8.4.7 Influence of head displacement on spectral measures ...... 99 8.4.7.1 Intra-group differences in head motion ...... 99 8.4.7.2 Inter-group differences in head motion ...... 100 8.4.7.3 Influence of head motion on voxel compartmentation and absolute metabolite quantitation ...... 103 8.5 Discussion ...... 105 8.5.1 Altered Glutamate-Glutamine Cycling in GTS ...... 105 8.5.2 The Role of Functionally Selective Modulators in the Adaptive Stabilization of Neurotransmitter Systems in GTS ...... 108 8.5.3 The Influence of Subcortical GABA-Glu-Gln Cycling Abnormali- ties on Dopaminergic Signalling and the Phenomenology of Tics .. 110 8.5.4 Methodological Limitations and Future Directions ...... 112 8.6 Conclusions ...... 113

9 Elemental investigation of pathophysiology 114 9.1 Abstract ...... 114 9.2 Introduction ...... 115 9.3 Materials and Methods ...... 118 9.3.1 Population Sampling ...... 118 9.3.2 Measurement of serum Ferritin ...... 118 9.3.3 Magnetic Resonance Imaging and Spectroscopy ...... 119 9.3.4 Quantitative Susceptibility Mapping ...... 119 9.3.5 Masking of Subcortical Matter Nuclei ...... 120 9.3.6 Quality Control ...... 121 9.3.7 Statistical Analysis ...... 122 9.4 Results ...... 124 9.4.1 Group diﬀerences in surrogate measures of iron ...... 124 9.4.2 Magnetic susceptibility correlations with ferritin, Gln:Gln and clinical measures ...... 127 9.5 Discussion ...... 130 9.5.1 Disturbed iron homeostasis in GTS ...... 130 9.5.2 Disruptions in iron homeostasis inﬂuence mechanisms of subcortical neurochemical signaling ...... 131 Table of Contents x

9.5.3 Limitations and future directions ...... 133

10 Investigation of the influence of aripiprazole on clinical status 134 10.1 Abstract ...... 134 10.2 Introduction ...... 135 10.3 Methods ...... 137 10.3.1 Population Sampling ...... 137 10.3.2 Clinical Assessment ...... 137 10.3.3 Statistical Analysis ...... 137 10.4 Results ...... 138 10.4.1 Patient characteristics at baseline ...... 138 10.4.2 Patient characteristics during treatment with aripiprazole ..... 140 10.4.2.1 Tics and premonitory urges ...... 140 10.4.2.2 Psychiatric comorbidities and quality of life ...... 140 10.4.3 Comparison of clinical characteristics of patients electing for- and against-treatment ...... 142 10.4.4 Adverse Effects and continuation of treatment ...... 143 10.4.5 Serum levels of aripiprazole ...... 143 10.5 Discussion ...... 143 10.5.1 Efficacy of aripiprazole on tics and premonitory urges ...... 144 10.5.2 GTS subgroup classification ...... 145 10.5.3 Efficacy of aripiprazole on associated comorbid conditions ..... 145 10.5.4 Influence of aripiprazole on quality of life ...... 146 10.5.5 Adverse effects of aripiprazole ...... 147 10.5.6 Decision factors for treatment with aripiprazole ...... 147 10.5.7 Characteristics of the Sample and serum levels of aripiprazole .. 148 10.5.8 Limitations ...... 148 10.6 Conclusions ...... 149

V CONCLUSIONS 150

11 Key findings and significance 151 11.1 Pathological glutamatergic neurotransmission in GTS ...... 152 11.2 Subcortical iron reductions associated with glutamatergic neurotransmission in GTS ...... 153 11.3 Aripiprazole improves associated comorbid conditions in addition to tics in GTS ...... 154 11.4 Significance ...... 156

VI Bibliography 157

VII APPENDIX 184

A Related publications 185 B Contribution of authours 188 C Research compliance certiﬁcates 190 D MR Imaging sequence parameters 197 E Author portfolio 213 List of Figures

2.1 Une leçon clinique á la Salpêtriére...... 7 2.2 Georges Gilles de la Tourette ...... 8 2.3 Clinical manifestations of GTS ...... 10 2.4 Major cortico-striato-thalamo-cortical circuits in the human brain ..... 17 2.5 Selective dysfunction of basal ganglia subterritories ...... 18 2.6 Interactions between the major neurochemical systems in the cortico- striato-thalamo-cortical pathway...... 19

3.1 Illustration of the longitudinal study design of the project...... 28

4.1 Basic Physics of the NMR signal ...... 33 4.2 Selective excitation of an image slice by applying a shaped RF pulse and field gradient at the same time...... 35 4.3 Conventional MRI pulse sequence diagrams ...... 37 4.4 MR image reconstruction from k-space ...... 37 4.5 The relationship between TR/TE and the encoding of physio-chemical tissue properties as image contrasts...... 39 4.6 Chemical shift properties of N-Acetylaspartate ...... 41 4.7 The effect of echo time on metabolite detectability ...... 43 4.8 1H-MRS time domain signal processing ...... 44 4.9 Referencing methods for absolute metabolite quantitation ...... 46 4.10 The linear combination basis spectrum fitting model ...... 47 4.11 GABA, Glutamate, Glutamine cycling ...... 49 4.12 Magnetic Susceptibility of diamagnetic and paramagnetic material .... 50 4.13 Tissue phase imaging ...... 52 4.14 Susceptibility-Weighted Imaging ...... 53 4.15 Quantitative Susceptibility Mapping Image processing ...... 55 4.16 Correspondence between Perls’ stain and QSM in revealing iron deposition in the deep grey matter nuclei ...... 57

6.1 Test-retest reliability of commonly used tissue segmentation methods ... 66 6.2 Test-retest reliability of 1H-MRS absolute metabolite quantitation .... 67

7.1 Tissue phase estimation using adaptive coil combination and ESPIRiT- SVD from multichannel data...... 71 7.2 QSM diﬀerences following multi-channel coil combination with the adaptive and ESPIRiT-SVD algorithms ...... 72

8.1 Voxel localization and spectral data pre-processing ...... 83 8.2 Spatial overlap of test-retest voxel localization ...... 88 8.3 Spectral localization, ﬁtting and statistical analysis ...... 90

xi List of Figures xii

8.4 Representative cingular 1H-MRS spectra of the frequency and phase-drift corrected data...... 93 8.5 Representative thalamic 1H-MRS spectra of the frequency and phase-drift corrected data...... 94 8.6 Representative striatal 1H-MRS spectra of the frequency and phase-drift corrected data...... 95 8.7 Correlation between absolute metabolite concentrations and clinical measures...... 96 8.8 1H spectra achieved at 3T and 7T to inspect Glutamate/Glutamine separation...... 98 8.9 Consistency of head motion across scanning sessions ...... 100 8.10 The eﬀect of subject movement on within voxel tissue content and metabolite concentration...... 104 8.11 LCModel individual metabolite ﬁtting ...... 106 8.12 Astrocytic-neuronal coupling and the homeostasis of glutamatergic and GABAergic neurotransmission ...... 109 8.13 Local circuit model of subcortical connectivity ...... 111

9.1 Processing and analysis framework utilized to obtain high-quality quantitative susceptibility maps and 1H-MR spectra...... 120 9.2 QQ plots of the multivariate outlier detection technique implemented via squared Mahalanobis distance...... 123 9.3 QSM Data quality...... 125 9.4 Nucleus segmentation quality and group comparison statistics...... 126 9.5 Ferritin group diﬀerences and correlations with susceptibility...... 128 9.6 Correlational analysis between susceptibility and Gln:Glu ...... 129

10.1 Distribution of YGTSS-TTS scores at baseline and followup in the different subgroups ...... 140 10.2 Psychiatric comorbidity subclassification at baseline and following treatment with aripiprazole ...... 141 10.3 Prevalence of comorbidities in patients that elected for- and against-treatment with aripiprazole...... 142 10.4 Adverse effects reported by the patients following the administration of aripiprazole...... 143 10.5 Prevalence of comorbidities in patients that elected for- and against-treatment with aripiprazole...... 144 List of Tables

6.1 Test-retest tissue fraction estimates of SPM, FSL and Freesurfer (FSU) .. 65

7.1 Statistical comparison between magnetic susceptibility values achieved with the adaptive and ESPIRiT-SVD coil combination methods ...... 71

8.1 Demographic and clinical characteristics of the 1H-HMRS study sample included in the final analysis ...... 80 8.2 Water T1 and T2 relaxation times and relative water content (α) in GM, WM and CSF ...... 85 8.3 Test-Retest Reliability of absolute metabolite quantitation ...... 87 8.4 Control vs. GTS group comparison of absolute metabolite concentrations 91 8.5 GTS Off- and On-treatment group comparison of absolute metabolite concentrations ...... 92 8.6 Influence of frequency/phase drift correction on spectral measures. .... 102 8.7 Comparison of frequency/phase corrected spectral data between high and low motion control groups ...... 102

9.1 Statistical comparisons of magnitude image data quality metrics ...... 122 9.2 Demographic and clinical characteristics of the QSM study sample included in the ﬁnal analysis ...... 124 9.3 Statistical comparison of magnetic susceptibility within general regions of interest ...... 127 9.4 Statistical comparison of magnetic susceptibility within distinct subcortical nuclei ...... 127

10.1 Subgroup classiﬁcation of the whole study sample based on comorbidities 138 10.2 Clinical characteristics of the whole study sample at baseline and following treatment with aripirazole ...... 139

xiii Abbreviations

3D three-dimensional AAH auto-align head ADHD attention deficit hyperactivity disorder aMCC anterior mid-cingulate cortex ANTS advanced normalization tools BAI beck anxiety inventory BDI-II beck depression inventory II CAARS conners’ adult ADHD rating scales COV coefficient of variation Cho choline compounds Cre (phospoho)creatine CRLB cramér-rao lower bound CSF cerebrospinal fluid CSTC cortico-striato-thalamo-cortical D1 dopamine type-1 like receptor D2 dopamine type-2 like receptor DOF degrees of freedom DOPA dihydroxyphenylalanine EFC entropy focus criterion EPI echo planar imaging ESPIRiT eigenvalue approach to autocalibrating parallel MRI FASTESTMAP fast, noniterative shimming of spatially localized signals FD framewise displacement FID free induction decay FLASH fast low-angle shot

xiv Abbreviations xv

FWHM full width at half maximum GABA γ-aminobutyric acid Gln glutamine Glu glutamate Glx glutamate plus glutamine GM grey matter

GPe globus pallidus external segment

GPi globus pallidus internal segment GTS gilles de la tourette syndrome GRE gradient echo MADRS montgomery-äsberg depression rating scale m-Ins myo-inositol MP2RAGE magnetization-prepared 2 rapid gradient echo MR magnetic resonance MRS magnetic resonance spectroscopy MSN medium-sized spiny neuron NAA n-acetylaspartate NMDA n-methyl-D-aspartate NMR nuclear magnetic resonance OCB obsessive- compulsive behavior OCD obsessive-compulsive disorder OCI-R obsessive-compulsive inventory, revised PET positron emission tomography PRESS point-resolved spectroscopy PUTS premonitory urge for tics scale RCT randomized controlled trial ROI region of interest rs-fMRI resting-state functional magnetic resonance imaging RVTRS rush video-based tic rating scale SD standard deviation SHARP sophisticated harmonic artifact reduction on phase data SMA supplementary motor area

SNc substantia nigra pars compacta Abbreviations xvi

SNr substantia nigra pars reticulata SNR signal-to-noise ratio STEAM stimulated echo acquisition mode SVD singular value decomposition SVS single voxel spectroscopy TCA tricarboxylic acid cycle TKD thresholded k-space division WM white matter WURS-k wender Utah rating scale Y-BOCS yale-brown obsessive compulsive scale YGTSS yale global tic severity scale YGTSS-GS ygtss global score YGTSS-TTS ygtss total tic score QI1 quality index 1 QOL quality of life scale QSM quantiative susceptibility mapping Symbols

B0 static magnetic ﬁeld

B1 applied magnetic ﬁeld

cm metabolite concentration 0 cw water concentration in bulk water

dx,dy, dz translational displacement from the center along the x-, y-, and z-axis

fε relative volume fraction of tissue type ε within a voxel

Im metabolite signal intensity

Iw water signal intensity

Nm number of protons in the molecule contributing to the metabolite signal P error probability

Rε relaxation factor of tissue type ε r correlation coeﬃcient S signal

Sref reference signal

T1 longitudinal relaxation time

T2 transverse relaxation time

Te echo time

Tr repetition time α, β, γ rotation angles about the x-, y-, and z-axis

aε relative water content in tissue type ε ξ scaling factor accounting for partial voluming ϕ phase ω larmour frequency χ magnetic susceptibility

xvii Symbols 1 Part I

INTRODUCTION

2 Chapter 1

Overview

Gilles de la Tourette syndrome (GTS) is a multifaceted and anomalous disorder that wavers along the fine margin between neurology and psychiatry with an enshrouding set of symptoms. The pathophysiology of GTS remains enigmatic, as tics — the hallmark feature of GTS — are usually joined by related symptoms such as echo- and copro- phenomena, as well as comorbid conditions that include attention deficit/hyperactivity disorder (ADHD), obsessive-compulsive disorder (OCD) and depression. Although a range of pharmacological, behavioral and surgical therapeutic approaches are currently being used to manage the symptomatology of GTS, there is currently no cure for the manifesting motor and non-motor features. This is largely due to the lack of a comprehensive pathophysiological model of GTS, which may be attributable to the presence of a large number of inconsistent findings and the dearth of studies utilizing multi-parametric approaches.

Considering the complex phenomenology of GTS, genetic analyses have not revealed a precise abnormality exhibiting Mendelian inheritance, but have rather pointed to a more-complex polygenic pattern of inheritance that leads to variabilities in a number of systems. Observations from (a) pharmacological investigations, (b) histopathological analyses of post mortem specimens, (c) neurophysiological investigations, (d) surgical and pharmacological animal model studies, and (e) several morphometric, functional and nuclear imaging studies have provided support for the pathological involvement of the basal ganglia and cortico-striatal-thalamo-cortical (CSTC) pathways.

At the neurochemical level, abnormalities in dopaminergic neurotransmission are widely considered as a primary abnormality in GTS. However, in view of the strict spatio- temporal synergy exhibited between excitatory, inhibitory and modulatory neurotransmitter systems that drive typical motor and non-motor behaviour, several groups have posited that other neurochemical systems may exhibit perturbations as well. Along this

3 Overview 4 line of reasoning, more recent studies have implicated other neurotransmitter systems that most prominently include the γ-Aminobutyric acid (GABA) system. Although GABA exhibits both spatio-temporal as well close metabolic links to glutamate, investigations on the role of the glutamatergic system in GTS have not yet been undertaken in vivo.

At the elemental level, one unifying feature exhibited by the dopaminergic, GABAer- gic and glutamatergic neurotransmitter systems is that the enzymes involved in their metabolism and the production of their receptors/transporters require iron for typical function. Along this line, preliminary work investigating serum ferritin concentrations have indicated that iron may be related to GTS pathophysiology. Nevertheless, in vivo investigations of the role of iron in GTS pathophysiology are currently lacking.

With recent advances in quantitative Magnetic Resonance Imaging (MRI) techniques, a unique perspective that allows the in vivo measurement of biochemical and physio- chemical tissue properties has been afforded, thus paving the road for undertaking more incisive investigations of disease specific changes at multiple-scales. Proton Magnetic Resonance Spectroscopy (1H-MRS) and Quantitative Susceptibility Mapping (QSM) are two such approaches that allow the measurement of steady-state metabolite quantities and the intrinsic physical property of magnetism in matter. Using 1H-MRS, for example, several neurochemicals could be quantified, including surrogate measures related to glutamatergic signalling. On the other hand, QSM is a novel contrast mechanism that provides surrogate quantitative measures of specific biomarkers such as iron.

Consequently, the overarching goal of this thesis is to investigate GTS pathophysiology using multi-parametric, quantitative MRI approaches at multiple-scales, in order to oﬀer novel perspectives of pathophysiology that could potentially pave the road for the design of new therapeutic approaches. In this regard, the secondary aim of this work was to investigate the inﬂuence of an established pharmacological agent (aripirazole) on neurochemistry and clinical status in GTS.

This thesis is composed of 5 separate Parts that are organized into 11 Chapters as follows:

• Part I: INTRODUCTION: The state of the art of the clinical phenomenology and pathophysiology of GTS are ﬁrst introduced (Chapter 2) in order to qualify the hypotheses and the objectives of this work (Chapter 3).

• Part II: METHODS: A succinct overview of the utilized imaging and clinical methods is presented in Chapter 4 and Chapter 5, respectively. Overview 5

• Part III: METHODOLOGICAL INVESTIGATIONS: Two methodological investigations were initially conducted to improve the accuracy of the performed magnetic resonance measurements. These included:

– Chapter 6: Investigation of the test-retest reliability of 1H-MRS absolute metabolite quantitation with partial volume correction. The results of this investigation led to improvements in the accuracy of absolute metabolite quantiation and served as a benchmark for the neurochemical investigation of pathophysiology presented in Chapter 8. – Chapter 7: Investigation of the influence of different coil-combination algorithms on QSM. This work led to conclusion that vendor provided phase data should not be used for QSM reconstruction since they contain an artifact that may bias group comparisons. Therefore, this work led to significant enhancements that permitted the elemental investigation of pathophysiology presented in Chapter 9.

• Part IV: PATHOPHYSIOLOGICAL INVESTIGATIONS: The main objectives of this work are presented as three separate investigations of GTS pathophysiology and are organized as follows:

– Chapter 8: Investigation of the role of the glutamatergic system in the pathophysiology of GTS using single voxel 1H-MRS. – Chapter 9: Investigation of the role of iron in the pathophysiology of GTS using QSM and serum ferritin analysis as surrogate measures of iron. – Chapter 10: Investigation of the inﬂuence of treatment with aripiprazole on the motor and non-motor features manifested in GTS.

• Part V: KEY FINDINGS AND SIGNIFICANCE: A summary of all the key ﬁndings and a discussion of their signiﬁcance is presented in the concluding chapter (Chapter 11). Chapter 2

Gilles de la Tourette Syndrome

This chapter provides an overview of Gilles de la Tourette syndrome pathophysiology with a speciﬁc focus on neurochemical aspects.

2.1 A brief history

The 19th century was a remarkable period in the history of neuroscience that saw the characterization of novel disorders of the brain and the clear documentation of previously recognized ones. Much of this is owed to the work of Jean-Martin Charcot (1825– 1893) who spent 33 years studying the nervous system and teaching students at the Pitié-Salpêtriére Hospital in Paris. His reputation attracted many bright students who, themselves, later become pioneers in various ﬁelds. This list contains names that include Sigmund Freud, Joseph Babinski, Pierre Janet and George Gilles de la Tourette, among others (Figure 2.1).

Under his mentorship, Gilles de la Tourette (1857–1904) began studying various neurological disorders before focusing on "obscure" movement disorders. At the age of 28, Gilles de la Tourette published a landmark article [1] about a bizarre condition that exhibited stereotyped movements, phonic symptoms, premonitory sensations, echo- and copro-phenomena, which he referred to as that "maladie de tics" (Figure 2.2). Owing to Tourette’s pioneering work in which he documented this condition as a distinct neurological disorder, Charcot bestowed the eponym "Gilles de la Tourette syndrome" (GTS) in his honor. Interestingly, there is some discourse in the current academic arena of who should be the true bearer of the eponym. Twelve years before his 1885 publication, descriptions of symptoms similar to that of Tourette’s were outlined in a monograph by Armand Trousseau (1801–1867) [2]. The true nature of events and reason why Charcot

6 Gilles de la Tourette Syndrome 7

Figure 2.1: Une leçon clinique á Salpêtriére. André Brouillet’s famous 1887 painting entitled "A Clinical Lesson at the Salpêtriére" is one of the best known paint- ings in the history of medicine. The painting illustrates Jean-Martin Charcot (mid- right) while he delivers a clinical lecture on the symptoms of hysteria to Gilles de la Tourette (mid-left) and his colleagues at the Pitié-Salpêtriére hospital in Paris [3]. bestowed the eponym upon Gilles de la Tourette is open for speculation. Nevertheless, these symptoms were ﬁnally brought to the fore of medical discourse. For centuries, individuals exhibiting tics had been largely outcast, persecuted and victimized by groups within society. On the other hand, some individuals who exhibited these symptoms were incredibly successful and the most notable example includes Samuel Johnson, the author of the modern dictionary.

Following Gilles de la Tourette’s landmark work, little progress was made in explicat- ing the pathophysiological basis of GTS or in developing new treatment strategies. In the early 20th century, GTS was primarily treated as a psychological disorder, where patients were usually told that their own psychological flaws were to blame for their symptoms. Nonetheless, the field took a positive turn in 1968 following Arthur and Elaine Shapiro’s success in suppressing the tics of a 24-year-old patient with the drug haloperidol, a dopamine receptor blocker [4]. The Shapiros heavily criticized the psycho- analytic approach and paved the way for further progress in understanding the pathophysiology of the disorder and treating it. However, it could be argued that their success in the treatment of tics is as significant as their work in establishing the first organized Tourette Syndrome Society in 1972. With the help of many patients and their families, the Shapiros established the American Tourette Syndrome Association (TSA), which was Gilles de la Tourette Syndrome 8

Figure 2.2: Georges Gilles de la Tourette. Georges Albert Édouard Brutus Gilles de la Tourette (30 October 1857 – 26 May 1904) was a French physician and the namesake of Gilles de la Tourette syndrome. He began his medical studies at Poitiers in 1873 before relocating to Paris where he became a student of the inﬂuential neurologist Jean-Martin Charcot, the director of the Salpêtriére Hospital. Tourette studied and lectured in psychotherapy, hysteria and hypnosis, before describing the symptoms of what he called the "maladie des tics" in his work "Étude sur une aﬀection nerveuse caractérisé e par de l’incoordination motrice accompagnée d’ècholalie et de coprolalie" (https://en.wikipedia.org/wiki/Georges_Gilles_de_la_Tourette). instrumental in promoting information and increasing the public discourse about GTS in the following decades.

Nevertheless, funding the study of GTS remained scarce, as it was understood to be a rare disorder with very low prevalence rates. Largely owing to the significant public response towards several published articles highlighting GTS, the 1980s saw a marked increase in research funding and the inclusion of the disorder in the third edition of the Diagnostic and Statistical Manual of Mental Disorders. Notwithstanding the large number of published works investigating GTS over the past three decades, current funding opporunities and scientific output are still overshadowed by other disorders such as Parkinson’s disease and Autism. In the year 2000, a pan European Society for the study of GTS (ESSTS) was established by Marie Robertson and Anne Korsgaard with the aims of increasing collaborative research and public awareness across Europe. The establishment of ESSTS was a significant step forward towards securing new funding opportunities and establishing new European guidelines for the (a) clinical characterization [5], (b) pharmacological Gilles de la Tourette Syndrome 9 treatment [6], (c) behavioral treatment [7], and surgical treatment [8] of patients with GTS and other tic disorders. As a result of the endeavors of many scientists and physi- cians working within such societies, we now have a much better understanding of the pathophysiology of GTS and have much better ways of treating it. Current knowledge of clinical, etiological, pathological and interventional aspects of GTS are summarized in the sections below.

2.2 Clinical Phenomenology

In popular culture, Gilles de la Tourette syndrome (GTS) is frequently regarded as a disorder in which individuals utter obscene, socially inappropriate or derogatory statements without any volitional control. This is partly true as only a minority of patients suffering from GTS exhibit this involuntary swearing condition known as coprolalia. In essence, GTS is a multifaceted disorder that exhibits both motor and non-motor clinical features. Specifically, GTS is a childhood-onset, neuropsychiatric movement disorder with relatively high heritability and prevalence rates [9] estimated at about 1% of the general population, thus rendering it as one of the most common movement disorders [10]. Tics are the cardinal features of GTS and are defined as rapid, non-rhythmic, stereotyped involuntary movements or utterances that are misplaced in context and time [9, 11]. Tics usually follow a waxing and waning course of severity and usually improve or go into complete remission in adulthood [12, 13]. Most tics are preceded by unpleasant sensory sensations (premonitory urges) which are relieved by executing the tic [14]. Psychiatric comorbidity in GTS is very common with lifetime prevalence rates of any psychiatric comorbidity estimated at 86% [15, 16]. These comorbid conditions most commonly include attention deficit hyperactivity disorder (ADHD) (≈50%), obsessive compulsive disorder (OCD) (20-60%), depression and anxiety, leading many to suggest an overlap in their pathophysiological mechanisms [17, 18] (Figure 2.3).

In the clinical setting, GTS is diagnosed if an individual exhibits: (a) multiple motor tics and one or more vocal tics; (b) a waxing and waning course of tic intensity, frequency and severity; (c) the persistence of tics for at least one year following ﬁrst onset; (d) tic onset before the age of 18; and (e) the presence of tics that are not caused by another medical condition or substance (DSM-5, 307.23, ICD-10 F95.2) [19]. The typical age of onset of tics is between 6 to 8 years of age, where peak severity is usually reached between ages 10-14. While the majority of patients exhibit signiﬁcant improvements over the course of their adolescent period, a small number of patients do not improve and in most of these cases, the patients take on a negative course of illness in which comorbid psychiatric conditions surface. In addition to the common conditions associated with Gilles de la Tourette Syndrome 10

Figure 2.3: Clinical manifestations of GTS. While tics are considered as the cardinal features of GTS, they are just the tip of the iceberg. The majority of patients exhibit other non-motor features that include obsessive-compulsive, attention defecit/hyperactivity, anxiety and depressive behavior among others. The ﬁg- ure was retreived from the media department of the Tourette association of America (https://www.tourette.org/resource/iceberg-illustration-poster/). Gilles de la Tourette Syndrome 11

GTS (ADHD, OCD, depression, anxiety), other conditions that could surface include self-injurious behavior, impulsivity, rage attacks, sleep problems and learning disorders [18]. Such associated conditions often impair a patients’ qualify of life and are usually the reason why patients seek treatment.

Diﬀerent therapeutic methods are currently being used to manage the symptomatology of GTS. These include: (a) behavioral therapy, (b) pharmacological treatment, (c) deep brain stimulation, and (d) alternative medicine [20]. Nevertheless, current treatment strategies are strictly palliative and are often ineﬀective and unsatisfactory. Conse- quently, a better understanding of the pathophysiological mechanisms of GTS could pave a new road for designing better treatments. Currently, there is no generally accepted etiological or pathophysiological model of GTS. In general, current data suggests that GTS has a complex genetic background where it is likely caused by genetic susceptibility factors that interact with the environment to confer the total risk of acquiring a complex phenotype [21].

2.3 Etiological Basis

The etiology of GTS is still not well characterized though much research is currently being undertaken in this area. Research into the genetic basis of GTS has been marked by difficulties in the replication of original findings, largely due to the low power of individual studies. Nevertheless, the heredity of GTS is undisputed in current literature [22, 23]. Akin to other complex psychiatric disorders exhibiting a diverse symptomatology, current data indicates that GTS is not caused by single gene effects that exhibit classic Mendelian patterns of inheritance, but exhibits a complex etiological basis likely caused by the complex interaction of multiple genetic variants with environmental factors. The complex interplay of multiple genetic variants is believed to be a primary contributor to the diverse phenotypes exhibited in GTS [22, 23], though an increasing body of evidence has provided support for a role of environmental factors in the onset and natural course of the disease (e.g. smoking during pregnancy, delivery complications etc.) [24, 25]. Nevertheless, there remains much debate on the role of environmental influences in GTS. For example, there has been much discourse and speculation on the role of post-streptococcal infection in tic onset [24, 26, 27]. Gilles de la Tourette Syndrome 12

2.4 Pathophysiology: functional anatomy

On a systems level, major strides have been achieved in the ﬁeld over the past two decades, though a generally accepted pathophysiological model of GTS remains elusive. Although evidence from structural and functional Magnetic Resonance Imaging (MRI) studies is varied, these studies have repeatedly highlighted abnormalities within the limbic, associative and motor cortico-striatal-thalamo-cortical (CSTC) networks [28]. A summary of structural and functional neuroimaging studies is highlighted in the following subsections.

2.4.1 Structural neuroimaging studies

Despite the abundance of structural MRI studies, there are many inconsistencies in the literature. The diversity of results can be explained by variabilities in sample size, patient characteristics (e.g. age, presence of comorbidities, neuroleptic use), imaging methods (e.g. region-of-interest based vs. whole brain), and statistical analysis methods [29, 30]. Nonetheless, there are some consistencies in structural investigations of GTS patients and these include: (a) a reduction in the volume of the caudate in both children and adults [31–34], and (b) cortical thinning in the sensorimotor, prefrontal and cingulate cortices [32, 35–37]. It is of note to mention that the specificity of cortical thinning seems to be dependent on the phenotype. For example, Worbe et al. [37] showed that patients with simple tics (cortical thinning in primary sensory and motor cortices) exhibit a different structural profile relative to patients with complex tics (cortical thinning in premotor, prefrontal and associative areas). Overall, structural MRI indicate that GTS is associated with dysfunction in associative, limbic and motor CTSC regions [28, 29, 38].

2.4.2 Functional neuroimaging studies

Evidence from task based functional MRI (fMRI) studies on the involvement of specific regions in the pathophysiology of GTS is relatively scarce and ambiguous. Task fMRI studies can be classified into four groups; investigations of motor behaviour, tic generation, tic suppression and executive function. Groups investigating mechanisms of tic generation and control in GTS have compared (a) simple vs. complex tics [39]; (b) tic imitation vs. real tics [40]; (c) periods of tic suppression vs. periods of free ’ticking’ [41]; (d) pre-monitory urge periods and tic generation periods [42]; (e) suppression of eye blink in healthy controls vs. GTS patients [43]. Although these studies show that an extensive network of premotor, primary motor and sensorimotor networks are involved Gilles de la Tourette Syndrome 13 in tic generation, a critical distillation of the studies only ties the Supplementary Mo- tor Area (SMA) to tic generation [29]. A few other groups have aimed at investigating motor performance in GTS patients. These studies are limited and conflicting, as some groups showed over activations in the SMA and the sensorimotor cortex during a finger tapping task [6, 44], while others show no group differences for the same task [45]. There is much debate on whether executive function is impaired in GTS. For example, (a) some behavioral studies have demonstrated impairments in several domains of executive function (response inhibition, selective attention, cognitive flexibility) [46–48]; (b) others have argued that such impairments are driven by comorbid conditions [49]; (c) while a third group of studies presented evidence of enhanced performance by GTS patients on cognitive tasks [50, 51].

Moreover, studies that implemented fMRI paradigms designed to investigate response inhibition in GTS patients also exhibited inconsistencies. First, using a Stroop task to investigate the neural correlates of response inhibition, Marsh et al. [52] demonstrated that tic severity and the persistence of GTS symptoms may be due to disrupted matura- tion of fronto-striatal circuits involved in self-regulatory control, since prefrontal activity correlated with tic severity and there was no age related change in frontostriatal circuitry as exhibited by the control group. Second, also using a Stroop task in addition to a go/no-go task, Debes et al. [45] demonstrated that there were no differences between healthy controls and patients, however, activity in the posterior cingulate and superior temporal gyrus correlated with OCD symptoms. This suggests that response inhibition symptoms may be due to comorbidity as argued by Ozonoff [49] and [53]. Third, similar to Debes [45] using a go/no-go task, Hershey et al. [54] also failed to find any group difference between controls and patients. Fourth, using the Simon task, Raz et.al [55] showed that increased activation in frontostriatal circuitry was associated with increased task related frontostriatal activity associated with better task performance. However, in contrast to the results of Marsh [52], Raz et al. [55] showed that frontostriatal activity correlates with age and tic severity. In conclusion, fMRI studies on response inhibition are highly inconsistent and further investigation of the neural correlates of GTS patients is warranted.

2.4.3 Key points gleaned from neuroimaging studies

Despite the substantial potential benefit afforded by structural and functional neuroimaging methods to further our understanding of the pathophysiological mechanisms underlying the GTS spectrum, a series of recently published papers have indicated that the quality of neuroimaging data is key in performing unbiased group comparisons [56–64]. One crucial factor that must be considered for the faithful evaluation of brain imaging Gilles de la Tourette Syndrome 14 data is related to motion artifacts that ensue as a result of displacements in head position within the head-coil during MRI data acquisition. As patients with GTS are essentially characterized by movement, the disruption of the acquired MRI signal leading to motion contaminated data is inevitable. The presence of motion artifacts in MRI data has been shown to: (a) lead to spurious correlations in estimates of functional connectivity [60, 65, 66] and (b) influence cortical thickness and grey-matter-volume morphometric estimates [56, 67]. While the inconsistencies in the published works investigating GTS pathophysiology may have been driven by variabilities in: (a) sample sizes, which range between N=10–60; (b) the clinical characteristics of the study samples; and (c) the status of psychoactive drug use, the inclusion of low-quality data may have also led to significant biases in the observed results.

Notwithstanding the aforementioned limitations which may have led to the inconsistent observations between diﬀerent studies, these studies were crucial in implicating speciﬁc brain regions that exhibit altered structural or functional characteristics in GTS. Key points gleaned from structural and functional MRI studies investigating GTS pathophysiology are summarized below:

• Both children and adult patients with GTS exhibit reductions in the volume of the caudate, which correlates negatively with tic-severity.

• Patients with simple tics exhibit cortical thinning in primary sensory and motor cortices, while patients with complex tics exhibit cortical thinning in premotor, prefrontal and associative areas.

• Results from fMRI studies of motor and cognitive control are generally inconsistent and may be biased by motion artifacts, small sample size and variabilities in the clinical characteristics of the study samples.

• Premonitory urges are associated with abnormalities in the supplementary motor area. Gilles de la Tourette Syndrome 15

2.5 Pathophysiology: neurochemical aspects

A large body of imaging, spectroscopic and post-mortem studies investigating neurochemical aspects of GTS pathophysiology have been published to date. Given that initial success of the dopaminergic antagonist haloperidol in reducing tic symptoms, these studies have mainly focused on the dopaminergic system. More recent works, however, have highlighted a role for other neurochemical systems in GTS. In the following section, an overview of neurochemical interactions in the basal ganglia in relation to GTS pathophysiology is presented.

2.5.1 Mico-Circuitry Of the Basal Ganglia

The basal ganglia are group of subcortical nuclei that act as a cohesive functional unit allowing the tight regulation of motor, cognitive and limbic functions. The main components of the basal ganglia are the Striatum (caudate and putamen), the Substantia

Nigra (pars compacta — SNr and pars reticulata — SNr), the Subthalamic Nucleus (STN) and the Globus pallidus interna (GPi) and externa (GPe) (Figure 2.4)[68]. With the exception of the excitatory glutamatergic (— ) STN projections, all intrinsic and output projections of the basal ganglia are γ-Aminobutyric acid (GABA)ergic (—) and inhibitory. The striatum and the STN are the main input structures and they receive glutamatergic excitatory signals from the cerebral cortex, the brainstem, and the limbic system. On the other hand, the GPi/SNr complex is the main output system project- ing GABAergic inhibitory neurons onto the thalamic sub-nuclei which feed back to the cortex. The GPe and the SNc hold intrinsic functions and provide the striatum with important modulatory signals [69].

In current anatomical models of internal basal ganglia circuitry, two antagonistic striato- thalamic pathways exist. A direct and an indirect pathway are organized in a way that allows for the selection or inhibition of competing actions [70]. In this model, the striatum is linked to the GPi/SNr output complex via a monosynaptic direct pathway emanating from set of distinct GABAergic medium-sized spiny neurons (MSNs) within the striatum. In the direct pathway, excitatory cortical input to the striatum results in the release of inhibitory signals from the GPi/SNr to the thalamus. This releases the ’breaks’ oﬀ the thalamus allowing it to emit its excitatory signals that facilitate the execution of movement. The direct pathway can be summarized as follows:

Cortex — Striatum — SNr/GPi ···· Thalamus — Cortex. Gilles de la Tourette Syndrome 16

The indirect pathway on the other hand, is polysynaptic where a diﬀerent population of striatal MSNs delivers input to the GPi/SNr output complex via the GPe and its intercalated STN. In this case, cortical excitability of the striatum leads to a decrease in the tonic inhibition of the STN by the GPe. As such, the STN is free to release its glutamatergic excitatory signals to the output nuclei furthering the inhibition of the thalamus and leading to a net hypokinetic state. The indirect pathway can be summarized as follows:

Cortex — Striatum — GPe ···· STN — SNr/GPi — Thalamus···· Cortex. Overall, the indirect pathway leads to the inhibition of movement while the direct pathway leads to the facilitation of movement.

Macroscopically, early models of basal ganglia physiology have posited a "funneling and selection" function of its nuclei [71]. In the late 1980s, however, this model has been replaced by a segregated circuit model in which the basal ganglia are seen as components of ﬁve segregated and parallel re-entrant loops, over which information sent from speciﬁc cortical areas is processed topographically and is integrated within its internal circuitry [69, 72]. These circuits fall into three major domains (motor, limbic and associative) and include the sensorimotor, oculomotor, dorsolateral prefrontal, lateral orbitofrontal, and anterior cingulate limbic loops [69, 70, 72] (Figure 2.4).

Disequilibrium in the natural mechanics of the inbound CSTC loops is a major factor in the emergence of both motor and non-motor features exhibited by various neuropsychiatric and movement disorders [74, 75]. Along this line, previous work has indicated that spatially focalized alterations in neurochemical ratios play a major role in the emergence of diverse motor, limbic and associative features (Figure 2.5). Keeping in mind that GTS is not only centered on tics, where the majority of the patients exhibit a varied motor and behavioral symptomatology, early hypotheses have postulated that the expression of tics and its accompanying behavioral features is a result of atypical neurotransmission within speciﬁc sub-territories of the basal ganglia, which consequentially lead to the aberrant integrative interplay of CSTC circuitry [76–78]. As the basal ganglia provide a mechanism for the selection of an action from a competing response [72, 79], fundamental alterations in the functional dynamics of CSTC circuitry could cause selection errors that are believed to lead to the expression of motor and non-motor features as exhibited in GTS [75, 78]. Gilles de la Tourette Syndrome 17

Figure 2.4: Major cortico-striato-thalamo-cortical circuits in the human brain. The ﬁgure illustrates pseudo-anatomical arrangement of the motor, associative and limbic pathways. Figure retrieved from [73].

2.5.2 Dopamine

Information incoming from different cortical regions is integrated at the level of the striatum in order to build a motor pattern that is based on current and past experience [82]. The striatum can be subdivided in two zones: the dorsal zone (matrisomes) which is predominantly involved in motor behavior [78, 82], and a ventral zone (striosomes) which receives its main input from limbic and associative areas [83, 84]. At this level, the specific trajectory of excitatory striato-thalamic circuitry is regulated by the neurotransmitter dopamine (Figure 2.6). Dopamine shapes a wide variety of psychomotor functions and exhibits its potent modulatory influence over striatal MSNs by acting on Dopamine- 1 (D1) or Dopamine-2 (D2) receptors which lie along the presynaptic dendritic shafts [85–87]. D1 stimulation facilitates N-methyl-D-aspartate (NMDA) mediated glutamate transmission promoting MSN excitability and activation of the direct pathway [86, 88]. In contrast, D2 stimulation inhibits NMDA glutamate transmission leading to net thalamic excitability [85]. Gilles de la Tourette Syndrome 18

Figure 2.5: Selective dysfunction of basal ganglia subterritories. Utilizing local injections of the GABAergic antagonist bicuculline, a causal link between the neurochemical activity of speciﬁc functional subterritories of the basal ganglia and clinical manifestations observed in movement and behavioral disorders were demonstrated [75, 80, 81]. The authors showed that the emergence of symptoms associated with GTS (e.g. tics, stereotypy, attentional deﬁcits) is dependent on the spatial location of the neurochemical alteration. Figure retrieved from [75].

Several lines of evidence have suggested a strong role for dopamine in GTS pathophysiology [31, 89]. Dopaminergic dysfunction in GTS is supported by (a) initial clinical observations of improvements in tic-like symptoms following the administration of dopaminergic antagonists (pimozide, ﬂuphenazine, haloperidol, risperidone, aripiprazole), synthesis blockers (α-methyl-para-tyrosine) or monoamine depletion drugs (tetrabenazine); and the exacerbation of symptoms following the administration of dopaminergic stimulants (L-DOPA, central stimulants) [20, 90–94]; (b) results from varied nuclear imaging studies that show alterations in dopamine transporter and receptor function in striatal and extra-striatal regions [95–110]; (c) increased dopamine concentrations in cerebrospinal ﬂuid [89, 111] and (d) altered dopamine levels as revealed by post-mortem studies [100, 110, 112]. Given this data, multiple dopaminergic hypotheses have thus been posited on the role of dopamine in GTS pathophysiology. These hypotheses include: (a) dopaminergic hyper-innervation within the striatum; (b) a presynaptic abnormality in aromatic L-amino acid dihydroxyphenylalanine (DOPA) decarboxylase, which is involved in the catalysis of L-DOPA to dopamine; (c) super-sensitivity of postsynaptic Gilles de la Tourette Syndrome 19

Figure 2.6: Interactions between the major neurochemical systems in the cortico-striato-thalamo-cortical pathway. See text for details. Figure retrieved from [31]. . striatal receptors; and (d) excessive intrasynaptic release of DA as a result of an imbal- ance between tonic and phasic levels [28, 31, 77]. Given that current models indicate that dopaminergic signalling in functionally compartmentalized into a tonic and phasic system, in which the dynamic spatio-temporal interplay between dopamine transporters, receptors and excitatory/inhibitory aﬀerent systems play a major role in output [113– 121], a critical distillation of the literature suggests a dysregulation in the ﬁring patterns of central monoaminergic nuclei [31].

However, given that (a) dopamine exhibits extensive interactions with other neurochemical systems that may inﬂuence its release; and (b) the heterogeneity of clinical phenotypes manifested in GTS, it is highly likely that more than one neurotransmitter system is implicated in GTS [28, 31]. Some studies have indeed suggested the involvement of GABAergic, cholinergic, glutamatergic, serotonergic, and histaminergic systems in GTS pathophysiology [31]. Gilles de la Tourette Syndrome 20

2.5.3 GABA

GABA is the primary inhibitory neurotransmitter in the human brain. There are two main types of GABAergic neurons that include ( a) projections neurons and ( b) interneurons [122]. Although the majority of GABAergic striatal neurons are of the spiny projection type (i.e. MSNs), striatal GABAergic interneurons produce a strong inhibitory inﬂuence over MSNs, regulating their output to the GPi and the GPe in the direct and indirect pathways, respectively [123].

Considering that GTS is in essence a disorder exhibiting pathological mechanisms of inhibitory control, various authors have posited that the primary inhibitory neurotransmitter GABA may have an important role in GTS pathophysiology [31]. For example, the burst-like disinhibition of thalamo-cortical projections which would ultimately facilitate hyperkinetic behaviour (ie. tics), could be a result of alterations in GABAergic striatal projection neurons (ie. MSNs) or striatal as well cortical interneurons. Given that pallidal GABAergic neurons exhibit a potent inﬂuence over dopaminergic neurons of the substantia nigra, abnormalities in GABAergic pallidal projection could also have an important role in GTS pathophysiology.

Although preliminary post mortem work investigating GABA levels in various brain regions failed to find any significant differences between patients and controls [124], more recent work has demonstrated that GABA does exhibit abnormalities in GTS. Utilizing unbiased immunocytochemical techniques, Kalanithi et al. [125] demonstrated profound alterations in the density and distribution of a specific type of GABAergic interneurons. Specifically, higher densities of parvalbumin-positive GABAergic interneurons were found in the GPi, whereas lower densities were observed in the GPe, caudate and putamen. The authors were able to replicate and extend their work in another study, in which they showed that different subtypes of interneurons also exhibit reductions [126]. In another study, Lerner et al. [127] utilized Positron Emission Tomography with injection of the ra- 11 dioligand [ C]flumazenil to visualize the distribution of GABAA receptors. The authors demonstrated that patients with GTS exhibit widespread abnormalities in the GABAer- gic system, as they observed (a) reductions in the binding capacity of GABAA receptors in the ventral striatum, thalamus, amygdala and right insula, and (b) reductions in the substantia nigra, periaqueductal gray, posterior cingulate cortex and cerebellum. More recent work utilizing Magnetic Resonance Spectroscopy expanded on these findings by demonstrating GABAergic reductions in (a) striatal, cingulate and sensorimotor regions in pediatric sample [128, 129], and (b) elevations in the supplementary motor area in an adult sample [130]. In general, this work provides strong support for an abnormality in GABAergic neurotransmission in GTS. Gilles de la Tourette Syndrome 21

2.5.4 Glutamate

Given its wide distribution as the brain’s primary excitatory neurotransmitter and its essential role in the normal mechanics of CSTC circuitry, a pathophysiological role for glutamatergic neurotransmission in GTS has recently been postulated [131]. A role for glutamate in GTS pathophysiology can be indirectly drawn from the fact that (a) glutamate and GABA exhibit close metabolic links, in which the penta-carbon skeleton of glutamate acts a precursor for the synthesis of GABA and (b) glutamate and dopamine exhibit extensive interactions at different levels of cortico-striato-thalamo-cortical circuitry [131] (Figure 2.6). With respect to the interactions of glutamate and dopamine, it has been suggested that the co-transmission of both neurotransmitters is possible in central monoaminergic neurons such as in the ventral tegmental area (VTA) [132]. Second, mesocortical dopaminergic inputs from the VTA are able to directly and indirectly alter pyramidal neuron excitability in prefrontal cortical regions [133]. Third, descending cortical glutamatergic afferents, modulate dopaminergic VTA and SNc neurons which feed back to the striatal matrisomes and striosomes, respectively [134, 135]. Fourth, cortical glutamatergic afferents and dopaminergic SNc projections converge towards striatal MSNs, where dopamine is able to modulate glutamatergic neurotransmission depending on the type of receptor it targets.

On the other hand, direct evidence of glutamatergic abnormalities in GTS is drawn from pathophysiological and genomic studies. Pathophysiologically, early postmortem studies demonstrated a reduction in glutamate levels in the GPi, GPe, and SNr [124]. This evidence correlates with volumetric MRI analysis that indicated a reduction in the size of the left GP [136], although a direct link was never made. Moreover, two genetic studies have also highlighted a role for glutamate in GTS. One large multigenerational family genome scan identiﬁed a role for 5p13 - an area that overlaps with glial glutamate transporter gene1 [137]. Another genome scan using sibling pairs identiﬁed a missense variant in E219D, a highly conserved residue that results in increased glutamate uptake [138, 139]. Additionally, inference can also be drawn from studies that investigated the role of glutamate in OCD since there is a high degree of pathophysiological overlap. Other genetic studies also implicated SLC1A1 gene (glutamate transporter) and GRIN2B gene (glutamate related gene expressing a subunit of the NMDA glutamate receptor) in OCD [140–142], which exhibits a phenomenological overlap with GTS pathophysiology.

In a Magnetic Resonance Spectroscopy study conducted on an adolescent patient sample, DeVito et al. [143] did not find any differences in the concentration of glutamate in various brain regions. However, the authors demonstrated reduced levels of N-acetylaspartate and choline in the left putamen, and reduced creatine levels in the putamen bilaterally. Nevertheless, this study had several limitations as (a) 50% of their patient group was Gilles de la Tourette Syndrome 22 sedated using chloral hydrate and this was not accounted for; (b) the impact of comorbidities and medication was not included in the analysis; (c) only a male population was investigated. In conclusion, the authors suggested that further studies are needed to replicate their results and to investigate the impact of comorbidity and the effect of psychotropic medication on metabolite levels.

In conclusion, though insuﬃcient, some published work points to a role for glutamate in GTS. It is not known whether GTS patients exhibit a hyper- or a hypo-glutamatergic state. For example, tics can be induced either by (a) excess dopaminergic stimulation of the striatum or by (b) excess cortical glutamatergic input onto the striatum. In both of these scenarios, hyperkinetic eﬀects could be induced [131]. Interrogating the role of glutamate in GTS will possibly have direct implications on novel interventional strategies [131].

2.5.5 Other neurotranmistters

Considering that typical motor and non-motor behaviour is driven by strict spatio- temporal interactions between various neurochemical systems, it is possible that neurotransmitter systems other than the primary excitatory, inhibitory and modulatory systems may be involved. Though preliminary, some work has indicated links between the cholinergic, serotonergic and histaminergic neurotransmitter systems and GTS. The strongest possible link is related to the cholinergic system. Initial immunocytochemical work indicated that patients with GTS exhibit alterations in the density and distribution of cholinergic interneurons [126], which may have drastic downstream consequences on the GABAergic projection and interneuron populations that they inﬂuence. This notion was demonstrated by the targeted ablation of striatal cholinergic neurons in a rat model, in which tic-like stereotypies were induced following acute stress of d-amphetamine chal- lenges [144].

Studies investigating the role of serotonin in GTS have revealed: (a) reductions of serotonin metabolite levels in the cerebrospinal ﬂuid [111, 145, 146] and the basal ganglia [124] in some but not all patients; (b) normal metabolite levels in the cortex [112]; (c) a decrease in postmortem density of 5HT-1A receptor levels in frontal and occipital areas [110]; (d) a reduction in the binding capacity of 5HT transporters in association with OCD [147]; (iv) a negative correlation between vocal tics and 5HT transporter binding potential in the midbrain and thalamic areas [148]. Other nuclear imaging studies investigating serotonin receptor binding exhibited variables results, where some showed no change [108] to increases in multiple regions [149]. In summary, these ﬁndings indicate Gilles de la Tourette Syndrome 23 that patients with GTS plus comorbid OCD may exhibit an abnormality in serotonergic signalling.

Related to the histaminergic system, though in vivo investigations of histaminergic levels in have not yet been explored in GTS, genetic ﬁndings have indicated that dysregula- tions in histaminergic neurotransmission may represent a rare cause for tourette syndrome [31, 150]. In general, proposals for the underlying mechanism of abnormality have included most neurochemical systems, however, the strongest evidence published to date indicates abnormalities in dopamine and GABA. Further cross-sectional and longitudinal in vivo investigations of other neurochemical systems may further elucidate the pathophysiological mechanisms of GTS.

2.6 Pathophysiology: elemental aspects

Iron, most commonly stored in ferritin and transported via transferrin, is an essential trace element that is involved in varied biochemical processes. [151]. Iron has been shown to be most densely concentrated in extrapyramidal structures, with greatest concentrations found in basal ganglia nuclei, speciﬁcally in the striatum, nucleus accumbens and the globus pallidus, in addition rostral brain nuclei including the substantia nigra, the red nucleus, and the dentate nucleus [151, 152]. Considerable evidence has linked alterations in the concentration of brain iron to several developmental and movements disorders that include: Parkinson’s disease, restless leg syndrome, autism spectrum disorder, hemochromatosis, multiple sclerosis among others [151, 153, 154].

Functionally, iron plays an important role in the maintenance of various neurobiological processes. More specifically iron (a) is involved in oxidation, hydroxylation, peroxida- tion mechanisms; (b) is a component of different enzyme families; (c) is essential for neurotransmitter synthesis and transport, (d) is involved in myelination and (e) has a significant role in the transport of oxygen [151, 153, 155]. As a result, iron levels have been consistently linked to dopaminergic concentrations. Along this line, various groups have shown that iron deficiency (a) affects dopaminergic neurotransmission during development [156]; (b) alters dopamine metabolism in the striatum of iron deficient rats [157, 158]; (c) alters dopamine transporter function in the rat striatum [159]; and (d) modulates the sensitivity and number of D1 and D2 receptors [159, 160]. Nevertheless, other work has also indicated that iron plays a role in the metabolism and neurotransmission of GABA and glutamate [161] (See Chapter 9 for more details).

The role of cerebral non-heme iron has been rarely interrogated in GTS. Gorman et al. [162] made indirect links to GTS by measuring the concentration of serum ferritin and Gilles de la Tourette Syndrome 24 calculating regional brain volumes [162]. The authors showed that ferritin levels were lower in GTS patients and that they correlated positively with volumes of sensorimotor, mid-temporal and sub-genual cortical areas. Nonetheless, the study had several limitations as (a) only serum measures were obtained where the relation between ferritin and iron in the brain is unclear; and (b) subjects were not assessed for dietary related iron depletion [162]. In another recent epidemiological study on tic disorders and their relationship to trace elements, the authors found a signiﬁcant decreases in iron and zinc levels in GTS [163]. In general, preliminary work indicates that abnormalities in iron metabolism may be related to GTS pathophysiology, though future in vivo investigations of its role in the GTS are warranted.

2.7 Treatment

Currently, there is no cure for GTS as treatment strategies are strictly palliative and are often ineffective or unsatisfactory. As a result, there is an urgent need for uncovering novel treatment strategies that (a) cause less adverse effects, (b) are more effective in treatment resistant patients and (c) could ameliorate both tics and behavioural symptoms. While much information has recently been uncovered on GTS, there remains an urgent need for further exploring its complex etiology and pathophysiology to pave the road for new treatment strategies.

A variety of psychotropic drugs have been used to treat tics in GTS. To date, only haloperidol has been licensed for the treatment of GTS in Germany (BfArM). Nonethe- less, it has been established that long term use of haloperidol is not so favorable as it induces severe side effects that include sedation, cognitive dulling, dystonic reactions, dyskinesias and weight gain [164]. Although a variety of other (a)typical antipsychotics (e.g. pimozide, quetiapine, ziprasidone and olanzapine) used for the treatment of GTS patients lead to a reduction of tic scores, many of them induce significant adverse effects similar to haloperidol [165]. According to both German and European guidelines for the treatment of tics, risperidone, sulpiride, tiapride (children) and aripiprazole are recommended as first line treatment [6].

Aripiprazole, a relatively recently developed drug [166], has been demonstrated by a number of groups to lead to significant improvements in tics while inducing minimal side effects [167–170]. While it is not officially approved for the treatment of GTS, it is still recommended as a first choice medication in the treatment of tics [6]. This is largely due to the fact that (a) it is better tolerated and causes less side effects relative to other antipsychotic drugs, and (b) is effective in many severely affected and treatment-refractory patients [170]. Studies exploring the use of aripiprazole in GTS have mainly focused on Gilles de la Tourette Syndrome 25 characterizing motor behaviour in child and adolescent patient samples. However, investigations of the consequences of aripiprazole treatment on both motor and non-motor behaviour are lacking. Neurochemically, aripiprazole has an adaptive pharmacological profile that targets multiple neurochemical systems. Though it mainly influences the dopaminergic and serotonergic systems, studies have shown that it also exhibits an influence on the GABAergic and glutamtergic systems [171]. As such, the utility of a longitudinal study design to investigate the effect of aripiprazole on (a) metabolite levels, as well as (b) motor and non-motor behaviour may provide crucial information on the pathophysiology of GTS and how to treat it. Chapter 3

Objectives

The main objectives of this thesis and details of the overall study design are presented in this chapter. The presented work reﬂects partial material published in a peer-reviewed perspective article that describes the pursued hypotheses within the framework of a Euro- pean consortium aiming for the study of Gilles de la Tourette syndrome pathophysiology: Forde NJ and Kanaan A.S. et. al., TS-EUROTRAIN: A European-wide investigation and training network on the aetiology and pathophysiology of Gilles de la Tourette Syn- drome. Frontiers in Neuroscience 2016, 10: 1-9[172].

3.1 Primary objectives

The work presented in this thesis is part of an EU-funded Marie Curie Initial Training Network specific to the study of Gilles de la Tourette Syndrome (GTS) (TS-EUROTRAIN- FP7-PEOPLE-2012-ITN, Grant Agr. No. 316978) [172]. The overarching goal of the study is to further clarify the pathophysiological mechanisms of GTS with a specific focus on the roles of the neurochemical glutamate, the element iron and the influence of treatment on neurochemistry and clinical status. Based on the current state of the art, our primary objectives were to investigate the following questions:

• Do GTS patients exhibit abnormalities in the glutamatergic system?

• Can subtle diﬀerences in glutamatergic metabolite concentrations be quantiﬁed with a certain degree of reliability using single voxel Magnetic Resonance Spec- troscopy in brain regions implicated in GTS?

• Do GTS patients exhibit an iron deﬁcient status in subcortical regions?

• Do alterations in the concentrations of glutamate or iron exhibit correlations with behavioral measures? 26 Objectives 27

• Is the pharmacological treatment of tics related to the glutamatergic system? I.e. Does aripiprazole act as a tic suppressing agent by inducing changes in the glutamatergic system?

• How does treatment with aripiprazole inﬂuence the clinical characteristics of the patients?

3.2 Study Design

To undertake a nuanced investigation of possible neurochemical and elemental abnormalities, participants were invited to partake in two sub-studies, in which data was collected cross-sectionally and longitudinally. Baseline data was collected from up to 44 patients with GTS and up to 40 healthy controls (See Chapters 6-10 for further details of population sampling). Patients undergoing any psychoactive drug treatment underwent a four-week washout period before data collection. All participants were invited to partake in the longitudinal study design, in which 23 healthy controls and 17 patients agreed to participate. For the patients, the longitudinal study design entailed undergoing a pharmacological treatment with aripiprazole for a minimum period of four-weeks, in an open-label, uncontrolled treatment as usual. For the controls, longitudinal study data was collected to assess the test-retest reliability of neurochemical quantitation. All participants were recruited by word of mouth, local-advertisement and by coordination with Germany Tourette Society (Tourette Gesellschaft Deutschland e.V.). A visual illustration of the study design is outlined in Figure 3.1.

It is important to note that the final sample sizes analyzed in the methodological and pathophysiological investigations (Chapter 6-10) exhibited variabilities due to: (a) differences in the processing and analysis schemes, and (b) differences in data quality control criteria. The framework implemented for achieving the final samples is detailed in the respective chapter. Objectives 28

Figure 3.1: Illustration of the longitudinal study design of the project. In stage 1, clinical and MRI data were collected from patients and controls at baseline. Patients undergoing any psychotropic drug use were asked to undergo a four week washout period before data acquisition. In stage 2, both patients and controls were recruited again for further MRI and clinical measurements. In the stage, the patients underwent a standard pharmacological treatment using the antipsychotic aripiprazole to investigate the inﬂuence of treatment on their neurochemical proﬁles. Control subjects were tested again for test-retest reliability measurements.

3.3 Data acquisition

To answer the questions outlined in the previous Section 3.1, neuroimaging data, behavioral data and blood-samples were collected from the large majority of recruited subjects. In a few cases, data collection of the recruited subjects was not fully completed for various reasons that included (a) ineligibility due to the presence of magnetic resonance contraindications, (b) current abuse of drugs or alcohol, (c) claustrophobia or (d) fatigue. Overall the data collection procedure entailed two 90-minute sessions for the acquisition of imaging and clinical data. Unless otherwise indicated, all imaging data were acquired by the author and all clinical data were acquired by Sarah Gerasch. More details on the collected data are outlined below:

• Magnetic Resonance Spectroscopy: Single voxel 1H-MRS data were collected to assess the neurochemical proﬁle of the study sample with a speciﬁc focus on the quantitation of glutamate and glutamine concentrations in cortico-striatal- thalamo-cortical regions implicated in GTS pathophysiology (See Chapters 6 and 8).

• Quantitative Susceptibility Mapping: High-resolution gradient-echo MR data were collected for the calculation of quantitative susceptibility maps which served as a surrogate measure of iron levels in deep grey matter nuclei (See Chapters 7 and 9). Objectives 29

• T1 weighted MR data: T1 weighted anatomical images were collected and used for various purposes that included: spectroscopic voxel localization, image segmentation and registration, estimation of tissue compartmentation and the masking of deep grey-matter nuclei.

• Clinical assessment: A comprehensive battery of clinical assessment and self- reported tests were collected to assess the degree of severity of core GTS symptoms (tics, premonitory urges) and comorbid conditions (e.g. OCD, ADHD, depression, anxiety) (See Chapter5 and 10 for further details). In brief the tests included:

– Clinical tools: ∗ M.I.N.I.: M.I.N.I. International Neuropsychiatric Interview [173] ∗ YGTSS: Yale Global Tic Severity Scale [174]. ∗ RVTRS: Modiﬁed Rush Video-Based Tic Rating Scale [175]. ∗ YGTSS: Yale Global Tic Severity Scale [174]. ∗ PUTS: Premonitory Urge for Tics Scale [176]. ∗ GTS-QOL: Gilles de la Tourette syndrome-quality of life scale [177]. ∗ Y-BOCS: Yale-Brown Obsessive Compulsive Scale [178]. ∗ MADRS: Montgomery-Äsberg Depression Rating Scale [179]. ∗ DSM-V-Symptom list for ADHD [19].

– Self reported tests: ∗ CAARS: Conners’ Adult ADHD Rating Scales [180]. ∗ BDI-II: Beck Depression Inventory II [181, 182]. ∗ BAI: Beck Anxiety Inventory [181]. ∗ OCI-R: Obsessive-Compulsive Inventory - Revised [183]. ∗ AQ: Autism spectrum Quotient [184].

• Blood samples: Blood samples were collected for the quantiﬁcation of (a) serum ferritin concentrations as a secondary measure of iron, and (b) aripiprazole concentrations to investigate patient compliance in drug-intake.

3.4 A note on treatment with aripiprazole

Although a variety of drug classes are currently being used for the treatment of GTS, aripiprazole (Abilify R ) was used as the drug of choice for the work presented in this thesis for the following reasons: Objectives 30

• Aripiprazole is recommended as the ﬁrst line of treatment by German and European GTS guidelines [6].

• Aripiprazole is one of the most commonly used pharmaceuticals for the treatment of tics, although it is not oﬃcially licensed yet.

• Aripiprazole exhibits a favorable side eﬀect proﬁle in comparison to other antipsychotics [170].

• Aripiprazole is a functionally selective drug that exhibits an adaptive pharmacological proﬁle that is dependent on the endogenous ligand. While it mainly targets the dopaminergic and serotonergic systems, it additionally exhibits an inﬂuence on elements involved glutamatergic/GABAergic signaling, thus rendering its use pertinent to this study [171].

3.5 Ethics

All studies were cleared by the local Ethics Committees of Hannover Medical School and the Medical Faculty the University of Leipzig (see Appendix C). All subjects were given the option to leave without any consequences for any reason during the study. The principal physician also reserved the right to withdraw any subjects from the study for urgent medical reasons. Part II

METHODS

31 Chapter 4

Magnetic Resonance Imaging and Spectroscopy

This chapter provides an introduction to the imaging methods used in this work. The basic principles of Magnetic Resonance Imaging [185–187], Magnetic Resonance Spectroscopy [188–191] and Quantitative Susceptibility Mapping [192–195] are presented succinctly in order to qualify the methodological and pathophysiological investigations presented in the subsequent chapters. Unless otherwise indicated, all the Magnetic Resonance Imaging and Spectroscopic data presented in this thesis were acquired by the author.

4.1 Nuclear Magnetic Resonance Imaging

4.1.1 Principles of Nuclear Magnetic Resonance Imaging

Magnetic Resonance Imaging (MRI) is a powerful scientific measuring instrument that allows the non-invasive photography of in-vivo anatomy and the quantitative evaluation of tissue properties in health and disease [185, 186]. The MRI technique is fundamentally based on principles of Nuclear Magnetic Resonance (NMR), which is a physical phenomenon that occurs when magnetically active nuclei are introduced into a static external magnetic field (B0). Elements that contain an odd number of protons or neutrons exhibit an angular momentum or ’spin’ that aligns itself with the static magnetic field while oscillating at a specific resonant frequency. In simple terms, the frequency of oscillation, or ’Larmour frequency’ (ω), depends on the number of unpaired proton- s/neutrons within the nucleus and the strength of the static magnetic field. If a second external magnetic field is applied (B1), for example, by radiofrequency stimulation, the atom oscillates in the direction of the newly applied magnetic field, causing the nuclei to

32 Magnetic Resonance Imaging and Spectroscopy 33

move out of alignment with the B0 field (Figure 4.1). Once the B1 field is turned off, the nuclei return to their pre-stimulated state in a process termed relaxation. Additionally, the transverse component of the magnetization induces a voltage according to Faraday’s law that can be detected by a radio-frequency antenna or a receiver coil tuned to the Larmor frequency. This detected signal is the NMR signal, which exhibits decay as the net magnetization (M) of nuclei return to their state equilibrium in a process termed free induction decay (FID).

NMR was first described independently by Edward Purcell and Felix Bloch in 1946 and was initially used by physicists to study the magnetic properties of atomic nuclei [185]. As a magnetically reactive nucleus exhibits defined mass and spin that are dependent on the number of particles within the atomic nucleus, different isotopes exhibit different NMR properties. Examples of magnetically reactive isotopes include 1H, 13C and 31P. Given that hydrogen is the most abundant nucleus in biological tissue, exhibiting strong and sensitive magnetic interactions, the NMR of protons (1H) is the most widely used technique to map properties of biological tissue. With the introduction of magnetic fields gradients in the 1970s, Paul Lauterbur as well as Peter Mansfield and Peter Grannel were able determine the spatial location of the induced signal, which served as a contrast for the reproduction of an image. By varying the parameters of radio-frequency pulse sequence, different contrasts could be generated mapping differences between tissues based on the relaxation properties of the hydrogen atoms.

Figure 4.1: Basic Physics of the NMR signal. (a) Nuclei with an odd number of protons and neutrons (e.g H1) exhibit a nuclear spin or angular momentum that aligns itself with an external magnetic field (B0). (B) When another magnetic field (B1) is applied, for example, by a radio-frequency pulse, the net magnetization vector is flipped at angle which produces two magnetization components (longitudinal and tranverse). Once the radio-frequency pulse is switched off, T1 recovery and T2/T2* decay occur. The figure was adapted from [185]. Magnetic Resonance Imaging and Spectroscopy 34

4.1.2 NMR properties of biological tissue environments

In general, the term relaxation describes how signals change with time and return to baseline. In NMR, the deteriorating signal of the nuclei returning to the pre-stimulated state can be described in terms of two separate processes, each with its own time constant and relationship to the B0 component vector [185, 186]. The T1 relaxation time is the time constant that describes the recovery of 63% of magnetization along the longitudinal

vector (i.e. along the axis of the B0 ﬁeld). In other words, the T1 relaxation time — or spin-lattice relaxation time — represents the time taken to regain the ability to create

an NMR signal. While the T1 relaxation time of cerebrospinal ﬂuid is about 4000ms [196], it can be inﬂuenced by the microenvironment of biological tissue varying between ≈1800ms for gray matter and ≈1000ms for white matter at 3T [197].

On the other hand, the T2 relaxation time is responsible for the broadening of the signal and is the time constant that describes the deterioration of the transverse magnetization

(i.e. perpendicular to the axis of the B0 field) to 37% [185, 186]. In tissue, T2 is usually much shorter that T1 by a factor of up to 1/10, as its mechanism involves the process of dephasing, or the oscillation of different hydrogen atoms at variable frequencies occurring as a result of the local magnetic field effects induced by neighboring atoms [185].

Nevertheless, another source of dephasing ensues as a result of local variation in the B0 field, which occurs mainly due to shielding of the main magnetic field by biological tissue structure variations (e.g. air tissue or bone tissue interfaces) and inhomogeneities of the magnetic field [185]. This source of dephasing causes the transverse magnetization to decay much faster than predicted by natural mechanisms and is usually represented by

the T2* relaxation time. T2* describes the ’effective’ or ’observed’ T2, which considers the two sources of dephasing induced by local magnetic field effects from neighboring

atoms and B0 non-uniformities.

4.1.3 NMR signal localization

Put simply, a magnetic resonance image is a representation of the NMR signal from hydrogen atoms in tissue. Initially, NMR signals were acquired from an individual element (pixel) or a column of elements in one dimension. The introduction of magnetic ﬁeld gradients allowed excitation of a select set of elements, thus paving the way for the acquisition of NMR signals from a plane (ie. 2 dimensional) or volume of elements (ie. 3 dimensional) [185, 186]. For example, for acquiring NMR data from an axial slice, a magnetic ﬁeld gradient is placed perpendicular to the slice (ie. in the z-direction, head

to foot), before a radio-frequency pulse (B1) with a narrow frequency range is applied to stimulate the plane (Figure 4.2). As a result, only those nuclei in the slice in which the Magnetic Resonance Imaging and Spectroscopy 35 range of frequencies are available would be excited, thus allowing the spatial encoding of hydrogen atoms in a specific plane where they are isolated from neighboring regions. However, this slice selective excitation strategy provides spatial resolution in one direction. Imaging a plane can be accomplished by using a second gradient field, which is turned on following the B1 pulse, causing nuclei to resonate differently along the gradient, i.e. they are spatially encoded in a second dimension. To obtain the third dimension, phase encoding gradients are applied for a limited time before signal acquisition. The phase encoding gradient modifies the spin resonance frequencies of the nuclei, in which the magnetization of external elements (voxels) will precess either faster or slower compared to the central element. The change in the frequency of precession changes the phase of the signal relative to the signals from neighboring positions. Repeated phase encoding directions can thus be used to determine the location of the precessing spins to create a three-dimensional image.

Figure 4.2: Selective excitation of an image slice by applying a shaped RF pulse and ﬁeld gradient at the same time. Adapted from [185, 186]. Magnetic Resonance Imaging and Spectroscopy 36

4.1.4 Pulse sequence design

An MRI pulse sequence is a programmed set of changing events related to magnetic field gradients, radio-frequency pulses and signal readout periods [185, 186]. Careful optimization of sequence parameters and the order of events is key in emphasizing specific tissue characteristics and generating different types of tissue contrast. Signals are typically collected when the NMR signal exhibits its strongest most coherent ’echo’. The time between the application of the radiofrequency excitation pulse and peak signal induced in the coil is generally referred to as the Echo Time (TE). As radio-frequency pulses are usually applied at specific intervals to excite different elements, the time between successive radio-frequency pulses, referred to as the Repetition Time (TR), is also an important parameter in emphasizing specific tissue characteristics. For example, a long

TR allows the protons in all of the tissues to relax back into alignment with the B0 ﬁeld.

At short TR, on the other hand, the protons do not fully relax back into B0 alignment before the next measurement is made, thus decreasing the signal from tissue.

Two types of conventional pulse sequences that are used to generate echoes are Spin echo (SE) and Gradient Echo (GRE) (Figure 4.3)[187]. SE sequences are the most commonly used pulse sequences. After a 90◦ radio-frequency pulse is applied, precessing nuclei usually go out of phase. In SE sequences, a 180◦ refocusing pulse is applied to reverse phase shifts that occur due to static field inhomogeneities and chemical shifts. Since the nuclei that now have a large negative phase are still oscillating at the same rate of decay, they begin to catch up with each other to go back into phase generating the peak echo signal. On the other hand, GRE sequences differ in that their flip angles are usually below 90◦ and also lack a 180◦ radio-frequency refocusing pulse. Utilizing low-flip angle excitation leads to more pronounced signal decay of the magnetization vector, thus allowing shorter TR/TE, shorter scan times and generation of new tissue contrasts.

4.1.5 Image reconstruction

To generate an actual image from an NMR signal, data is sampled multiple times measuring a large number of echoes to build the signal-to-noise ratio. The measured NMR signal is encoded in k-space which is a mathematical construct consisting of a matrix onto which frequency and phase data is mapped [185]. Typically, frequency and phase information are mapped along the x- and y-axis, respectively. To put this into perspective, in a conventional SE sequence, one echo generates one line of data in k-space corresponding to a single phase-encoding step. Although information at the peripheral parts of the k-space matrix contain ﬁne details with high spatial-resolution, they exhibit low-image signal. On the other hand, central elements in the matrix contain high-signal Magnetic Resonance Imaging and Spectroscopy 37

Figure 4.3: Conventional MRI pulse sequence diagrams (A) Spin echo and (B) Gradient Echo pulse sequence diagrams. Adapted from [187] information, though at lower resolution. Typically, a pulse sequence is designed in a way to maximize the signal-to-noise ratio to achieve an image with ﬁne detail and good contrast. Frequency information within k-space is then rendered into image space using Fourier transform decomposition (Figure 4.4).

Figure 4.4: MR image reconstruction from k-space. Plot of the k-space of a T1 weighted MRI image of a pineapple. The image was reconstructed from all the spatial frequences of k-space using inverse fourier transformation. Data was acquired by the author. Magnetic Resonance Imaging and Spectroscopy 38

4.1.6 Encoding physio-chemical properties in images

In general, T1, T2 and hydrogen distribution (also known as Proton Density), in addition to the adjustable parameters TE and TR are all considered to ultimately achieve a particular tissue contrast (Figure 4.5)[187]. Different tissues have different T1 and T2 times and as such they will exhibit different amplitudes of free induction decay depending

on the TR and TE parameters. Since white matter tissue has a shorter T1 (≈1000ms at 3T) than gray matter (≈1800ms at 3T) and cerebrospinal ﬂuid (≈4000ms at 3T) [196, 197] it recovers its longitudinal magnetization faster after the application of the radio frequency pulse, and thus generates a stronger NMR signal [187]. A stronger longitudinal magnetization leads to a stronger transverse magnetization and stronger signal on the next pulse. Therefore, a shorter TR emphasizes the diﬀerence between tissue

with diﬀerent T1 relaxation times, such that tissues with short T1 (e.g. white matter) will exhibit stronger a signal and a higher intensity on the reconstructed image, whereas

tissues with long T1 relaxation times (e.g. cerebrospinal ﬂuid) will exhibit smaller signals

and appear darker. As grey matter has an intermediate T1, it will produce a moderate intensity thus accentuating the diﬀerence between the cortex and white matter. Images

that emphasize diﬀerences based on the T1 relaxation times are designated T1-weighted images.

On the other hand, T2-weighted images are dependent on TE values [187]. Since white

matter has a shorter T2 than cerebrospinal ﬂuid and loses its transverse magnetization more rapidly, making the TE longer allows for the transverse magnetization to decay

more thus emphasizing diﬀerences based on T2 relaxation times [187]. In T2-weighted

images, tissues with shorter T2 times will appear darker, while tissues with longer T2

times appear brighter. If TE is short and TR is long, neither T1 nor T2 tissue differences are emphasized. Such an imaged is labeled as a Proton Density (PD) image since it emphasizes the differences in the distribution of hydrogens within tissue. PD images provides good contrast between gray (bright) and white (darker gray) matter, with little contrast between tissue and cerebrospinal flui. Pathological processes, such as demyelination or inflammation, often increase water content in tissues, which (a) decreases the

T1-weighted signal such that pathological white matter appears darker; (b) increases the

T2-weighted signal such that pathological white matter appears brighter making subtle changes easier to detect; and (c) increases the PD signal such that white matter also

appears brighter. In general, T1, T2 and PD weighted images provide good contrast and are commonly used to segment diﬀerent tissue and non-tissue classes. Magnetic Resonance Imaging and Spectroscopy 39

Figure 4.5: The relationship between TR/TE and the encoding of physio- chemical tissue properties as image contrasts. Adapted from [187]. Magnetic Resonance Imaging and Spectroscopy 40

4.2 Magnetic Resonance Spectroscopy

4.2.1 Principles of in-vivo NMR Spectroscopy

Magnetic Resonance Spectroscopy (MRS) is an analytical method that enables the non- invasive identification and quantification of biochemical substances within tissue [188]. Similar to MRI, MRS exploits the magnetic properties of atomic nuclei, however, instead of generating images of in-vivo anatomy, MRS is used to generate spectra from NMR sensitive isotopes providing physiological and biochemical information. Although many nuclei including 31P, 19F, 13C, 23N, could be used to obtain MR spectra, proton MRS (1H- MRS) is the most widely implemented technique used to probe tissue biochemistry given the natural abundance of hydrogen atoms in human tissue. Whereas MRI measures the distribution and interaction of water hydrogen atoms with tissue to map a single peak, 1H-MRS analyzes the signal of hydrogen atoms at different locations within a molecule.

Fundamentally, 1H-MRS is based on the chemical shift properties of atoms [191, 198]. Once placed in an external magnetic field, protons at different location within a molecule experience variable resonance frequencies, or chemical shifts, since they exhibit non- identical chemical environments (Figure 4.6)[198]. Shielding effects experienced by nuclei are caused by the electric field surrounding a molecule leading to different chemical

shifts of protons. Considering methane (CH4) as a simple example, the valence electrons around the methyl carbon generates a magnetic ﬁeld that shields the nearby protons

from experiencing the full force of the B0 ﬁeld. As a result, methane protons experience

an effective field that is slightly weaker than B0. Given the symmetric chemical structure of methane, its four protons experience chemical shifts at 0.23 ppm (relative to Tetram- ethylsilane). For larger molecules containing other moieties (e.g. Oxygen, Nitrogen), different protons experience different chemical shifts depending on their location within the molecule. As such, nuclei in different chemical environments can be distinguished on an NMR spectrum based on their resonant frequencies. Chemical shifts are usually represented in parts per million (ppm) of the resonant frequency, measured relative to a reference compound. In addition to experiencing chemical shifts, molecules also experience spin-spin coupling (or J-coupling) which results since nuclei experience the magnetic field of neighboring nuclei through the polarization of the electrons in the molecular bonds between them. Spin-spin coupling effects leads to a splitting of the NMR signal into two or more peaks (doublets or triplets), thus providing more information about the sample molecule.

Virtually all MRS studies are performed by collecting the NMR signal following the application of either a 90◦ pulse or an echo-type sequence [191]. The resonances from diﬀerent molecules within a tissue samples are collected simultaneously as a time-domain Magnetic Resonance Imaging and Spectroscopy 41

Figure 4.6: Chemical shift properties of N-Acetylaspartate. Shielding effects experienced by nuclei are caused by the electric field surrounding a molecule leading to different chemical shifts of protons. The figure demonstrates the chemical structure and the frequency domain spectrum of N-Acetylaspartate (NAA) acquired with TE=30ms at 3T. The three protons of the CH3 group (blue shading) exhibit similar shielding effects which causes their individual signals to add up leading to the prominent peak at 2.02ppm. Since the protons of the NH, CH and CH2 groups (green shading) are in close proximity, they interact via J-couling (red arrows) spiliting the peaks and leading to a more complex pattern of multiple peaks. The data was acquired by the authour and the figure was adapted from [191].

FID signal that is decomposed into a visually interpretable frequency domain spectrum via Fourier transform. Spectra in the frequency domain reflect the constituents of a tissue sample, in which the chemical shifts are used to distinguish different molecules within the sample, while the signal intensity represents the concentration of a specific metabolite.

4.2.2 Acquisition of 1H-MRS spectra

A typical MRS experiment usually begins with the acquisition of an anatomical image onto which a tissue sample could be localized for the acquisition of 1H-MRS spectra [189]. In general, the main objective of MRS studies is to detect weak signal of metabolites from a defined tissue sample. As such, the quality of the spectrum is critically dependent on pre-scan procedures that are applied to calibrate various aspects of MR system function. Typically, this consists of (a) adjusting transmitter and receiver gains, (b) setting the scanner frequency to be on-resonance with water; (c) suppressing the water signal which overlaps with almost all relevant metabolite peaks distorting that spectral baseline; and (d) enhancing the homogeneity of the magnetic field by adjusting the shim currents in Magnetic Resonance Imaging and Spectroscopy 42 a procedure called shimming [189]. Optimizing water suppression and the homogeneity of the magnetic field are critical in increasing the signal-to-noise ratio and resolution (linewidth), thus increasing the sensitivity and the specificity of the spectrum.

Both single-voxel and multi-voxel spectroscopy techniques are usually used for acquisition of spectra from localized regions of interest. Depending on the objective of the study, both long and short TEs could be used to probe specific metabolites or improve the quality of the results [191]. In single-voxel spectroscopy (SVS), NMR signals are usually acquired from a tissue samples with a resolution of approximately of 1–8cm3 using a combination of slice-selective excitation in three dimensions while the radio-frequency pulse is turned on. Two main techniques used for the acquisition of SVS spectra include point-resolved spectroscopy (PRESS) and stimulated echo acquisition mode (STEAM) [188]. PRESS is the most used SVS technique in which spectra are acquired using one 90◦ pulse followed by two 180◦ pulses. Pulses are applied at the same time as the different field gradients, thus the emitted signal is a spin-echo. On the other hand, STEAM utilizes three 90◦ pulses that are also applied simultaneously with the different field gradients. Although PRESS offers a better signal-to-noise ratio, STEAM allows shorter TEs to compensate for reduced SNR. STEAM is usually used as an alternative to PRESS when short echo times and chemical shift artifacts are of concern. For both sequences, spoiler gradients are used to dephase signals external to the localized region of interest thereby reducing their signal.

On the other hand, multi-voxel spectroscopy — also known as MRS imaging (MRSI) or chemical shift imaging — is used to simultaneously acquire multiple spatially arrayed spectra from slices or volumes [191]. MRSI implements both spectroscopic and imaging techniques to produce spatially localized spectra which could be used to generating maps encoding metabolite quantities. Unlike MRI however, phase encoding gradients are used to map spatial information with the absence of frequency-encoding gradients, such that chemical shift information is retained to generate a spectroscopy grid. MRSI data is acquired with sequences similar to PRESS and STEAM, except that phase encoding gradients are used in one, two or three dimensions to sample k-space after the application of the radio-frequency pulse [191]. The main advantage of MRSI is that it allows the examination of larger regions of interest with spatial encoding of metabolic quantities in many volume elements, although it suﬀers from longer acquisition times and imprecise quantitation due to voxel bleeding.

In general, MRS data is either obtained at short TE (≈20-40ms) or long TE (≈135- 180ms) (Figure 4.7)[191]. Data acquired with short TEs exhibit a high signal-to-noise ratio leading to spectra with more metabolite peaks (e.g. myo-inositol and glutamate). In contrast, data acquired at long TEs exhibit lower signal-to-noise ratio and variable Magnetic Resonance Imaging and Spectroscopy 43

amount of T2 weighting, but are usually better resolved spectra with ﬂatter baselines. Whereas short TE allows the detection of more metabolites, acquisition at long TE allows the detection of metabolites that exhibit large overlaps with other metabolites (e.g. lactate).

Figure 4.7: The effect of echo time on metabolite detectability. (a) PRESS data acquired at short TE generates more complex spectra that contain more metabolite peaks that are not visible at longer TEs (b,c). Longer TEs are usually used when the detection of Lactate is particularly important. Cre=Creatine, Cho=Choline compounds, Gln=Glutamine, Glu=Glutamate, Lac=Lactate, mI=Myo-Inositol, NAA= N- acetylaspartate. The figure was retrieved from [191] with modifications.

4.2.3 Time domain signal processing

Typically, in-vivo NMR spectroscopy data is recorded as RAW time-domain FID signals that must undergo a number of signal processing steps before quantiﬁable spectra are achieved [189, 190]. Following time-domain signal processing, fourier transform decomposition is applied to obtain frequency domain spectra which usually undergo further phase and baseline correction before quantitation. Time-domain signal processing steps include: Magnetic Resonance Imaging and Spectroscopy 44

• Apodization (or time domain ﬁltering): which is performed to improve the spectral resolution and sensitivity of the signal, and is accomplished by removing high- frequency noise which would be detrimental to spectral peak detection.

• Zero-ﬁlling: which involves appending zeroes to the end of the FID resulting in great improvements in the resolution of the spectrum.

• Eddy-current correction: which eliminates artifacts in the spectral lineshape caused by transient currents induced with the application of the pulsed ﬁeld gradient by using a reference signal (e.g. unsuppressed water).

• Frequency and phase correction: which limits artifactual broadening of spectral peaks, distortion of spectral lineshapes, and reductions in the signal-to-noise

ratio, which may be caused by rapid bulk motion and temporal drifts in the B0 ﬁeld.

Figure 4.8: 1H-MRS time domain signal processing. Following a number of preprocessing steps that most commonly include apodization, zero-filling, eddy-current correction and frequency and phase correction, quantifiable frequency domain spectra are achieved via the Fourier transformation of the pre-processed RAW time-domain FID signal. The figure was retrieved from [189] with modifications. Magnetic Resonance Imaging and Spectroscopy 45

4.2.4 Spectral quantitation

In general, the primary aim of spectral analysis is to determine the concentration of compounds present in the spectra. As the area under a spectral peak is proportional to the metabolite concentration, spectral quantitation involves measuring the intensity of spectral peaks and converting these measure to a metabolite concentration estimates [190]. However, peak intensities are meaningless quantities with arbitrary units that are inﬂuenced by multiple experimental factors including voxel size and localization, radiofrequency coil sensitivity, receiver gain, and others. As such, the conversion of the arbitrary peak intensity into a meaningful value is critical in obtaining an absolute quantitative estimate of a given metabolite. This is mainly achieved by using the peak intensity of a reference metabolite whose concentration is known. The absolute concentration of the metabolite of interest can thus be calculated based on the relative peak intensities based on:

|C | I N Met = Met · Met (4.1) |CRef | IRef NRef where C, I and N denote the absolute concentration (in molar or molal units), spectral peak intensity and the number of protons, respectively, for both the metabolite of interest (Met) and the reference compound [190].

There are a number of techniques used for signal referencing including internal referencing (metabolite or water), external referencing, phantom replacement and electrical referencing [190]. Given their feasibility and ease of implementation, internal metabolite referencing methods are widely used as they do not require additional data acquisition. This involves expressing the signal intensity of the metabolite of interest (e.g. glutamate) as a ratio to the signal intensity of another metabolite with a prominent peak (e.g. total creatine). However, internal metabolite referencing methods provide a relative measures that is not absolute and suﬀer from serious limitations if the reference compound exhibits instability due to pathology. Since the water signal is relatively stable in many pathologies, internal water referencing methods overcome this limitation and usually involve the acquisition of water-unsuppressed spectra from the same region of interest. Given the abundance of water in brain tissue, it exhibits a signal that is about 100,000 greater than that of other metabolites, therefore, only a few signal averages are required. This allows acquisition in a short period of time, making the internal water referencing method a feasible and common technique for accurate 1H-MRS metabolite quantitation.

In the external referencing methods, similar principles are applied, but instead of measuring specific peaks from a biological sample, signals are acquired from an in-vitro sample Magnetic Resonance Imaging and Spectroscopy 46 with a known concentration [190]. However, since the reference signal is not usually acquired from the spatial location, the method may suffer from drawbacks as the reference signal and the metabolite of interest may experience variable magnetic homogeneities. The phantom replacement method has an advantage in which the electrical properties of the sample and the position within the magnetic field are carefully considered to acquire data with similar parameters. Although both methods have an advantage in that the concentration of the reference is accurately known, they are difficult to implement and are usually regarded as infeasible for clinical studies.

Figure 4.9: Referencing methods for absolute metabolite quantitation. The ﬁgure was adapted from [190].

In general, 1H-MRS spectra exhibit complex line-shapes, multiple overlapping peaks and signal contamination from macro-molecules [190]. Therefore, estimating the peak intensity for a specific metabolite is not as trivial as measuring the area under the curve as in peak integration methods. This method is severely limiting in 1H-MRS since it cannot effectively separate the contributions of overlapping or partially-overlapping peaks. Other methods developed to achieve reliable estimates of metabolite peaks include model peak fitting and basis spectrum fitting. In the model peak fitting technique, a model function that best describes the shape of a peak of interest is chosen to fit the data. Commonly used model functions include the Lorenzian, Gaussian and Voigt lineshapes. Fits are usually improved with the inclusion of prior knowledge and the quality of fitting is assessed by inspecting the difference between the data and the fitted curve (i.e. residual), which should be kept to a minimum.

Basis spectrum fitting is the most sophisticated and most commonly used spectral fitting technique [190] (Figure 4.10). In essence, this technique is intuitive and is based on the logical principle that the spectrum can be modeled as the linear combination of the spectrum of each individual metabolite if the shape is known [190]. In other words, the entire spectrum is fit to a series of basis functions that are determined either exper- imentally or via simulations. Both types have their advantages and disadvantages, but in both cases, the basis set is specific to the pulse sequence and timing parameters of Magnetic Resonance Imaging and Spectroscopy 47

Figure 4.10: The linear combination basis spectrum fitting model. Demonstration of the individual metabolite peaks achieved for single-voxel 1H- MRS PRESS (TE=30ms) data acquired from the anterior cingulate cortex. The spectrum was fit using the LCModel software [199] with internal referencing to the water signal. Asp=Aspartate, Cre=Creatine, tCho=Choline compounds, GABA= γ-Aminobutyric acid, Gln=Glutamine, Glu=Glutamate, Gua=Guanine, Lac=Lactate, mIno=Myo-Inositol, MM=Macromolecules, tNAA= N-acetylaspartate plus N-acetylaspartateglutamate, Tau=Taurine. The resented data was acquired by the author at 3T and fit with LCModel [199]. the acquisition. Once basis sets are obtained, fitting is performed in the the time- or frequency- domain as a linear combination of the metabolite spectra, where the relative metabolite concentrations are estimated based on the amplitude weightings in the linear combination producing the best fit. Basis spectrum fitting has been shown to outper- form the model peak fitting technique, leading to higher quantitation accuracy due to its relative insensitivity to overlapping peaks. Magnetic Resonance Imaging and Spectroscopy 48

4.2.5 1H-MRS metabolites of the human brain

1H-MRS spectra of the human brain typically reveal four major peaks in addition to water. The major metabolites [188] visible in 1H-MRS spectra at short echo time are discussed below:

• N-acetylaspartate (NAA): NAA is the second most concentrated neurochemical in the human brain. It is mainly found in neurons [200] and to some degree in oligodendrocytes and myelin [201] and has several functions that include ﬂuid balance and the contribution of precursors for energy production and the synthesis of myelin and the neurotransmitter N-acetylaspartyl-glutamate. On an 1H-MRS spectrum, NAA exhibits the most prominent peak at 2.02ppm. Given its function and the prominence of its peak, NAA is considered as one of the most reliable indicators of neuronal integrity. For example, neuronal degradation as a result of malignant neoplasms leads to visible decrease in the NAA peak [202].

• Creatine: Creatine (main peak at 3.02ppm) occurs in both phosphorylated and unphosphorylated forms and plays an important role in storing phosphate groups and supplying energy to myocytes and neurons [203]. As a result, creatine is considered a marker for energetic systems and intracellular metabolism. Since its concentrations are relatively stable, creatine is commonly used as an internal reference standard, although it may exhibit alterations in speciﬁc conditions such as brain tumours.

• Choline: As an essential nutrient, choline has multiple roles that include the synthesis of phosphatidylcholine for cell membrane components and serves as a precursor for the neurotransmitter acetylcholine [204]. Choline’s prominent peak occurs at 3.22 and its quantitation represents the sum of choline containing compounds which include phosphocholine and glycerophosphocholine with small contributions from acetylcholine and citicoline. Choline is considered as a marker for cell membrane turnover reﬂecting cellular proliferation and is altered in a variety of conditions. For example, regions with acute demyelination in multiple sclerosis exhibit increases in choline concentrations.

• Myo-inositol: Myo-inositol exhibits a peak resonance at 3.65ppm that reﬂects a pool of inositol containing compounds. It is primarily synthesized in astrocytes and as such is considered a marker for the integrity of glial cells [205]. Myo-inositol is involved in the activation of protein C kinase and functions as an osmolyte. Its elevations are considered a biomarker in Alzheimer’s disease. Magnetic Resonance Imaging and Spectroscopy 49

• Glutamate and Glutamine: Glutamate is the primary excitatory neurotransmitters and is the most concentrated neurochemical in the human brain [206]. On the other hand, glutamine is a non-neuroactive substance that exhibits close metabolic links to glutamate. Given that neurons lack the necessary enzymes to synthesize glutamate, astrocytic-neuronal coupling mechanisms help maintain the glutamate-glutamine metabolic cycle to maintain an adequate supply of glutamate in neurons for further release. As the amide- to carboxy-group conversion presents the only diﬀerence between the chemical structures of glutamate and glutamine (Figure 4.11), their methelyne and methine J-coupled resonances produce multiplet peaks (2.08, 2.34, 3.74 ppm for glutamate; 2.12, 2.44, and 3.75 ppm for glutamine) that are diﬃcult distinguish on spectra with a limited linewidth and signal-to-noise ratio. Therefore, the composite signal of glutamate and glutamine (Glx) is often used as a marker of the integrity of glutamatergic neurons and astrocytes.

• γ-aminobutyric acid (GABA): GABA is the primary inhibitory neurotransmitter in the human and also exhibits a similar structure (Figure 4.11) and close biochemical link to both glutamate and glutamine [205]. A GABA multiplet peak occurring at 3.01 ppm is normally obscured by the creatine signal at 3.03 ppm. As a result of this overlap, GABA peaks are diﬃcult to quantify reliably using conventional 1H-MRS. Specialized editing pulse sequences are usually used to isolate the GABA peaks from overlapping resonance for reliable quantitation.

Figure 4.11: GABA, Glutamate, Glutamine cycling. Red/green shading indicates chemical structure diﬀerences between glutamate/glutamine and glutamate/- GABA. See Chapter 8 for further details. Magnetic Resonance Imaging and Spectroscopy 50

4.3 Quantitative Susceptibility Mapping

4.3.1 Magnetic Susceptibility

Magnetic susceptibility (χ) is a physical quantity that describes the extent to which a substance becomes polarized in an external magnetic field [207]. While the acquired NMR signal in MRI and MRS originates from nuclear particles, the magnetization experienced by circulating electrons is the source of the magnetic susceptibility signal. In essence, magnetic susceptibility is an intrinsic elemental property that is dependent on the distribution of an atom’s electrons. As atomic orbitals can occupy a maximum of two electrons, paired electrons exhibit different spins which cancel out and lead to no induction of a dipole moment. On the other hand, the presence of an unpaired electron leads to the induction of a net dipole moment. Materials are usually classified as exhibiting diamagnetic or paramagnetic properties based on the strength of their polarization when applied to external magnetic fields (Figure 4.12). Diamagnetic material have a negative susceptibility originating from currents induced by their paired circulating electrons. This leads them to exhibit a net polarization that opposes the applied field and causes a reduction in the effective field within the object. As a result of their unpaired electrons, paramagnetic material have a positive susceptibility, leading to their polarization in the same direction as the static field and augmenting the effective field within the object. Biological tissues can be either diamagnetic or paramagnetic depending on their atomic microstructure.

Figure 4.12: Magnetic Susceptibility of diamagnetic and paramagnetic material As a result of their intrinsic distribution of their electrons, diamagnetic material exhibit a negative susceptibility that leads to their polarization (J) opposite to the magnetic field. On the other hand, paramagnetic material exhibit a positive susceptibility that leads to their net polarization in the direction of the magnetic field, thus augmenting the effective field experienced by the object. Adapted from http://mriquestions.com/what-is-susceptibility.html Magnetic Resonance Imaging and Spectroscopy 51

Despite the fact that magnetic susceptibility is a central intrinsic property of matter that may be useful as a contrast mechanism similar to T1 and T2 relaxation constants, it was long considered as a nuisance signal rather than a contrast mechanism in MRI [192]. This was the case since minute variations in magnetic susceptibility between tissues lead to local field distortions that cause MR signal loss and image artifacts [208, 209]. Field perturbations and Larmour frequency shifts and are largely due to significant differences −6 between the susceptibility of water (χwater = −9.05 × 10 SI units) and air (χair = 0.37 × 10−5 SI units, e.g. nasal cavity and ear canals) and the presence of paramagnetic deposits [195]. Nevertheless, recent MRI methods have capitalized on this susceptibility- induced signal loss to reconstruct data from the phase signal, exposing new contrasts that reveal important details of tissue anatomy.

4.3.2 Phase Imaging

In general, the measured k-space signal results in two data streams with 90◦ phase diﬀerence which are digitized as complex data with real and imaginary parts [185]. These signal are usually corrupted by a noise signal exhibiting Gaussian probability distribution that persist even following Fourier transform to image space. As such, both the real and imaginary parts of the complex data are used to reconstruct images of diﬀerent clinical utility. Typically, the magnitude image is used to maximize the signal-to-noise ratio and is calculated after the Fourier transform as pReal2 + Imaginary2 at each voxel which yields an image with a Rician noise probability distribution [210].

On the other hand, the phase image measures local frequency offset, which reflects the magnetic field perturbation, and is calculated as arctan(Imaginary/Real) at each voxel [195, 211]. Since the sine and cosine functions are periodic with a period of π, the reconstructed phase images exhibit phase wrapping such that any angle outside the range between - π and + π will be folded back. This phase wrapping is a source of severe artifacts that limits the utility of the phase images without further processing. Additionally, phase values within the brain are also influenced by magnetic field inhomogeneities, susceptibility differences between tissue and air, and also by the long-range magnetic field generated by the human body (dipole field) [195]. These sources of background phase increase the phase wrapping and limit local tissue contrast. To mitigate the visible artifacts present in the phase image data, varied algorithms have been proposed for both phase unwrapping and background field removal (see section 4.3.4). The application of these preprocessing techniques enhances phase contrasts to allow the differentiation of tissue types based on susceptibility 4.13. Magnetic Resonance Imaging and Spectroscopy 52

Figure 4.13: Tissue phase imaging 3T magnitude (a) and phase (b) data acquired from a Gilles de la Tourette syndrome patient. Phase wraps and image distortion artifacts due to magnetic field inhomogeneities are clearly visible on the RAW phase image. Following phase unwrapping (c) and background field removal (d), the processed phase image directly represents susceptibility induced local frequency offsets reflecting the magnetic field perturbation. The data was acquired by the authour.

4.3.3 Susceptibility Weighted Imaging

The visualization of magnetic susceptibility tissue differences is most commonly achieved via gradient echo data acquired using a single- or multi-echo spoiled-gradient-recalled- echo (GRE) sequence [195, 212]. Although filtered phase images exhibit illustrative tissue contrasts, the differentiation of neighboring tissues with large susceptibility differences remains limited. The Susceptibility-Weighted Imaging (SWI) technique capitalizes on the contrast inherent in the magnitude and phase images to improve susceptibility contrast by combining both images [194]. Initially, SWI was described as MR-Venography given that the the aim of obtaining these images was to enhance the visualization of small vessels by taking advantage of the paramagnetic properties of deoxyhemoglobin [209]. In this approach a brain tissue phase mask is constructed with amplitudes between 0 and 1 to suppress signal intensities of certain values. The phase mask is then multiplied

by the T2* weighted magnitude image a number of times to generate an SWI image with an optimal contrast between tissues of diﬀerent susceptibilities, where the number of multiplications is usually dependent on phase diﬀerences and the contrast-to-noise ratio [195, 209]. SWI could be used to enhance contrast between grey-/white-matter and water/fat, in addition to enhancing the contrast of paramagnetic elements exhibiting high densities in the brain (e.g. iron). SWI has a number of applications in the clinical setting including the diagnosis of cerebral vascular pathology and the detection of abnormal accumulation of mineral deposition. Magnetic Resonance Imaging and Spectroscopy 53

Figure 4.14: Susceptibility-Weighted Imaging The multiplication of the magnitude image (a) with the unwrapped and ﬁltered phase yields a susceptibility weighted image (b) that is sensitive to venous blood, hemorrhage and iron storage. The technique was initially named MR venography as cerebral venous vessels could be visualized as illustrated in the minimum intensity projection map (c) over a stack of 10 slices. Data was acquired from a Gilles de la Tourette syndrome patient at 3T.

4.3.4 Quantitative Susceptibility Mapping

The non-locality and orientation dependence of phase are an intrinsic limitation that lead to variations in signal phase, such that certain biological tissues can have positive or negative values depending on their orientation in the B0 field [195]. Additionally, due to the convoluting effect of the long-range dipole field, it is difficult and not always reliable to differentiate tissues with diamagnetic and paramagnetic properties. As such, SWI is in essence a qualitative approach that does not provide a quantitative measure of magnetic susceptibility.

Quantitative Susceptibility Mapping (QSM) is a recently established technique that allows the determination of the intrinsic magnetic susceptibility properties of tissues based on signal phase [193, 213]. Considering the magnetization experienced by an imaging voxel, the magnetic field can be expressed as the superposition of the combination of dipole fields generated by all voxels [195]. Since the superposition of the magnetic field is linear and the dipole fields do not change from from one voxel unit to another, the relationship between the spatial distribution of the experienced magnetization (i.e. susceptibility) and the magnetic field can be expressed as a convolution [195]. However, due to the presence of zeroes in the convolution kernel, the mapping of susceptibility distribution from magnetic field is an ill-posed inverse problem. Multiple techniques have Magnetic Resonance Imaging and Spectroscopy 54 been proposed to solve the ill-posed inverse problem. These are mainly based on the use of single-orientation regularization approaches (e.g. HEIDI, MEDI, TKD) or the use of repeated measurements with subject rotation within the magnetic field (e.g. COS- MOS) [192]. Given the difficulty of acquiring GRE data with multiple orientations in the clinical setting, single orientation approaches are more commonly used due to their feasibility.

Overall, the calculation of quantitative susceptibility using single orientation approaches maps involves a number of sequential steps (Figure 4.15) that are summarized below:

• Coil combination: MR signal detection is typically achieved via multi-channel phase array coils. Following acquisition, a single coil-combined image is reconstructed from with MR data detected from each separate channel [192]. Though a number of methods have been proposed, some coil-combination algorithms suﬀer from the occurrence of pole artifacts or open-ended fringe lines in the phase images (see Chapter7). Therefore, care in employing an artifact free coil-combination methods is essential for the accurate calculation of susceptibility.

• Field map estimation: Since phase aliasing results in abrupt artificial jumps of 2π, phase wraps should be accounted for to obtain a phase map representative of susceptibility induced field offsets reflecting magnetic field perturbations. A number of phase unwrapping techniques have been utilized including path-following approaches, minimum norm solving, or filtering [192].

• Background field removal: The magnetic field perturbation experienced by a given voxel is influenced by the static magnetic field, magnetic field inhomogeneities, and the magnetic susceptibility distribution from outside sources [192]. Such background field contributions are approximately an order of magnitude stronger than the internal field contributions, and therefore, efficient strategies for the removal of background field should be employed for accurate susceptibility calculation. Traditional high-filtering operations are used to separate background phase from brain tissue phase with the assumption that background phase is smooth with low-spatial frequencies [195]. However, traditional highpass-filtering approaches are not without their limitations as they may also remove a substantial amount of the low-frequency tissue phase components along with background phase. Other proposed solutions that overcome these limitations include the So- phisticated Harmonic Artifact Reduction for Phase Data for phase data (SHARP) and its variants [192, 214–216], the projection onto dipole field method (PDF) [217], and harmonic phase removal using the laplacian operator (HARPRELLA) [218]. Magnetic Resonance Imaging and Spectroscopy 55

Figure 4.15: Quantitative Susceptibility Mapping Image processing. See text for details. Adapted from [192].

• Field-to-susceptibility inversion: Though the relationship between filtered phase image, which reflects the magnetic field perturbation, and magnetic susceptibility can be expressed as a simple convolution, the presence of zeroes in the convolution kernel renders the field-to-susceptibility inversion as an ill-posed inverse problem. Multiple approaches have been developed to solve this inverse problem using single- and multi-orientation MRI acquisition approaches. Single-orientation approaches mainly incorporate regularization and can be broadly divided into (a) non-iterative k-space techniques; iterative image space-based optimization approaches applying regularization to (b) all Fourier coefficients and (c) only to Fourier coefficients close to the magic angle cone [192].

• Tissue referencing: Given that the k-space origin of the computed susceptibility map is undefined, the maps have an unknown offset such that QSM data provide relative rather than absolute values of magnetic susceptibility [192]. As a result, the choice of a suitable reference tissue with comparable susceptibility across subjects and disease states is essential for unbiased group comparisons. Reference tissues that have been used in the past include cerebrospinal fluid, white-matter, the internal capsule, and the whole brain. Recent work, however, has indicated that the internal capsule and cerebrospinal fluid appear to be the most suitable reference tissues as they exhibit minimal age- and disease-dependent variations [219] Magnetic Resonance Imaging and Spectroscopy 56

4.3.5 Sources of magnetic susceptibility contrast

Fundamentally, the susceptibility values measured via the QSM technique are determined by the molecular composition of an imaging voxel, which contain an ensemble of molecules within complex cellular environments exhibiting diﬀerent susceptibilities [195]. As cells are fundamentally made of similar constituents (e.g. lipid membrane, cytosol, organelles), the relative portion of cellular contents within diﬀerent tissues, particularly those with a strong contrast, are what determines the susceptibility contrast [220].

Considering that the majority of human brain tissue exhibits susceptibilities within ±20% of the susceptibility of water, the two major sources of susceptibility that lead to striking contrasts and visible tissue diﬀerences are myelin and iron [192]. Myelin, which is mainly composed of lipids (e.g. pholipids and sphingolipids) with susceptibilities that are more diamagnetic than water, has been demonstrated to contribute to bulk susceptibility in white matter by a number of studies [192, 215]. For example, Liu et al. [221] demonstrated the disappearance of the strong grey and white matter contrast by utilizing a transgenic mouse model characterized by the loss of myelin.

Iron, on the other hand, is sequestered by ferritin and hemosiderin (non-heme iron), and is also bound to heme-proteins (heme iron) which are mainly paramagnetic [222, 223]. While heme iron is concentrated in the blood dominantly influencing the susceptibility of vessels (see Figure 4.14C), the seminal work of Hallgren and Sourander [224] demonstrated that non-heme iron is preferentially concentrated in deep grey matter nuclei (see Chapter 9 for further details). Despite the presence of other paramagnetic elements (e.g. manganese, copper), the paramagnetic contribution of non-heme iron is about 30 times greater than the combination of other trace elements [225, 226], thus rendering QSM as an interesting application for mapping iron densities (Figure 4.16). The use of QSM as a surrogate measure to estimate iron concentrations in deep matter nuclei has been validated by a number of post moretem studies that include Perls’ stain [225, 227], X-ray fluorescence [228] and proton induced X-ray emission [229]. Consequently, QSM offers the possibility of investigating brain iron in various neurological and psychiatric disorders. Magnetic Resonance Imaging and Spectroscopy 57

Figure 4.16: Correspondence between Perls’ stain and QSM in revealing iron deposition in the deep grey matter nuclei. The Perls’ stain method uses a mixture of potassium ferrocyanide and hydrochloric acid to stain tissue [230]. Iron stored in ferritin reacts with the ferrocyanide to form an insoluble blue dye as depicted in the left panel. Greater intensities of blueness are clearly visible in deep grey matter nuclei and exhibit a correspondence with QSM as illustrated in the right panel. QSM data was acquired by the author and Perls’ stain images were retrieved from [230]. Chapter 5

Clinical Assessment

This chapter provides a succinct summary of the clinical methods used in this study to assess both motor and non-motor features of GTS, in addition to the methods used to classify the patients into diﬀerent subcategories. Collection of the clinical data was performed by Sarah Gerasch.

To undertake a thorough evaluation of the clinical characteristics of patients with GTS, a variety of clinical rating scales are usually used to assess a patients’ motor features and psychiatric comorbidities. In the clinical setting, patients usually undergo a neuropsychiatric interview in addition to a comprehensive clinical assessment battery to gather information on tics, premonitory urges, quality of life and psychiatric comorbidities (OCD, ADHD, depression, anxiety, and autism). In order to achieve a reliable diagnosis of psychiatric comorbidities and avoid a bias due to erroneous self-report, both self- and physician-rated assessment tools were used to diagnose each condition as described in the following sections.

5.1 Tics, premonitory urges and quality of life

The frequency, intensity and complexity of tics, in addition to overall impairment were assessed using the Yale Global Tic Severity Scale (YGTSS) [174]. The YGTSS-Total Tic Score (YGTSS-TTS, range 0-50) includes the motor-tic score (range 0-25) and the vocal-tic score (range 0-25). The YGTSS global score (YGTSS-GS, range 0-100) includes the YGTSS-TTS and the overall impairment score (range, 0-50). As the YGTSS assesses tic severity over a period of 7 days, we additionally collected tic-severity data using the Modiﬁed Rush Video-Based Tic Scale (RVTRS) on the day of MRI data acquisition [175] 58 Clinical Assessment 59 in order to measure current tic frequency exactly on the day of the neurochemical and elemental measurements. Premonitory urges were assessed using the Premonitory Urge for Tics Scale (PUTS, range 0-36) [176], and health related quality of life was assessed using the Gilles de la Tourette syndrome Quality of Life Scale [177].

5.2 Obsessive Compulsive Disorder

Three different rating scales were used to diagnose OCD. The diagnosis of OCD was made, if the patient (a) fulfilled the M.I.N.I. International Neuropsychiatric Interview 5.0 ‘OCD current’ category [173] and (b) reached the cut-off values of the Yale-Brown Obsessive Compulsive Scale (Y-BOCS) [178] and/or the Revised Obsessive-Compulsive Inventory (OCI-R) [183].

5.3 Attention defecit-hyperactivity disorder

As suggested by previous work [231, 232], the diagnosis of ADHD was made based on results obtained from (a) DSM-IV symptom list for ADHD [233]; (b) Wender Utah Rat- ing Scale (WURS-k) [234, 235]; (c) and Conners’ Adult ADHD Rating Scale (CAARS) [180]. Patients were diagnosed with ADHD if they satisﬁed respective cut-oﬀ values of WURS-k or DSM-IV symptom list and of ≥4/8 CAARS categories.

5.4 Depression

For the diagnosis of depression, results of the (a) M.I.N.I., (b) Montgomery Äsberg Depression Rating Scale (MADRS) [179], and (c) Beck Depression Inventory II (BDI-II) [181, 182, 236] were taken into account as suggested by previous work [237–239]. Patients were diagnosed as depressive if they (a) fulﬁlled the M.I.N.I. category ‘major depressive episode current’ and reached a BDI-II score ≥18 and/or a MADRS score ≥7 or (b) reached both a BDI-II score ≥18 and a MADRS score ≥7.

5.5 Anxiety

For the diagnosis of anxiety disorder, results of both M.I.N.I. and the Beck Anxiety Inventory (BAI) [240, 241] were taken into consideration [242]. Anxiety disorder was diagnosed if the patients (a) reached a BAI score ≥26; or (b) reached a BAI score Clinical Assessment 60

≥8 while fulﬁlling the M.I.N.I categories ‘panic disorder current’, ‘agoraphobia current’, ‘social phobia current’, and/or ‘generalized anxiety disorder current’.

5.6 GTS subgroup classiﬁcation

Depending on the presence of comorbid diagnoses as defined above, the following subgroup classifications were used: GTS-only (tics without any other comorbid conditions) and GTS-plus (tics with one or more associated comorbidity). Within the GTS-plus category, the following sub-classifications were constructed based on type and number of the associated psychiatric comorbidities: (i) GTS-OCD (excluding ADHD, but possibly other comorbidities), (ii) GTS-ADHD (excluding OCD, but possibly other comorbidities), and (iii) GTS-OCD-ADHD (and possibly other comorbidities). Part III

Methodological Investigations

61 Chapter 6

Absolute metabolite quantitation

This chapter presents an 1H-MRS methodological investigation that was conducted to assess the test-retest reliability of absolute metabolite quantitation with partial volume correction. This study was conducted initially to interrogate the sensitivity of the spectral quantitation technique and to optimize its accuracy before the statistical comparison of patient/control metabolite levels (Chapter 8). To generalize the utility of the technique, the method was tested on a test-retest healthy control sample acquired during the course of the thesis work and another unrelated test-retest dataset that was acquired as part of a previous study [243]. The results led to improvements in accuracy of spectral quantitation and served as benchmark for (a) absolute metabolite quantitation with partial volume correction and (b) the assessment of the repeatability of quantitation as Chapter 8. The presented work is an expanded format of an abstract that underwent a peer-review process and was presented at the 2015 annual meeting of the International Society of Magnetic Resonance Imaging in Medicine in Toronto, Ontario, Canada: Kanaan AS, et al., Test- retest quantitation of absolute metabolite concentrations with partial-volume correction using diﬀerent segmentation methods, Proc. Intl. Soc. Mag. Reson. Med. 23, 2015, Program No. 1974 [244].

6.1 Introduction

The use of water as an internal concentration standard, is a feasible and common techniques for the accurate quantitation of 1H-MRS absolute metabolite concentrations. However, large spectroscopic voxels commonly contain a mixture of grey-matter, white matter and cerebrospinal ﬂuid at variable proportions across subjects. Additionally, metabolite signal only arise from the grey-matter and white-matter tissue compartments,

which also exhibit diﬀerent T1 and T2 relaxation times [245, 246]. Therefore, to achieve reliable estimates of absolute metabolite concentrations, diﬀerences in within voxel tissue compartmentation, relative water content and relaxation properties should be taken into account. Typically, relative water content and relaxation times can be taken from the

62 Absolute metabolite quantitation 63 literature, however the determination of within voxel tissue content involves the segmentation of anatomical images of appropriate contrast into the three main tissue classes. Previous work has indicated that different segmentation algorithms yield variable estimates of tissue fractions, which would influence estimates of the absolute metabolite concentration within grey-matter and white-matter [247, 248]. In this work, we investigated the test-retest reliability of the three most commonly used segmentation algorithms and the influence of partial volume correction on absolute metabolite estimates.

6.2 Methods

To investigate the variability of segmentation algorithms and their influence on quantitation correction using different acquisition schemes and at different time-points, anatomical and spectroscopic data were acquired from two independent samples at two different time-points. The study was approved by the local Ethics Committee and all participants had given informed written consent prior to the examination.

Dataset 1: The ﬁrst dataset included 20 healthy controls (age = 37.76±10.2 years, three female) with anatomical data acquired via MP2RAGE (TR=5s, TE=3.93ms, FOV=192 mm, inversion times=0.7/2.5s, 256x256 acq. matrix, 1.0mm3 resolution) and spectroscopic data acquired via PRESS (TE= 30ms,TR=3000ms, 80 supp acq., 16 unsuppressed acquisitions, FASTESTMAP shimming) from the anterior mid-cingulate cortex (aMCC, 6.4ml) and the bilateral thalamus (THA, 7.2ml) (3T MAGNETOM Verio, Siemens, Er- langen, Germany). This dataset was acquired by the authour for the purposes of this thesis and comprehensive imaging parameters are outlined in Appendix D. The inbuilt AutoAlignHead Siemens sequence was used to align the geometry of the voxel to a standard [249]. On the retest scan, the voxels were automatically localized to the same region-of-interest using the saved voxel geometry information from the ﬁrst scan.

Dataset 2: The second dataset included 10 healthy controls (age = 28.3±2.3 years, all male) with anatomical data acquired via a Magnetization Prepared Rapid Acquisition GRE (MPRAGE, TR=1s, TE=2.7ms, FOV=192mm, 256x256 acq. matrix, 1.0mm3 resolution) sequence and 1H-MRS acquired via PRESS (TE= 30ms, TR=5000ms, 3.0mL, 128 supp averages, 16 unsuppressed acquisitions, 8-channel head coil.) from a left frontal white matter voxel (3T MAGNETOM Trio, Siemens, Erlangen, Germany) [243]. Position and orientation were adjusted to contain white matter as much as possible. For the retest scan, the voxel was manually prescribed to the same region of interest informed by images of voxel location of the initial test scan. Absolute metabolite quantitation 64

Data Processing: Binary masks representing voxel limits were mapped onto anatomical space by calculating the transformation matrix from the RAW file header [250, 251]. Voxel overlap was calculated via the Sørensen-Dice coefficient [252, 253]. Three different Segmentation algorithms were tested including SPM12 NewSegment (SPM; http://www.fil.ion.uc l.ac.uk/), FSL-FAST (FSL; https://fsl.fmrib.ox.ac.uk/fsl/fslwiki), Freesurfer (FSU; http s://surfer.nmr.mgh.harvard.edu/). Grey-matter, white-matter and cerebrospinal fluid tissue percentages were calculated within the limits of the MRS binary mask. Probabilistic maps were binned at 0.5 to make tissue concentrations add up to 100%. Scanner-averaged frequency domain spectra were analyzed with LCModel [199], which implements a fully automated quantitation algorithm which does not account for partial volume effects. Inclusion criteria for good quality spectra were signal-to-noise ratio higher than 15, and line-widths lower than 12Hz, and Cramer Rao lower bounds lower than 20%. Compartmentation within the MRS voxel was considered for the quantitation of absolute metabolite concentrations by applying equation (2) in Gussew et al.

[254] while ignoring relaxation eﬀects of metabolites since they have similar T1/T2 times in grey-matter and white-matter and are approximately accounted for by LCModel. Ab- solute metabolite concentrations were calculated using tissue fraction percentages generated using SPM, FSL and Freesurfer. Test-retest statistical comparisons were performed using a paired sample t-test with a signiﬁcance level set at P<0.05 (uncorrected)).

6.3 Results

Re-localization Accuracy The relocalization procedure across the diﬀerent time-points yielded Sørensen-Dice coef- ﬁcients of 0.76 ±0.14 for the cingular voxel; 0.81 ±0.13 for the thalamic voxel; and 0.12 ±0.14 for the white-matter voxel.

Segmentation Consistency No significant differences were observed in tissue fraction estimates between sessions using the three algorithms (paired t-tests) (Figure 6.1). SPM and FSL estimates were similar (Table 6.1), however, Freesurfer grey-matter estimates were significantly different as they were ≈20-30% lower in the cingulum, ≈30-40% higher in the thalamus and ≈5- 10% lower in white-matter. SPM exhibited the lowest variation (coefficient of variation) and highest consistency in tissue fraction estimates across sessions. Importantly, we note the freesurfer does not segment cerebrospinal fluid in the cingulate cortex.

Absolute metabolite quantitation and correction With LCmodel ﬁtting, ﬁve metabolite concentrations were considered. These included total N-acetylaspartatyl compounds (tNAA; i.e. N-acetylaspartete, NAA, plus NAAG), Absolute metabolite quantitation 65

Table 6.1: Test-retest tissue fraction estimates of SPM, FSL and Freesurfer (FSU)

GM WM CSF SPM FSL FSU SPM FSL FSU SPM FSL FSU Mean 72.9 59.6 44.1 13.0 23.1 25.9 14.1 17.3 1.1 ACC SD 2.4 5.5 3.7 2.5 7.7 5.7 0.9 3.8 1.0 Mean 51.7 45.1 76.0 40.5 39.0 5.3 7.8 15.9 8.1 THA SD 2.7 4.6 6.4 3.0 5.7 1.8 0.6 5.9 1.9 Mean 11.8 10.1 5.1 88.1 89.1 94.5 0.1 0.8 0.1 WM SD 6.1 5.2 3.1 6.2 5.8 3.3 0.1 0.7 0.0

(phospho)creatine (Cre), choline compounds (Cho), myo-inositol (m-Ins), glutamate (Glu), glutamine (Gln), and Glu+Gln (Glx). There were no significant differences observed between sessions for the uncorrected metabolite concentrations (Figure 6.2). Mean coefficients of variance (COV) were lower than 10% for most metabolites. Partial volume correction with SPM tissue estimates exhibited the highest reliability between-sessions (lowest COV%) for different anatomical sequences and different voxels. Quantitation correction with SPM decreased between-session variance for Glx and Cho in the thalamic voxel. Increases in variance for other metabolites in different regions were slight and nonsignificant.

6.4 Discussion

In this study, we examined the test-retest reliability of 1H-MRS absolute metabolite quantitation with partial volume correction. Test-retest data was acquired from three regions of interest to inspect the consistency of tissue segmentation algorithms and the effect of utilizing these estimates for absolute metabolite quantitation with partial volume correction. An automated realignment technique was implemented to localize the cingualte and thalamic voxeles. This automated relocalization technique resulted in relatively high Sørensen-Dice coefficients see [250], thus indicating high repeatability of the localization technique utilized for the purposes of this thesis. For the frontal white- matter voxel, which was previously acquired as part of another study, manual placement of retest voxel resulted in low spatial overlap (Sørensen-Dice coefficient = 12%). Given (a) that this voxel was prescribed with the goal of maximizing white matter content (b) the voxel included approximately 90% white matter (Table 6.1)) and (c) the the white matter region was assumed to exhibit similar metabolite concentrations [255], the exact reproduction of the voxel’s position was not an issue in this case.

Our results indicated that sophisticated and commonly used segmentation algorithms yield diﬀerent regional estimates of grey matter, white matter and cerebrospinal ﬂuid. We report that SPM yielded the highest consistency between the three tissue segmentation Absolute metabolite quantitation 66

Figure 6.1: Test-retest reliability of commonly used tissue segmentation methods. The left panel depicts the localization of the voxels and the respective segmentation outputs for the (a) anterior mid-cingulate cortex, (b) bilateral-thalamus and (c) frontal white matter regions of interest. Plots of the inter-session variability of the three segmentation algorithms are illustrated in the left panel. The results clearly indicate that SPM has the lowest inter-session variability. COV=Coeﬃcient of variance. methods, which led to the lowest between session variance of partial volume corrected absolute metabolite estimates. Although consistency does not imply validity, we observed that SPM yielded the lowest between session variance across diﬀerent samples, regions-of- interest, and scanning protocols. In comparison with LCModel quantitation (i.e. without partial volume correction), we report a decrease in variance of key metabolites estimates when considering SPM tissue fraction estimates. Absolute metabolite quantitation 67

Figure 6.2: Test-retest reliability of 1H-MRS absolute metabolite quantitation with partial volume correction. The overlap of the test-retest spectra for (a) anterior mid-cingulate cortex, (b) bilateral-thalamus, and (c) frontal white matter voxels is illustrated in the left panel. The inter-session variance for absolute metabolite quantitation is illustrated in the right panel. Correction with SPM outputs led to reductions in the inter-session variance of absolute concentrations of key metabolites.

6.5 Conclusions

To interrogate changes of relevant metabolites in the longitudinal setting in psychiatric or neurological disorders, the utility of SPM to correct for 1H-MRS partial volume eﬀects is recommended. Chapter 7

Coil Combination

This chapter presents a methodological study in which the effect of different coil combination algorithms on Quantitative Susceptibility Mapping (QSM) results were investigated. This study was initially conducted to correct for and assess the influence of an artifact present in the tissue phase maps generated by the MR system (Siemens Magnetom Verio, VB17) used in the acquisition of the imaging data in this thesis. This investigation led to the conclusion that vendor provided phase maps lead to a bias in the QSM data reconstructed from the tissue phase maps, and must therefore be corrected before the statistical comparison of patient/control data for an objective interpretation of changes of subcortical magnetic susceptibility as a surrogate measure of iron content (Chapter9). The presented work is an expanded format of an abstract that underwent a peer review process and was presented at the 2017 annual conference of the International Society of Magnetic Reso- nance Imaging in Medicine in Honolulu, Hawaii, USA: Metere and Kanaan et al., Effects of coil combination algorithms on Quantitative Susceptibility Mapping, Proc. Intl. Soc. Mag. Reson. Med. 25, 2017, Program No. 2433 [256].

7.1 Introduction

Quantitative Susceptibility Mapping (QSM) data is calculated from the phase evolution of the signal generated via gradient-echo imaging sequences [257] by exploiting the local relationship between the phase and the magnetic susceptibility in the Fourier domain [258]. The MR signal is often collected through phased array coils which oﬀer increased signal-to-noise ratio (SNR) compared to volume coils [259] and allow for the use of parallel imaging techniques [260–262]. Typically, the complex images obtained separately from each channel are combined together to obtain the ﬁnal image of the sample.

Since each channel is subject to speciﬁc modulations and oﬀsets, a linear combination of the signals produces a spatial distribution of constructive and destructive interference, resulting in large portions of the image having an extremely low SNR. To overcome

68 Coil Combination 69 this issue, several coil combination algorithms have been proposed [259, 263–266] and compared [267]. However, under typical experimental conditions at 3T and 7T, the phase images reconstructed with some of these algorithms may exhibit pole artifacts or open-ended fringe lines [268]. Of note, the widely-used adaptive coil combination algorithm produces accurate and SNR-optimized images without the need for an additional reference scan [263] but is not robust against pole artifacts.

Recently, it has been shown that the eigenvector analysis introduced in ESPIRiT for the purpose of parallel imaging [262], can be used for estimating absolute-phase maps using virtual conjugate coils [269] by explicitly exploiting the conjugate symmetry in k-space. A similar approach using singular value decomposition (SVD) on the coil elements and ESPIRiT to estimate the coil sensitivity was independently proposed [270]. The ESPIRiT-SVD coil combination method, like the adaptive one, does not require additional scanning nor additional echoes, but allows for accurate tissue phase estimate without any pole artifacts.

To our knowledge, the eﬀect of pole artifacts in QSM results has not been systematically investigated yet. In this work we compare the QSM results obtained from the phase images calculated with both the adaptive (producing pole artifacts) and the ESPIRiT- SVD (without pole artifacts) algorithms in a cohort of healthy controls.

7.2 Methods

Magnetic resonance measurements were performed on a 3T MAGNETOM Verio (Siemens Healthcare, Erlangen, Germany) equipped with a 32-channel head coil following approval of the local Ethics Committees and obtaining informed consent. Magnetic resonance imaging data included sample of N=22 healthy control subjects (Age= 38.59±11.75 years, 4 female) as described in detail in Chapter 9. For inspecting diﬀerence of QSM values in deep grey matter nuclei, high resolution anatomical images were acquired for segmentation purposes using an Magnetization-Prepared 2 Rapid Gradient Echo (MP2RAGE)

[271] with the parameters TR, = 5 s; TE = 3.93 ms; inversion times=0.7/2.5s; sagittal slab orientation; acquisition matrix 256 256 176, and 1 mm3 nominal isotropic resolution. Susceptibility weighted data were acquired using a 3D-flow compensated, spoiled, gradient recalled echo sequence with the parameters: TR=30s; TE=17ms; 256×256 matrix; flip-angle=13◦; 0.8mm isotropic nominal resolution. Comprehensive imaging parameters are provided in Appendix D. For each acquisition, combined phase images from the single-channel complex data were reconstructed using (a) the adaptive coil combination algorithm [263]; and (b) the ESPIRiT-SVD coil combination algorithm [270]. The adaptive combine method was implemented directly at the scanner by the manufacturer, and Coil Combination 70 was applied with the standard settings. For the ESPIRiT-SVD method, N = 16 principal components were used. The ESPIRiT calibration to obtain the virtual sensitivities of the principal components was performed using the Berkeley Advanced Reconstruction Toolbox (BART) v0.3.01 [272]. QSM was computed using the superfast dipole inversion approach which employs (a) sophisticated harmonic artifact reduction for phase (SHARP) to eliminate background field contributions and (b) threshold k-space division (TKD) for calculating magnetic susceptibility [273]. All QSM data were referenced to median cerebrospinal fluid susceptibility, which was computed within a subject specific mask of the lateral ventricles [219]. For the masking of subcortical nuclei, an MP2RAGE- QSM hybrid contrast image was created to yield more accurate segmentations of basal ganglia nuclei via FSL-FIRST. Robust co-registration between skull-stripped MP2RAGE and FLASH data was achieved using rigid-body linear transformation of the T1-weighted data onto N4-bias field corrected FLASH magnitude data. Brainstem and cerebellar nuclei were mapped onto native QSM space by non-linear transformation of masks that were carefully delineated on a QSM group-average template using the Advanced Normaliza- tion Tools (ANTs). Cortical regions were also extracted from the anatomical data via the automated freesurfer pipeline. Median susceptibility values were extracted from each region-of-interest (ROI) and Wilcoxon rank-sum tests were used to assess differences.

7.3 Results

Beyond the arbitrary phase offsets, the tissue phase maps generated using the adaptive and the ESPIRiT-SVD coil combination methods exhibited different spatial modulations. The most striking effect was the presence of pole artifacts in the adaptive method and their absence with the ESPIRiT-SVD method throughout the volume (Figure 7.1).

The difference images between the two algorithms demonstrated the presence of the artifact more clearly on (a) single subject native images (Figure 7.2A), (b) the standardized group average image (Figure 7.2B); and (c) standard deviation of the standardized group average image (Figure 7.2C). In all three cases, the differences were strikingly visible in the region where the phase singularity is localized. Considering the 2 dimensional histograms for each image, spreading around the identity line indicated that these differences are global, i.e. they are present throughout the volume.

Considering susceptibility differences between various regions of interest including the basal ganglia, brainstem and cerebellar nuclei as well cortical regions, significant differences were observed for the red nucleus, caudate and the cingulum (and trends for significance for the subthalamic nucleus) between the two methods (Table 7.1). It is Coil Combination 71

Figure 7.1: Tissue phase estimation using adaptive coil combination and ESPIRiT-SVD from multichannel data. Note that presence of the open-ended fringe line artifact in (a) the adaptive method and their absence in (b) ESPIRiT-SVD.

Table 7.1: Statistical comparison between magnetic susceptibility values achieved with the adaptive and ESPIRiT-SVD coil combination methods..

Adaptive ESPIRiT-SVD P-Value Substantia Nigra 103±40 112±25 0.178 Red Nucles 66±48 55±42 0.005 Subthalamic Nucleus 0±44 14±34 0.072 Caudate 24±16 18±13 0.006 Putamen 9±18 8±16 0.372 Pallidum 76±21 76±17 0.910 Thalamus -33±18 32±12 0.570 Cingulum -20±15 -25.6±8.4 0.039 Basal Ganglia 23±14 20±10 0.101 Brainstem 56±35 61±29 0.833 Susceptibility values are reported in parts per billion. The signiﬁcance levels was set at P <0.05 uncorrected.

of note that the QSM data reconstructed following the ESPIRiT-SVD coil combination method consistently yielded lower standard deviations.

7.4 Discussion

Apart from the region nearby the pole artifact, the QSM results from the adaptive and the ESPIRiT-SVD coil combinations methods were remarkably diﬀerent throughout the imaging volume. The corresponding phase images also appear very diﬀerent as a Coil Combination 72

Figure 7.2: QSM differences following multi-channel coil combination with the adaptive and ESPIRiT-SVD algorithms The difference between the two algorithms is illustrated on a (a) single subject, (b) the mean image of the standardized group data, (c) standard deviation of the standardized group data. Note that the differences are stronger near the location of the singularity as illustrated in the third pannel. result of substantially different low spatial frequency components. However, our QSM pipeline involves a background-field removal step, which effectively acts as a spatial high-pass filter. While this step is expected to perform poorly for the removal of pole artifacts, it should be adequately removing low spatial frequency components, provided that they fulfill the Laplacian conditions [267], but the presence of pole artifact may significantly degrade the global performances of this QSM step. Additionally, our QSM pipeline includes a field-to-source inversion step, which enables the determination of the (relative) magnetic susceptibility. However, the relationship between the phase evolution and the magnetic susceptibility is local in the Fourier domain [258], and therefore global in the image domain. For this reason, a local artifact in the (properly filtered) phase images should produce global effects in the resulting magnetic susceptibility. Therefore, Coil Combination 73 we speculate that the globally different QSM originate from the combination of these two aspects.

Results from the group studies indicated that the pole artifacts are consistently localized in the same area across subjects. This finding indicates that, to a considerable extent, the observed pole artifacts are determined by the coil geometry and performances. The difference in performances of the adaptive and ESPIRiT-SVD methods are also localized consistently and induce an overall systemic effect in QSM, as evidenced by the non- vanishing group average for the QSM results. The spatial distribution of group standard deviation for the QSM results was substantially lower for the ESPIRiT-SVD case, and this is an indication of overall more consistent results. The ROI-based analysis indicates that the susceptibility values are (statistically) significantly different for cortical and subcortical regions that are typically of interest. This is especially relevant for group studies where consistent magnetic susceptibility results are crucial, and the presence of such artifacts may degrade the otherwise statistically significant QSM results to the point that important scientific and clinical questions may be left unanswered or poorly addressed.

Several coil combination methods have been developed over the recent years, and excellent comparisons have been already presented in great detail [267, 268]. While the so-called Römer approach [259] is considered the gold standard, it has the drawback of requiring additional acquisitions and therefore it may not be feasible to apply under speciﬁc experimental settings, most notably retrospectively to existing data-sets. Many other proposed algorithms do present the same problem. In contrast, the ESPIRiT- SVD method does not require an extra reference scan nor a multi-echo acquisition, and therefore can be applied to existing data-sets where individual coil elements images are available.

7.5 Conclusions

This study indicates that pole artifacts significantly corrupt QSM results. The effects are stronger for regions of interest that are close to the pole artifacts (e.g. striatum), but are also significant throughout the whole volume. The utility of a coil combination algorithm that does not exhibit any pole artifacts such as ESPIRiT-SVD, is essential for the reliable estimation of cortical and subcortical susceptibility for precise interrogation of inter-group differences. Part IV

Pathophysiological Investigations

74 Chapter 8

Neurochemical Investigation of pathophysiology

This chapter presents a comprehensive analysis of the role of the neurotransmitter glutamate in the pathophysiology of GTS. The investigation was conducted via 1H-MRS at baseline and during treatment with the antipsychotic aripiprazole. The accuracy of glutamatergic metabolite quantitation was signiﬁcantly improved by information gleaned from the preceding methodological investigation presented in Chapter 6. Further methodological optimizations were conducted and are presented in this chapter. This work was published as a peer-reviewed journal article: Kanaan AS, et al. Pathological glutamatergic neurotransmission in Gilles de la Tourette syndrome. Brain: A Journal of Neurology 140:218-234, 2017 and is presented with the inclusion of additional details [274].

8.1 Abstract

Gilles de la Tourette syndrome (GTS) is a hereditary, neuropsychiatric movement disorder with reported abnormalities in the neurotransmission of dopamine and γ-aminobutyric acid (GABA). Spatially focalized alterations in excitatory, inhibitory and modulatory neurochemical ratios within functional subdivisions of the basal ganglia, may lead to the expression of diverse motor and non-motor features as manifested in GTS. Current treatment strategies are often unsatisfactory thus provoking the need for further elucidation of the underlying pathophysiology. In view of (a) the close synergy exhibited by excitatory, inhibitory and modulatory neurotransmitter systems; (b) the crucial role played by glutamate (Glu) in tonic/phasic dopaminergic signalling; and (c) the interdependent metabolic relationship exhibited between Glu and GABA via glutamine (Gln); we postulated that glutamatergic signalling is related to the pathophysiology of GTS. As such, we examined the neurochemical proﬁle of cortico-striato-thalamo-cortical regions in 37

75 Glutamate 76 well-characterized, drug-free adult patients and 36 age/gender-matched healthy controls via magnetic resonance spectroscopy at 3T. To interrogate the influence of treatment on metabolite concentrations, spectral data was acquired from 15 patients undergoing a four-week treatment with aripiprazole. Test-retest reliability measurements in 23 controls indicated high repeatability of voxel localization and metabolite quantitation. We report significant reductions in striatal concentrations of Gln, Glu+Gln (Glx) and the Gln:Glu ratio and thalamic concentrations of Glx in GTS in comparison to controls. On-treatment patients exhibited no significant metabolite differences when compared to controls but significant increases in striatal Glu and Glx, and trends for increases in striatal Gln and thalamic Glx compared to baseline measurements. Multiple regression analysis revealed a significant negative correlation between (a) striatal Gln and actual tic severity and (b) thalamic Glu and pre-monitory urges. Our results indicate that patients with GTS exhibit an abnormality in the flux of metabolites in the GABA-Glu-Gln cycle, thus implying perturbations in astrocytic-neuronal coupling systems that maintain the subtle balance between excitatory and inhibitory neurotransmission within subcortical nuclei.

8.2 Introduction

The basal ganglia are organized into distinct functional sub-territories that receive/emit projections from/to specific cortical regions, forming a number of parallel and integrative Cortico-Striato-Thalamo-Cortical (CSTC) loops that provide a mechanism for the selection of an action from competing responses [72, 79, 275]. Broad categories of behaviour are thought to map onto specific functional subdivisions within motor, associative and limbic functional domains [276]. At the neurochemical level, the dopaminergic modulation of glutamatergic cortico-striatal afferents, innervating distinct populations of striatonigral and striatopallidal Medium-sized Spiny Neurons (MSNs), is essential for the regulation of thalamo-cortical output to maintain the spatial selectivity of context- adapted behaviour [121]. Spatially focalized alterations in excitatory, inhibitory and modulatory neurochemical ratios within specific functional sub-systems may perturb the typical dopamine D1- and D2-receptor mediated modulation of striatonigral and striatopallidal outputs [75]. Such perturbations may lead to the expression of diverse motor and non-motor features as exhibited by various psychiatric and movement disorders.

Gilles de la Tourette Syndrome (GTS) presents an example of such a disorder with fundamental alterations in the functional dynamics of CSTC circuitry [28]. In essence, patients with GTS are characterized by the presence of motor and vocal tics, which have been defined as rapid, habitual, burst-like movements or utterances that typically mimic Glutamate 77 fragments of normal behaviour [9]. Patients often report unpleasant premonitory urge sensations preceding tics that are relieved by their execution [14]. The large majority of patients with GTS also present with other comorbid conditions that include Atten- tion Deficit/Hyperactivity Disorder (ADHD), Obsessive-Compulsive Behaviour/Disorder (OCB/D), depression and anxiety [16]. In current models of GTS pathophysiology, symptoms are thought to arise as a result of the inappropriate activation of specific clusters of striatal neurons, which lead to a burst-like disinhibition of thalamo-cortical output [76, 79]. Although the therapeutic spectrum for GTS has recently been expanding, current treatment strategies are often unsatisfactory [92], thus provoking the need for further elucidation of the underlying pathophysiology.

The bulk of current literature on GTS pathophysiology supports the hypothesis of a dysregulated dopaminergic system. This notion was initially supported by clinical evidence of improvements in tics following the administration of dopamine antagonists, synthesis blockers or depletion drugs and the exacerbation of symptoms following the administration of dopaminergic stimulants [92]. Essentially, the implication is that patients with GTS exhibit a hyper-responsive spike-dependent dopaminergic system [28]. Methodologically varied work has revealed that patients with GTS exhibit alterations in (a) D2 receptor binding; (b) dopamine active transporter density/binding; and (c) phasic dopamine transmission in striatal and cortical regions [31]. In summary, a critical distillation of the literature suggests a dysregulation in the ﬁring patterns of dopaminergic nuclei [31].

Dopaminergic signalling is functionally compartmentalized into two systems: (a) an asynchronous, extra-synaptic, ’tonic-ﬁring’ state driven by an intrinsic pacemaker potential that is gated by inhibitory pallidal γ-aminobutyric acid (GABA) input; and (b) a synaptically focused, burst-like, ’phasic-ﬁring’ state that is dependent on a glutamatergic excitatory drive via the activation of N-methyl-D-aspartate (NMDA) receptors [121]. The transient phasic release of dopamine into striatal regions has been shown to be driven by functionally relevant stimuli encoding reward-prediction errors [277], reinforcement- seeking behaviour [278] and the selection of habitual motor programs [279]; functions that have been implicated in the pathophysiology of GTS [280–282].

The interaction between tonic and phasic dopaminergic signalling is believed to po- tently modulate cortico-striatal input and response selection in ambiguous situations to most eﬀectively guide behaviour [121]. Current models indicate that the regulation of dopaminergic signalling is a dynamic process that is sculpted by a spatio-temporal synergy between (a) local feedback mechanisms that are regulated by the action of tonically- driven, extra-synaptic dopaminergic concentrations on somatodendritic and presynpatic Glutamate 78

D2 autoreceptors [118]; and (b) glutamatergic input that is impeded and gated by pallidal GABAergic and mesencephalic cholinergic afferents, respectively [119, 120]. Thus, the irregular afferent modulation of dopaminergic nuclei would have profound effects on tonic/phasic dopaminergic release in the striatum and the control of subsequent thalamo- cortical output [283, 284].

In view of (a) the close synergy exhibited between excitatory, inhibitory and modulatory neurotransmitter systems within the striatum and throughout the brain, and (b) the interdependent metabolic relationship exhibited between glutamate (Glu), and GABA via the non-neuroactive metabolic intermediate glutamine (Gln) [285]; one can postulate that if a dopaminergic abnormality were present, other neurotransmitter systems would exhibit perturbations as well [131]. Along this line of reasoning, separate groups have demonstrated that adult patients with GTS exhibit alterations within the GABAergic system in cortical regions using in-vivo 1H Magnetic Resonance Spectroscopy (1H-MRS) [130, 286] and subcortical regions using Positron Emission Tomography (PET) [127]. Additionally, previous ex-vivo work has also indicated that striatal and pallidal nuclei exhibit alterations in the distribution of fast-spiking GABAergic [125] and tonically- active cholinergic interneurons [126]. These results lend support to the notion of extant irregularities in the modulatory mechanisms governing cortico-striatal input and output selection in GTS.

Given the crucial role of the glutamatergic excitatory drive on the phasic dopaminergic response and the extensive interactions exhibited between the two systems, we postulate that patients with GTS exhibit alterations in glutamatergic signalling [131]. This hypothesis is further buoyed by: (a) a post-mortem report indicating reductions in Glu concentrations in pallidal and nigral regions in a sample of four patients [124]; (b) genomic-based studies indicating linkage to regions overlapping with Glu transporter genes [137–139]; (c) a gene-set-based analysis implicating astrocyte metabolic coupling mechanisms [287]; (d) a transcriptome analysis indicating downregulations of glutamatergic synapse genes [288]; (e) a study indicating that the disruption of NMDA-dependent dopaminergic burst ﬁring attenuates the acquisition of conditioned behavioural responses and cue-dependent learning [289].

As the role of glutamatergic signalling has not been thoroughly investigated in adult patients with GTS in-vivo, we aimed at investigating the neurochemical profile of cortical and subcortical CSTC regions in a well-characterized group of adult patients via 1H-MRS for the first time. To this end, we employed novel analysis techniques to mitigate the effects of motion artifacts and to obtain absolute concentration estimates for a more reliable evaluation of subtle metabolite changes. We additionally utilized a longitudinal study design to investigate the influence of treatment on Glu and Gln concentrations in Glutamate 79 a subset of patients that were treated with the atypical, second-generation antipsychotic aripiprazole.

8.3 Materials and Methods

8.3.1 Population Sampling

The study was approved by the local Ethics Committees. All participants gave written informed consent and received monetary compensation for their participation. Forty- three right-handed adult patients with GTS (7 female; 18-65 years) were recruited from the outpatient psychiatry clinic at Hannover Medical School. Demographic and clinical data of all subjects included in the final analysis are summarized in Table 8.1. Patients were deemed ineligible if they exhibited severe tics to the head and face, contraindications to Magnetic Resonance (MR) examinations, a history of other significant neurological disorders and current abuse of drugs and alcohol. Patients using any psychoactive substances underwent a four-week washout period before participation. After baseline data acquisition, a subset of patients (N=17) received treatment with oral aripiprazole, which was administered using a titration procedure commencing at 2.5mg/day up to maximum daily dosage of 30mg based on treatment response. All patients were diagnosed based on DSM-5 criteria and underwent a thorough clinical assessment battery. The Yale Global Tic Severity Scale (YGTSS) [174] and the modified Rush Video-based Tic Rating Scale (RVTRS) [175] were used to capture tic severity, while the Premonitory Urge for Tics Scale (PUTS) [176] was used to assess premonitory urges. Clinical information was also collected on comorbid features using various scales that include the Yale-Brown Obses- sive Compulsive Scale (Y-BOCS) [178] and the Revised Obsessive-Compulsive Inventory (OCI-R) for OCD; the Beck Depression Inventory (BDI-II) [181] for depression; and the Conners’ Adult ADHD Rating Scale (CAARS) [180] for ADHD. Psychiatric comorbidities were diagnosed as described in Chapter 5. Forty age- and sex-matched healthy control subjects (8 female, 18-65 years) without a history of neurological, psychiatric and tic disorders were also recruited and assessed in a similar manner as the patients. A subset of the control subjects (N=23) were invited for a second MR exam for test- retest reliability measurements. All subjects were instructed not to drink coffee or tea, and to abstain from smoking for at least 2h before the examination. Subjects were also instructed to adhere to a regular sleeping cycle the night before the scan. To minimize the variability that could arise from circadian physiological effects [290], the time of day of the MR exam was matched between patients and controls, with the majority of acquisitions conducted between 10 AM and 4 PM. Glutamate 80

Table 8.1: Demographic and clinical characteristics of the 1H-HMRS study sample included in the ﬁnal analysis

Test statistic Test statistic GTS Controls GTS-APZ (Controls vs. (GTS vs. (treated) GTS) GTS-APZ) 36 37 (15) 15 — — 38.3±11.1 t = 0.045, Age (years) 38.4±11.1 40.1± 13.1 71 — (39.9±13.1) P = 0.96 31/6 Odds ratio = 0.8, Gender (M/F) 29/7 — (13/2) P = 0.76 Handedness 36 right 37 right (15 right) 15 right — — 21.2±8.2 t = 1.70, YGTSS- TTS — 19.0±6.6 — 13 (21.6±6.9) P = 0.11 44.2±17.0 t = 3.60, YGTSS- GS — 35.7±16.6 — 13 (48.7±16.5) P = 0.003 9.0±5.0 t = 2.40, RVTRS — 8.1±4.0 — 13 (10.5±6.0) P = 0.03 20.1±5.8 t = 0.53, PUTS — 19.9±5.9 — 13 (20.5±6.0) P = 0.61 t = 0.84, QOL — 22.4±16.0 — 13 P = 0.42 3.3±5.5 t = 0.70, Y-BOCS — 4.9±5.6 — 13 (6.1±7.0) P = 0.50 14.7±12.0 t = 4.5, t = 0.054, OCI 4.8±5.2 16.7±13.5 71 13 (16.2±11.5) P < 0.0001 P = 0.96 52.0±14.1 t = 4.2, t = _0.25, CAARS 40.5±7.8 54.2±13.0 71 13 (54.2±14.8) P < 0.0001 P = 0.82 13.5±11.8 t = 5.5, t = 1.75, BDI-II 2.8±4.2 10.7±8.2 71 13 (14.2±10.8) P < 0.0001 P = 0.10 Medication history Drug-naive — 16 (6) — — — Nonpsychoactive — 12(4) — — — Stimulants — 2(2) — — — Antipsychotics — 2(2) — — — Benzodiazepines — 1(0) — — — triptane=1 (0), Other — — — — opipramol=2 (0) Unreported 1 (1) — — — Diagnosis GTS-only — 18 (6) 10 — — GTS+OCD — 7 (5) 2 — — GTS+ADHD — 8 (1) 1 — — GTS+OCD+ADHD — 4 (3) 2 — — All the recruited subjects were right handed. Medication history is reported for drugs taken at least four weeks before data acquisition. Abbreviations: ADHD = Attention Deﬁcit/Hyperactivity Disorder; BDI-II = Beck Depression Inventory; CAARS = Con- ners’ Adult ADHD Rating Scales; GTS = Gilles de la Tourette Syndrome; OCD = Obsessive-Compulsive Disorder; OCI-R = Obsessive-Compulsive InventoryâA˘ ¸SRevised; PUTS = Premonitory Urge for Tics Scale; QOL= Quality Of Life scale; RVTRS = modiﬁed Rush Video-based Tic Rating Scale; Y-BOCS = Yale-Brown Obsessive-Compulsive Scale; YGTSS-GS = Yale Global Tic Severity Scale Global Score; YGTSS-TTS = YGTSS Total-Tic Score. Glutamate 81

8.3.2 Magnetic Resonance Data Acquisition

Magnetic resonance measurements were performed on a 3T MAGNETOM Verio (Siemens Healthcare, Erlangen, Germany) equipped with a 32-channel head coil. The patients were instructed to remain still without actively suppressing their tics and thinking of nothing in particular. Overall, the scanning session lasted approximately 75 minutes and included the acquisition of self-report data on the degree of tic-urges and tic-suppression (see section 8.4.3). A landmark-based pre-scan gradient-echo sequence provided by the vendor (Auto-Align Head, AAH) was applied at the beginning of each scan for automatic detection of the crista galli and the tip of the occipital bone within the mid-sagittal plane [249]. The geometric information was subsequently applied on all imaging protocols and saved for the retest scans. T1-weighted images were acquired in the ﬁrst AAH space using a three-dimensional Magnetization-Prepared 2 Rapid Gradient Echo

(MP2RAGE) sequence [271]: repetition time, TR=5s; echo time TE=3.93ms; inversion times, T1=0.7/2.5s; sagittal slab orientation; acquisition matrix 256 × 256 × 176, and 1 × 1 × 1mm3 nominal resolution [291]. To minimize errors that arise from bulk head- displacement between the anatomical and spectral acquisitions, single-shot ‘dummy’ spectra were localized on the MP2RAGE image immediately after acquisition. The (fast) AAH sequence was applied again before each 1H-MRS voxel acquisition to co-register the ‘dummy’ scan geometry to the newly deﬁned space. For the retest scans, the same procedure was used to automatically re-localize the 1H-MRS voxel to the same region. Motivated by previous imaging, genetic and post-mortem work [30, 31, 124, 288], 1H spectra were acquired from three cuboid Regions of Interest (ROIs), with Point-RESolved

Spectroscopy (PRESS) [292] and TE=30ms; TR=3s; 1024 time-domain data points; 80 (ROIs 1 and 2; see below) or 128 (ROI 3) water-suppressed and 16 water-unsuppressed averages. A ﬁrst 25 × 16 × 16mm3 voxel was prescribed on the anterior Mid-Cingulate Cortex (aMCC) with the center on the level of the genu of the corpus callosum and an orientation parallel to the hippocampal axis (Figure 8.1A). A second 28 × 16 × 16mm3 voxel was centered on the bilateral thalamus while maximizing the amount of grey matter (GM) (Figure 8.1B). A third 20 × 15 × 20mm3 voxel was localized within the left corpus striatum as anterior as possible to include maximal portions of the caudate; as inferior as possible to include the most inferior portion of the putamen; and as medially as possible without including any portions of the lateral ventricles (Figure 8.1C). On average, the voxel contained GM portions from the putamen (63.2%), pallidum (22.3%), caudate (12.8%) and nucleus accumbens (1.7%). ROIs were shimmed automatically using FASTESTMAP with 4-5 iterations [293, 294]. Comprehensive imaging parameters are provided in Appendix D. Glutamate 82

8.3.3 Absolute Metabolite Quantitation

As MRS measurements involve the summation of multiple averages to build the Signal- to-Noise Ratio (SNR), subject motion and drifts in the magnetic field during acquisition cause frequency and phase errors [295]. Such drifts give rise to incoherent signal averaging, line broadening, lineshape distortion and reduced SNR, which may ultimately affect metabolite quantitation and group comparisons. As such, non-averaged time-domain raw data were exported from the scanner and a non-linear least-squares minimization operation was employed to fit each signal average, (where t is time) to a reference scan,

Sref(t) (here taken to be the ﬁrst average), by the adjustment of the frequency, ω, and phase, φ, of the signal according to [295, 296]:

minimize kR(t) − G(t, f, ϕ)k (8.1) f,ϕ∈R where 2π(ft+ φ ) G(t, f, ϕ) = S(t) · e 360 (8.2)

To enable fitting of complex data while avoiding non-physical parameter estimates, the vectors and Sref(t) were modified prior to minimization by concatenating their real and imaginary parts into a single real-valued vector. Motion corrupted outlier signals were additionally removed prior to spectral averaging. This was accomplished by calculating the root-mean-square of the difference spectrum between each individual acquisition and the average and discarding acquisitions deviating by more than three standard deviations (SD) from the mean [296]. The performance of the algorithm is demonstrated in Figure 8.1 (D-E).

The water signal from the non-suppressed spectra was used as a concentration reference. However, spectroscopic voxels in different subjects contain varying proportions of GM, white matter (WM) and cerebro-spinal fluid (CSF); the GM and WM compartments have different water concentrations with different T1 and T2 relaxation times for water; and negligible signals of most metabolites arise from the CSF compartment. To consider tissue compartmentation, the MRS voxel was first registered to anatomical space by calculating the transformation matrix from the raw file header, and a binary mask representing the voxel limits was then constructed to map the voxel onto the anatomical image [250, 251]. SPM12 New-Segment (http://www.fil.ion.ucl.ac.uk/spm) was used to automatically segment the brain into probabilistic GM, WM and CSF maps, which were binned at 0.5 to make the three tissue classes add up to 100%. The GM masks were optimized to include subcortical nuclei generated via FSL-FIRST’s Bayesian model-based Glutamate 83

Figure 8.1: Voxel localization and spectral data pre-processing. Console screenshots illustrating the prescription of the voxels in the (a) anterior mid-cingulate cortex (aMCC), (b) bi-lateral thalamus and (c) corpus striatum on MP2RAGE images in single subject. (d) The effect of frequency- and phase-drift (FPD) correction on data acquired from a striatal voxel of a GTS patient (top-panel) is clearly visible in the corrected data (bottom-panel). The performance of the non-linear least-squares minimization operation is illustrated on a single scan (the 25th average) superimposed on the reference signal (inset plots). For the aMCC spectra, which exhibited a high SNR, spectral registration was conducted on an approximate range between 1.8-4.2 ppm (red shaded area). For spectra with a lower SNR, the strength of the water signal was utilized for spectral registration on an approximate range between 4.2-7.5ppm (green shaded area). (e) The benefit of the correction on the linewidth and SNR is clearly visible on the red coloured spectrum. As some of the spectra exhibited a spurious residual signal in the ppm range between 3.6-4.0 ppm (blue shaded area), possibly due to insufficient spoiling, the LCModel fitting range was adjusted to 0.2-3.67ppm (orange shaded area). Glutamate 84 segmentation of subcortical nuclei [297]. Within-voxel tissue proportions were then calculated based on the optimized tissue segmentation masks. Averaged frequency- and phase-drift-corrected spectra were fit with LCModel [199]. As some spectra contained artifacts in the frequency range above 3.7 ppm, presumably due to insufficient spoiling, spectral fitting was performed in a 0.2-3.67 ppm range. Results for the default 0.2-4.0 ppm range exhibited correspondence and are not presented here. For final inclusion into the statistical models, a step-wise semi-automated quality-control method was utilized. Detectable metabolites were first identified for each voxel in the control sample if the relative Cramér-Rao Lower Bounds (CRLBs) of a given metabolite were below 100% in at least 50% of the subjects [298]. Cut-off values of absolute CRLBs (CRLBlim), defined as 50% of the mean metabolite concentration in the control sample) were then calculated for each metabolite to threshold cases with excessive fitting errors [299]. Final inclusion criteria were: (a) correct voxel prescription, (b) SNR > 10 (LCModel output), (c) full width at half maximum (FWHM) < 11Hz (LCModel output) for good quality spectra and (d) absolute CRLB < CRLBlim for individual metabolites. All surviving spectra were visually inspected to ensure the quality of included data. Within-voxel compartmentation was considered by applying Eq. 2 [254]:

Im 2 0 Cm = · · cw · ξ (8.3) Iw Nm with

f · RGM · α + f · RWM · α + f · RCSF · α ξ = GM w GM WM w WM CSF w CSF (8.4) fGM · RGM + fWM · RWM and

TR TE − T − T Rε = (1 − e 1,ε ) · e 2,ε (8.5)

w where cm is the metabolite concentration in tissue; co =55.6mol/L is concentration of bulk water [245]; I m and I w are the amplitudes of the metabolite and the water signal, respectively; and N m is the number of protons within the molecule contributing to the metabolite signal. Partial-voluming and the presence of non-water substances are accounted for by the scaling factor ξ, where f ε and αε are, respectively, the volume

fraction and relative water content of tissue ε = GM, WM, CSF), and Rε considers water

relaxation. Relaxation eﬀects of metabolites were ignored since they have similar T 1 and

T 1 in GM and WM [246]. Relaxation time and relative water tissue content values are outlined in Table 8.2. Glutamate 85

Table 8.2: Water T1 and T2 relaxation times and relative water content (α) in GM, WM and CSF

GM WM CSF

1 1 2 T1 1820 1084 4163 1 1 3 T2 99 69 503 α 0.814,5 0.714,5 14,5 1 (Stanisz et al., 2005) [197]; 2(Lin et al., 2001) [300]; 3(Piechnik et al., 2009) [196]; 4(Choi et al., 2006) [301]; 5(Norton et al., 1966) [302]

8.3.4 Statistical Analysis

Statistical analysis was performed in the Python programming language (Scipy v.0.15.1 and Statsmodels v.0.6.1) [303, 304]. All data exhibited a Gaussian distribution as assessed via the Kolmogorov-Smirnov test. Between- and within- group differences for all metabolites were assessed using two-way independent sample t-tests and paired-sample t-tests, respectively, with a significance level set at <0.05 (uncorrected). A multiple- linear-regression model accounting for age and gender was employed to examine the relationship between Gln/Glu concentrations and clinical measures (YGTSS total-tic score, RVTRS, PUTS, Y-BOCS and CAARS). Metabolite test-retest reliability measures were assessed in the control sample by calculating the coefficient of variation (COV), percent- age difference and paired-sample t-tests. The reliability of spatial re-localization was assessed using the Sørensen–Dice metric [252, 253]. Glutamate 86

8.4 Results

8.4.1 Demographic and clinical characteristics

Complete datasets were acquired from 37 of 43 recruited patients and 36 of 40 recruited controls. MR data was not collected from (a) three patients and three controls due to severe head motion during acquisition; (b) two patients and one control due to claustrophobia; and (c) one patient due to previously unreported MR contraindications. Seventeen of the 37 patients underwent a four-week aripiprazole therapy, and complete datasets were collected from 15 patients. Control and patient subjects were comparable in terms

of age (t71=0.045, P=0.96), gender (Fischer’s odds ratio 0.80, P=0.76) and handedness (Table 8.1). Control subjects significantly differed from the patients on obsessions/compulsions (OCI-R, Y-BOCS), ADHD (CAARS), and depression (BDI-II). In the subset of patients that received aripiprazole treatment, significant reductions were observed

in RVTRS (t13= 2.40, P=0.03) and YGTSS global impairment (t13=3.60, P=0.003). A comprehensive analysis of the clinical measures of the entire sample is presented in Chapter 10.

8.4.2 Test-Retest Reliability

The Sørensen–Dice metric quantifying the spatial overlap of test-retest localization yielded means and SDs of 0.80±0.10 for the aMCC, 0.83±0.10 for the thalamus, and 0.80±0.09 for the striatum, indicating high repeatability of our localization technique (Figure 8.2). Assessing the reliability of the spectral measurements (Table 8.3), we found no significant differences in (a) spectral quality parameters (FWHM, SNR) and (b) intra- voxel tissue proportions (WM, GM, CSF). Detectable metabolites (as defined above) included total N-acetylaspartatyl compounds (tNAA; i.e. N-acetylaspartete, NAA, plus N-acetylaspartylglutamate, NAAG), (phospho)creatine (Cre), choline compounds (Cho), Glu, Gln, glutamate plus glutamine (Glx), myo-inositol (m-Ins) and GABA in all voxels, in addition to Aspartate (Asp), Lactate (Lac) and Taurine (Tau) in the aMCC. Con- centration estimates were considered unreliable if they exhibited a significant test-retest group difference in the control cohort. Metabolites that failed this criterion included Asp and Lac in the aMCC and m-Ins in the striatum, and were excluded from further analyses. Glutamate 87

Table 8.3: Test-Retest Reliability of absolute metabolite quantitation. Sum- mary of spectral quality parameters, absolute metabolite concentrations and statistical results of test-retest healthy control 1H-MRS measurements

N Baseline Retest COV % Difference D 95 % CI Statistic P-Value

FWHM 20 3.5±0.7 3.7±0.6 8.8±8.1 17.7±16.2 — -0.54±0.29 t18 = -0.61 0.55 SNR 20 35.5±4.1 35.1±4.7 5.1±4.9 10.1±9.7 — -2.51±3.31 t18 = 0.37 0.72 GM 20 74.1±5.2 73.2±5.3 1.9±2.6 3.8±5.1 — -2.64±4.29 t18 = 0.78 0.45 WM 20 11.7±4.9 11.9±3.3 10.6±12.9 21.1±25.8 — -2.94±2.58 t18 = -0.14 0.89 CSF 20 14.2±5.9 14.8±5.5 6.9±9.0 13.8±18.0 — -4.4±3.1 t18 = -1.22 0.24 Asp 20 1.1±0.3 0.9±0.3 16.8±13.8 33.6±27.7 0.7 0.02±0.38 t18 = 2.36 0.03 Cre 20 7.0±0.5 6.8±0.3 2.6±2.3 5.2±4.6 — -0.08±0.43 t18 = 1.67 0.11 Cho 20 1.9±0.2 1.8±0.2 3.4±2.3 6.9±4.6 — -0.11±0.16 t18 = 0.74 0.47 mIno 20 5.5±0.5 5.5±0.5 3.4±2.9 6.9±5.8 — -0.27±0.39 t18 = 0.51 0.62 aMCC tNAA 20 8.7±0.4 8.6±0.5 2.7±2.1 5.3±4.2 — -0.27±0.38 t18 = 0.42 0.68 Glu 20 8.3±0.6 8.2±0.6 3.1±2.4 6.2±4.7 — -0.28±0.49 t18 = 0.72 0.48 Gln 20 2.0±0.4 2.0±0.4 8.2±6.0 16.4±12.1 — -0.25±0.26 t18 = 0.07 0.94 Glx 20 10.3±0.8 10.2±0.7 2.9±2.8 5.9±5.6 — -0.4±0.62 t18 = 0.59 0.56 GABA 20 1.4±0.4 1.3±0.2 13.9±9.6 27.9±19.2 — -0.18±0.21 t18 = 0.12 0.91 Tau 20 1.4±0.3 1.3±0.3 9.9±8.3 19.9±16.5 — -0.11±0.3 t18 = 1.09 0.29 Lac 20 0.6±0.2 0.5±0.2 11.7±8.2 23.3±16.3 0.45 -0.04±0.2 t18 = 2.6 0.02*

FWHM 21 3.5±0.7 3.6±0.6 8.4±8.1 16.9±16.2 — -0.51±0.28 t19 = -0.61 0.55 SNR 21 35.6±4.1 35.1±4.6 4.9±4.8 9.8±9.6 — -2.34±3.19 t19 = 0.42 0.68 GM 21 74.1±5.1 73.3±5.2 1.8±2.5 3.6±5.0 — -2.53±4.06 t19 = 0.76 0.46 WM 21 11.7±4.8 11.7±3.4 10.8±12.6 21.5±25.3 — -2.7±2.62 t19 = -0.04 0.97 CSF 21 14.2±5.7 14.9±5.4 7.0±8.8 13.9±17.6 — -4.31±2.83 t19 = -1.43 0.17

us Cre 21 7.0±0.5 6.8±0.3 2.6±2.2 5.2±4.5 — -0.1±0.4 t19 = 1.49 0.15 Cho 21 1.8±0.2 1.8±0.2 3.5±2.3 7.1±4.5 — -0.11±0.14 t19 = 0.47 0.64 mIno 21 5.5±0.5 5.5±0.5 3.4±2.8 6.8±5.7 — -0.28±0.35 t19 = 0.37 0.72

Thalam tNAA 21 8.7±0.4 8.6±0.5 2.6±2.0 5.3±4.1 — -0.28±0.34 t19 = 0.24 0.81 Glu 21 8.3±0.6 8.2±0.6 3.0±2.3 6.1±4.6 — -0.28±0.45 t19 = 0.62 0.54 Gln 21 2.0±0.4 2.0±0.4 8.3±5.9 16.5±11.8 — -0.22±0.28 t19 = 0.31 0.76 Glx 21 10.3±0.8 10.2±0.7 2.8±2.8 5.7±5.5 — -0.37±0.6 t19 = 0.64 0.53 GABA 21 1.3±0.3 1.3±0.2 13.8±9.4 27.7±18.7 — -0.19±0.18 t19 = -0.03 0.98

FWHM 18 8.0±1.2 8.2±1.3 6.8±3.9 13.6±7.9 — -1.04±0.67 t16 = -0.61 0.55 SNR 18 17.2±2.7 16.3±2.7 5.7±5.9 11.3±11.8 — -1.01±2.79 t16 = 1.39 0.18 GM 18 76.2±4.1 76.6±3.8 1.6±1.8 3.2±3.5 — -3.06±2.39 t16 = -0.38 0.71 WM 18 23.1±5.2 22.7±5.2 6.6±7.0 13.2±14.0 — -3.21±4.05 t16 = 0.44 0.67 CSF 18 0.7±2.2 0.7±2.5 — — — -1.73±1.57 t16 = -0.6 0.55 Cre 18 6.8±0.9 6.8±0.6 4.7±3.8 9.4±7.5 — -0.5±0.54 t16 = 0.08 0.94 Cho 18 1.6±0.2 1.6±0.2 4.9±4.9 9.8±9.8 — -0.21±0.1 t16 = -0.94 0.36 mIno 18 3.6±0.6 3.3±0.5 8.5±6.4 17.1±12.9 0.66 -0.01±0.77 t16 = 2.52 0.02* Striatum tNAA 18 7.4±1.1 7.1±0.7 5.7±5.1 11.4±10.1 — -0.39±0.96 t16 = 1.0 0.33 Glu 18 6.5±0.8 6.4±0.8 5.0±4.5 10.0±9.1 — -0.47±0.67 t16 = 0.46 0.65 Gln 18 2.9±1.0 2.6±0.8 18.3±13.6 36.6±27.3 — -0.26±0.95 t16 = 1.26 0.22 Glx 18 9.5±1.3 9.0±1.3 6.5±6.0 13.0±12.0 — -0.46±1.34 t16 = 1.22 0.24 GABA 18 1.9±0.3 1.9±0.4 11.7±9.6 23.4±19.2 — -0.22±0.29 t16 = 0.25 0.81 Metabolite concentrations are reported in mmol/L units. Abbreviations: aMCC = anterior mid-cingulate cortex; Asp = aspartate; CI =confidence interval; Cho = choline compounds; COV = coefficient of variation; Cre = (phospoho)creatine; CSF = cerebrospinal fluid (%); D = Cohen’s D; FWHM = full width at half maximum (Hz); GABA = γ-aminobutyric acid; GM = grey matter (%);Gln = glutamine; Glu = glutamate; Glx = glutamate plus glutamine; Lac = lactate; m-Ins = myo-inositol; NAA = N-acetylaspartate with an additional contribution from N-acetylaspartylglutamate; SNR = signal-to-noise ratio; Tau = taurine; WM = white matter (%). Glutamate 88

Figure 8.2: Spatial overlap of test-retest voxel localization Representative images illustrating the extent of spatial overlap (purple) achieved between visit 1 (red) and visit 2 (blue) using the auto-align re-localization technique for the (a) aMCC, (b) bi-lateral thalamic and (c) the left corpus striatal single voxel spectroscopy regions of interest.

8.4.3 Degree of tic-urges and tic-suppression during MR data acquisition

All patients were instructed to remain still without actively suppressing their tics and thinking of “nothing” in particular during MR data acquisition. Following the scans, the urge-to-tic and the active suppression of tics were verified via self-report. Specifically, the patients were asked: (a) “How strong was your urge to tic during scanning?” (0=no urge at all, 10=unbearable urge to tic) and (b) “How much effort did you expend to suppress tics?” (0=no effort was made to suppress tics; 10=maximal effort/attention was made to suppress tics). Median baseline values of the urge-to-tic and the effort expended to suppress tics were 3.5 and 3.0, respectively. In the on-treatment condition, similar Glutamate 89 median values were observed (urges=3.0 and suppression=2.5), indicating that intra- group comparisons of patients’ spectral measures negligible bias from active suppression of tics. Given the low observed median values of tic suppression at baseline, it seems further plausible to suggest that the mild suppression of tics may not have had a strong influence on inter-group comparisons.

8.4.4 Group Diﬀerences in Metabolite Concentrations

Group comparison of left striatal baseline measurements revealed signiﬁcant decreases in

Gln concentrations (t60=2.54, P=0.0119), Glx concentrations (t60=2.54, P=0.017) and the Gln:Glu ratio (t60=2.22, P=0.03) in GTS compared to controls (Table 8.4). The thalamus additionally exhibited decreases of Glx concentrations in GTS compared to controls at baseline (t61=2.44, P=0.018). Cohen’s effect sizes were relatively high (striatum: DGln=0.64; DGlx=0.63; DGln:Glu=0.56; thalamus: DGlx=0.61) indicating practical significance. In the subset of patients that underwent aripiprazole treatment, paired- group comparisons revealed significant increases in striatal Glu (t8= -2.30, P=0.047) and Glx concentrations (t8= -3.0, P=0.015) in on-treatment patients compared to baseline GTS (Table 8.5). In addition, we observed trends for increases in striatal Gln (t8=

-1.843, P=0.098) in thalamic Glx (t8= -2.133, P =0.064). Comparing baseline measurements of the control sample with GTS patients on-treatment, we did not observe any significant differences in Gln, Glu, Glx concentrations and the Gln:Glu ratio. Consider- ing all other metabolites, we only observed a difference in thalamic Cre concentrations

(t61= 2.39, P=0.02) when comparing controls to patients at baseline. For the aMCC, no diﬀerences were observed for any metabolites. Sample distributions and statistical results for Gln, Glu and Glx are illustrated in Figure 8.3. Representative spectra from ten subjects per voxel are illustrated in Figures 8.4–8.6. Glutamate 90

Figure 8.3: Spectral localization, fitting and statistical analysis.Left-panel: Sagittal, coronal and axial images illustrating the localization of the cingular (aMCC), thalamic (THA) and corpus striatal (STR) regions of interest. The reconstructed masks were generated based on geometric information extracted from the raw file header. Mid panel: Exemplary spectra illustrating LCModel fits, baselines and residual signals of frequency- and phase-drift-corrected data. The inset images demonstrate the location of voxels with respect to the GM, WM and CSF compartments, which were used to calculate within-voxel tissue proportions for absolute quantitation. A combination of segmentation outputs from SPM12 and FSL-FIRST was used for accurate masking of subcortical nuclei. Right panel: Plots illustrating the distribution of Gln, Glu and Glx concentrations in controls (green), GTS patients at baseline (red) and GTS patients following treatment with aripiprazole (blue); ** denotes significance at p<0.05; * denotes a trend for significance (p<0.1). Glutamate 91

Table 8.4: Control vs. GTS group comparison of absolute metabolite concentrations. Summary of spectral quality parameters, absolute metabolite concentrations and statistical results (GTS patients vs. controls) of 1H-MRS measurements.

N (HC/GTS) Controls Patients % Difference D CI(95 %) Statistic P-Value

FWHM 34/35 3.6±0.66 3.69±0.72 17.99±15.71 — -0.42±0.25 t67 = -0.49 0.622 SNR 34/35 35.21±3.72 33.74±5.29 14.54±15.85 — -0.77±3.7 t67 = 1.31 0.196 GM 34/35 73.38±5.46 71.6±7.95 8.84±7.98 — -1.56±5.11 t67 = 1.06 0.291 WM 34/35 12.73±5.07 14.77±6.68 48.6±27.35 — -4.94±0.86 t67 = -1.41 0.164 CSF 34/35 13.86±5.68 13.6±5.86 43.25±24.57 — -2.55±3.08 t67 = 0.19 0.853 Asp 34/34 1.09±0.3 1.1±0.4 31.46±33.16 — -0.19±0.16 t66 = -0.15 0.878 Cre 34/35 6.92±0.48 6.92±0.5 7.51±4.57 — -0.24±0.24 t67 = 0.0 1 Cho 34/35 1.86±0.18 1.93±0.21 10.99±8.92 — -0.17±0.02 t67 = -1.56 0.124 mIno 34/35 5.45±0.52 5.7±0.59 10.05±7.9 — -0.53±0.02 t67 = -1.86 0.067 NAA 33/35 7.95±0.44 7.9±0.5 5.82±3.67 — -0.19±0.28 t66 = 0.39 0.701

aMCC NAAG 33/35 0.67±0.23 0.68±0.23 44.27±49.67 — -0.13±0.1 t66 = -0.28 0.78 tNAA 34/35 8.62±0.41 8.59±0.49 5.14±3.55 — -0.19±0.25 t67 = 0.26 0.798 Glu 34/34 8.26±0.6 8.32±0.62 8.44±5.33 — -0.36±0.24 t66 = -0.4 0.692 Gln 34/34 1.99±0.46 1.96±0.45 26.81±24.1 — -0.19±0.26 t66 = 0.28 0.782 Glx 34/35 10.25±0.89 10.35±0.93 10.59±8.41 — -0.54±0.34 t67 = -0.45 0.653 GABA 34/35 1.33±0.31 1.33±0.27 21.05±17.66 — -0.14±0.14 t67 = -0.05 0.96 Tau 34/35 1.36±0.34 1.39±0.58 31.56±27.0 — -0.27±0.2 t67 = -0.28 0.779 Lac 33/34 0.59±0.2 0.56±0.19 30.46±32.4 — -0.07±0.12 t65 = 0.55 0.587 Gln:Glu 34/34 0.24±0.05 0.24±0.05 22.97±25.4 — -0.02±0.03 t66 = 0.36 0.717

FWHM 31/32 5.54±0.86 5.43±0.7 16.41±10.85 — -0.29±0.51 t61 = 0.54 0.592 SNR 31/32 17.68±3.47 18.0±3.21 17.33±12.2 — -2.03±1.39 t61 = -0.38 0.707 GM 31/32 83.81±11.3 82.91±8.82 13.22±15.72 — -4.28±6.08 t61 = 0.35 0.73 WM 31/32 7.31±10.16 10.02±8.71 104.02±59.52 — -7.55±2.13 t61 = -1.12 0.267 CSF 31/32 8.87±3.99 7.07±3.07 45.35±38.63 — -0.01±3.62 t61 = 1.99 0.051 Cre 31/32 6.5±0.48 6.13±0.69 11.18±11.4 0.6 0.06±0.67 t61 = 2.39 0.02* us Cho 31/32 1.84±0.18 1.81±0.28 15.91±13.06 — -0.08±0.16 t61 = 0.62 0.535 mIno 31/32 4.61±0.58 4.4±0.67 12.57±12.87 — -0.1±0.54 t61 = 1.35 0.182 tNAA 31/32 8.24±0.66 7.89±0.94 10.81±12.46 — -0.07±0.76 t61 = 1.68 0.099 Thalam Glu 31/32 6.35±0.85 6.01±0.71 14.59±11.45 — -0.05±0.75 t61 = 1.73 0.089 Gln 31/32 3.0±0.66 2.72±0.78 27.49±22.5 — -0.09±0.65 t61 = 1.51 0.135 Glx 31/32 9.35±1.04 8.72±0.96 12.37±12.87 0.61 0.11±1.14 t61 = 2.44 0.018* GABA 31/32 1.82±0.33 1.76±0.38 26.79±22.99 — -0.12±0.25 t61 = 0.72 0.473 Gln:Glu 31/32 0.48±0.13 0.46±0.17 28.21±23.68 — -0.06±0.09 t61 = 0.47 0.64

FWHM 30/32 8.05±1.45 7.72±1.57 20.57±13.88 — -0.45±1.11 t60 = 0.84 0.403 SNR 30/32 17.27±2.8 17.44±3.38 16.91±14.02 — -1.78±1.44 t60 = -0.21 0.833 GM 30/32 75.93±4.13 73.19±9.45 6.86±7.87 — -1.07±6.55 t60 = 1.44 0.155 WM 30/32 23.57±4.77 26.33±8.43 21.75±22.46 — -6.32±0.82 t60 = -1.54 0.128 CSF 30/32 0.49±1.73 0.48±1.99 — — -0.96±0.97 t60 = 0.01 0.991 Cre 30/32 6.77±0.77 6.61±0.73 12.23±10.73 — -0.23±0.54 t60 = 0.8 0.426 Cho 30/32 1.56±0.19 1.63±0.3 16.39±12.56 — -0.19±0.06 t60 = -1.0 0.319 mIno 30/32 3.48±0.6 3.61±0.6 18.85±17.07 — -0.44±0.18 t60 = -0.85 0.397

Striatum tNAA 30/32 7.36±0.9 7.09±0.7 8.9±6.88 — -0.15±0.68 t60 = 1.28 0.204 Glu 30/32 6.26±0.6 6.09±0.68 8.84±8.47 — -0.16±0.51 t60 = 1.06 0.294 Gln 30/32 2.75±0.97 2.16±0.84 53.96±39.6 0.64 0.13±1.06 t60 = 2.54 0.014* Glx 30/32 9.01±1.23 8.24±1.19 18.93±12.67 0.63 0.14±1.39 t60 = 2.46 0.017* GABA 30/32 1.91±0.28 1.86±0.41 19.74±17.78 — -0.14±0.23 t60 = 0.51 0.608 Gln:Glu 30/32 0.44±0.15 0.36±0.14 50.13±38.83 0.56 0.01±0.16 t60 = 2.22 0.03* Metabolite concentrations are reported in mmol/L units. Abbreviations: aMCC = anterior mid-cingulate cortex; Asp = aspartate; CI =confidence interval; Cho = choline compounds; COV = coefficient of variation; Cre = (phospoho)creatine; CSF = cerebrospinal fluid (%); D = Cohen’s D; FWHM = full width at half maximum (Hz); GABA = Î¸s-aminobutyric acid; GM = grey matter (%);Gln = glutamine; Glu = glutamate; Glx = glutamate plus glutamine; Lac = lactate; m-Ins = myo- inositol; NAA = N-acetylaspartate with an additional contribution from N-acetylaspartylglutamate; SNR = signal-to-noise ratio; Tau = taurine; WM = white matter (%). Glutamate 92

Table 8.5: GTS Oﬀ- and On-treatment group comparison of absolute metabolite concentrations. Summary of spectral quality parameters, absolute metabolite concentrations and statistical results of 1H-MRS measurements in GTS patients at baseline and during treatment

N GTS GTS-APZ % Difference D CI(95 %) Statistic P-Value

FWHM 12 3.79 ± 0.6 3.57 ± 0.86 11.28 ± 8.37 — -0.44 ± 0.87 t10 = 1.619 0.134 SNR 12 34.5 ± 3.3 34.42 ± 6.21 11.47 ± 11.72 — -4.32 ± 4.48 t10 = 0.061 0.953 GM 13 70.65 ± 8.63 69.66 ± 9.78 8.24 ± 7.71 — -6.78 ± 8.77 t11 = 0.524 0.61 WM 13 14.85 ± 7.01 16.18 ± 7.64 29.72 ± 31.16 — -7.5 ± 4.86 t11 = -0.54 0.599 CSF 13 14.47 ± 6.19 14.15 ± 4.05 18.83 ± 12.43 — -4.09 ± 4.73 t11 = 0.302 0.768 Asp 12 1.19 ± 0.52 0.9 ± 0.27 29.18 ± 31.52 — -0.08 ± 0.65 t10 = 1.701 0.117 Cre 12 7.02 ± 0.6 6.83 ± 0.31 5.6 ± 4.8 — -0.24 ± 0.61 t10 = 1.236 0.242 Cho 12 1.95 ± 0.24 1.87 ± 0.2 7.81 ± 6.73 — -0.11 ± 0.28 t10 = 1.372 0.197 mIno 12 5.88 ± 0.91 5.73 ± 0.45 10.81 ± 11.35 — -0.48 ± 0.78 t10 = 0.521 0.613 NAA 12 8.03 ± 0.63 7.77 ± 0.54 5.69 ± 5.29 — -0.26 ± 0.77 t10 = 1.547 0.15

aMCC NAAG 12 0.69 ± 0.22 0.68 ± 0.26 46.11 ± 55.72 — -0.2 ± 0.22 t10 = 0.085 0.934 tNAA 12 8.72 ± 0.67 8.46 ± 0.74 6.25 ± 7.98 — -0.36 ± 0.89 t10 = 1.121 0.286 Glu 12 8.4 ± 0.78 8.17 ± 0.53 7.44 ± 5.59 — -0.36 ± 0.82 t10 = 0.971 0.353 Gln 12 2.02 ± 0.46 2.04 ± 0.67 26.33 ± 16.57 — -0.53 ± 0.49 t10 = -0.107 0.917 Glx 12 10.42 ± 1.03 10.21 ± 0.98 8.05 ± 7.58 — -0.68 ± 1.09 t10 = 0.571 0.579 GABA 12 1.28 ± 0.22 1.43 ± 0.25 22.61 ± 19.18 — -0.36 ± 0.06 t10 = -1.349 0.205 Tau 12 1.43 ± 0.89 1.55 ± 0.65 44.11 ± 39.72 — -0.81 ± 0.56 t10 = -0.361 0.725 Lac 11 0.58 ± 0.18 0.52 ± 0.18 46.05 ± 25.4 — -0.12 ± 0.23 t9 = 0.555 0.591 Gln:Glu 12 0.24 ± 0.05 0.25 ± 0.08 25.64 ± 14.43 — -0.07 ± 0.05 t10 = -0.389 0.705

FWHM 10 5.24 ± 0.58 5.52 ± 1.17 10.93 ± 14.19 — -1.2 ± 0.63 t8 = -0.746 0.475 GM 11 84.92 ± 7.39 79.08 ± 16.19 13.17 ± 24.48 — -5.9 ± 17.57 t9 = 1.259 0.237 WM 11 9.17 ± 6.99 13.05 ± 11.01 66.96 ± 53.95 — -12.48 ± 4.72 t9 = -1.235 0.245 CSF 11 5.9 ± 2.16 7.85 ± 5.81 28.1 ± 38.59 — -6.04 ± 2.15 t9 = -0.996 0.343 SNR 10 18.0 ± 3.13 18.8 ± 2.75 18.05 ± 15.32 — -3.72 ± 2.12 t8 = -0.597 0.565 Cre 10 6.11 ± 0.47 6.49 ± 0.7 6.85 ± 8.64 — -0.97 ± 0.21 t8 = -1.835 0.1 us Cho 10 1.84 ± 0.24 1.97 ± 0.26 8.29 ± 11.43 — -0.38 ± 0.12 t8 = -1.655 0.132 mIno 10 4.46 ± 0.75 4.64 ± 0.6 11.92 ± 12.29 — -0.85 ± 0.49 t8 = -0.745 0.476 tNAA 10 8.23 ± 0.54 8.6 ± 1.03 8.96 ± 10.27 — -1.19 ± 0.45 t8 = -0.914 0.384 Thalam Glu 10 5.95 ± 0.62 6.65 ± 1.19 16.17 ± 13.34 — -1.64 ± 0.23 t8 = -1.609 0.142 Gln 10 2.78 ± 0.94 3.38 ± 0.85 38.74 ± 23.39 — -1.48 ± 0.29 t8 = -1.529 0.161 Glx 10 8.72 ± 0.96 10.03 ± 1.26 18.89 ± 15.11 — -2.41 ± -0.2 t8 = -2.113 0.064 GABA 10 1.84 ± 0.28 1.74 ± 0.48 27.32 ± 17.99 — -0.29 ± 0.49 t8 = 0.542 0.601 Gln:Glu 10 0.48 ± 0.2 0.52 ± 0.16 32.45 ± 26.92 — -0.23 ± 0.14 t8 = -0.597 0.565

FWHM 10 7.51 ± 1.56 7.8 ± 1.53 13.05 ± 10.27 — -1.83 ± 1.24 t8 = -0.742 0.477 SNR 10 16.7 ± 3.85 17.6 ± 3.01 14.83 ± 11.5 — -4.32 ± 2.52 t8 = -0.818 0.434 GM 11 75.35 ± 4.14 76.26 ± 3.33 3.04 ± 2.35 — -4.41 ± 2.6 t9 = -1.065 0.312 WM 11 24.63 ± 4.16 23.71 ± 3.33 9.3 ± 6.55 — -2.6 ± 4.43 t9 = 1.075 0.308 CSF 11 0.02 ± 0.04 0.04 ± 0.06 — — -0.07 ± 0.03 t9 = -0.803 0.441 Cre 10 6.84 ± 0.71 7.1 ± 0.58 7.22 ± 4.76 — -0.9 ± 0.38 t8 = -1.525 0.162 Cho 10 1.67 ± 0.31 1.73 ± 0.28 7.9 ± 7.66 — -0.36 ± 0.23 t8 = -1.246 0.244 mIno 10 3.93 ± 0.76 3.46 ± 0.72 17.56 ± 16.84 — -0.26 ± 1.2 t8 = 2.21 0.054

Striatum tNAA 10 7.33 ± 0.64 7.44 ± 0.75 5.97 ± 3.17 — -0.8 ± 0.58 t8 = -0.701 0.501 Glu 10 5.97 ± 0.7 6.51 ± 0.54 12.4 ± 7.97 -0.82 -1.15 ± 0.08 t8 = -2.302 0.047* Gln 10 2.01 ± 0.97 2.42 ± 0.48 38.69 ± 46.79 — -1.16 ± 0.35 t8 = -1.843 0.098 Glx 10 7.98 ± 1.35 8.92 ± 0.73 14.27 ± 10.17 -0.82 -2.01 ± 0.13 t8 = -2.997 0.015* GABA 10 1.75 ± 0.47 1.88 ± 0.48 24.73 ± 15.27 — -0.6 ± 0.34 t8 = -0.87 0.407 Gln:Glu 10 0.34 ± 0.16 0.37 ± 0.08 37.46 ± 46.03 — -0.17 ± 0.09 t8 = -0.893 0.395 Metabolite concentrations are reported in mmol/L units. Abbreviations: aMCC = anterior mid-cingulate cortex; Asp = aspartate; CI =confidence interval; Cho = choline compounds; COV = coefficient of variation; Cre = (phospoho)creatine; CSF = cerebrospinal fluid (%); D = Cohen’s D; FWHM = full width at half maximum (Hz); GABA = γ-aminobutyric acid; GM = grey matter (%);Gln = glutamine; Glu = glutamate; Glx = glutamate plus glutamine; Lac = lactate; m-Ins = myo-inositol; NAA = N-acetylaspartate with an additional contribution from N-acetylaspartylglutamate; SNR = signal-to- noise ratio; Tau = taurine; WM = white matter (%). Glutamate 93

Figure 8.4: Representative aMCC 1H-MRS spectra of the frequency and phase-drift corrected data (red) and LCModel ﬁts (black) Glutamate 94

Figure 8.5: Representative THA 1H-MRS spectra of the frequency and phase-drift corrected data (red) and LCModel ﬁts (black) Glutamate 95

Figure 8.6: Representative STR 1H-MRS spectra of the frequency and phase-drift corrected data (red) and LCModel ﬁts (black) Glutamate 96

8.4.5 Correlation of Metabolite Concentrations with Clinical Variables

The multiple regression model using age and gender as covariates revealed a signiﬁcant negative correlation between GTS baseline measurements of striatal Gln concentrations and RVTRS (r=-0.52, P=0.012) (Figure 8.7A). The analysis also revealed a signiﬁcant negative correlation between thalamic Glu and PUTS (R=-0.47, P=0.017) (Figure 8.7B).

Figure 8.7: Correlation of absolute metabolite concentrations with clinical measures. The multiple linear regression model revealed signiﬁcant negative correlations between (a) left corpus striatal Gln concentrations and post-scan measurements of actual tic severity (modiﬁed Rush Video-based Tic Rating Scale; RVTRS), in addition to (b) bi-lateral thalamic Glu concentrations and pre-monitory urges (Premonitory Urge for Tics Scale; PUTS). Glutamate 97

8.4.6 Comparison of Glutamate/Glutamine separation at 3T and 7T

To inspect whether we were able to achieve a reliable separation between Glu and Gln using our acquisition and analysis strategy — which was enhanced by the improved line-width and SNR following frequency/phase drift correction and the removal of single motion-affected averages — we collected additional data to inspect the separation that is achieved at a higher-field strength. Specifically, PRESS data from the aMCC was collected from an independent sample of N=5 healthy controls at 3T and 7T using the same procedures as outlined in Section 8.3. Mean values and standard deviations of absolute metabolite concentrations for were as follows:

(a) 1H-MRS at 3T: 2.0±0.4 mmol/L for Gln; 8.4±0.3 mmol/L for Glu; and 10.4±0.6 mmol/L for Glx.

(b) 1H-MRS at 7T: 1.2±0.4 mmol/L for Gln; 7.9±0.4 mmol/L for Glu; and 9.1±0.6 mmol/L for Glx.

The subtle variability in the concentration estimates at different fields is of the order of previously reported results [298]. While we cannot exclude a tendency for slightly higher estimates of Gln and Glu at lower field, the observed Glu/Glx ratios (0.81±0.03 and 0.87±0.04 at 3T and 7T, respectively) were remarkably similar. This seems to indicate that the outlined acquisition and analysis strategies achieved a reasonable separation between Glu and Gln, deeming it plausible to report both values. Representative 3T and 7T data from the 5 subjects is illustrated in Figure 8.8 Glutamate 98

Figure 8.8: 1H spectra achieved at 3T and 7T to inspect Glutamate/Glu- tamine separation Glutamate 99

8.4.7 Inﬂuence of head displacement on spectral measures

To investigate the effect of head motion on spectral measures, we extracted metrics of motion from an echo-planar imaging (EPI) resting state functional MR imaging (rs-fMRI) sequence, which was acquired for each subject during the scanning procedures. The EPI sequence was acquired following the MP2RAGE scan and before the spectroscopic acquisitions with the following parameters: TR= 1.4 s, TE= 30 ms, field of view 202 mm, flip angle 69◦, 422 volumes, nominal isotropic resolution 2.3 mm3 (See Appendix D for further parameter details). A rigid-body transformation model with 6 degrees of freedom (DOF) was used to spatially register every volume of the rs-fMRI time series to a reference volume (here taken as the first volume) using MCFLIRT [305]. As the total scan time of the rs-fMRI time series was approximately double that of the time required for acquiring the averaged spectrum from one voxel, only the first 211 volumes of the rs-fMRI time series were taken to approximately equal the total scan time of the spectroscopic data. Framewise displacement (FD)—a scalar quantity that indexes displacements in head position for one volume, , with respect to its preceding volume [65]—was calculated

from the translations, dx, dy, dzalong the Cartesian axes and rotations, α (pitch), β (roll), γ (yaw), extracted from MCFLIRT as follows:

|FDi| = |∆dx,i| + |∆dy,i| + |∆dz,i| + |∆αi| + |∆βi| + |∆γi| (8.6)

where ∆dx,i = dx,i–1 – dx,i and similarly for the other rigid body parameters. Rotational displacements were converted from degrees to millimeters by calculating the displacement on the surface of a sphere of radius equal to 50 mm, which is approximately the mean distance from the cerebral cortex to the center of the head.

8.4.7.1 Intra-group diﬀerences in head motion

Previous work has indicated that the tendency to move during MRI data acquisition is a generally stable trait across subjects [64]. To investigate whether this notion applies to our sample, the mean FD (µFD) was calculated for each subject in the healthy control test-retest datasets. Means and standard deviations of µFD for the test and retest healthy control datasets (N=21) were 0.14±0.05 mm and 0.13±0.04 mm, respectively. Mean FDs of the control sample were signiﬁcantly correlated across sessions (R= 0.98, P= 1.2×10-8; Figure 8.9), indicating that, on average, the control subjects exhibited stable and reproducible head motion. We followed the same procedure and tested whether the patients exhibited any changes in head motion as indexed by FD before and after treatment. Means and standard deviations of µFD for the GTS patients at baseline and Glutamate 100

Figure 8.9: Consistency of head motion across scanning sessions . Correlation of head motion as indexed by FD in the (a) healthy control and (b) GTS patient groups across the two scanning sessions. Both controls and patients exhibited significant correlations in µFD indicating that the subjects exhibited a stable and reproducible tendency to move in the head-coil. following treatment with aripiprazole (N=12) were 0.21±0.14 and 0.26±0.18, respectively, exhibiting no statistical differences using paired-sample t-tests. Similar to the healthy control subjects, head motion in the patients also exhibited significant correlations across the off- and on-treatment sessions (R= 0.88, P= 1.3×10-4; Figure 8.9). This result indicates that the patients also exhibited stable and reproducible head motion during the scans. It is, thus, plausible to suggest that the consistent tendency of head movement of the patients in the off- and on-treatment conditions may not have biased intra-group comparisons of absolute metabolite concentrations.

8.4.7.2 Inter-group diﬀerences in head motion

Next we were interested in whether there were inter-group diﬀerences of motion between the patients and controls. As would be expected, the GTS patient group, which is, in essence, characterized by movement, exhibited a relatively higher baseline µFD (0.20±0.14 mm) than the control sample (0.14±0.0.5 mm), reaching statistical signiﬁ-

cance (t67= -2.5; CI-95%=-0.024 to 0.024; P=0.015). If not accounted for, such varying motion tendencies may lead to frequency and phase errors that may give rise to incoherent signal averaging, artefactual line broadening, lineshape distortions and reduced SNR, which may ultimately lead to biased inter-group comparisons. Previous work has indicated that motion-related signal loss and lineshape deterioration in single-voxel 1H- MRS can be restored by correcting phase and frequency offsets in the acquired signals Glutamate 101 prior to spectral averaging [306]. Consistently, noticeable qualitative improvements were observed in all frequency/phase drift corrected spectra following fitting with LCModel (see Figure 8.1D for an example of striatal data from one patient). Quantitatively, we observed significant improvements in FWHM and SNR for both the control and patient group following frequency/phase drift correction for all voxels. Paired sample t-statistical comparisons between uncorrected and frequency/phase drift corrected data revealed significant increases in spectral quality measures and absolute metabolite concentrations for both the control and patient groups (Table 8.6). As reported in Table 8.4. FWHM and SNR were fairly similar between patients and controls for all voxels and exhibited no statistical differences. This indicated that the quality of the compared spectra was similar and was unlikely to bias inter-group comparisons.

We further investigated whether frequency/phase drift corrected data of two control samples with significant differences in head motion would exhibit differences in FWHM, SNR, and absolute metabolite concentrations. As such, based on a threshold of µFD=0.14 mm, we divided the control sample (first examination; striatal voxel data) into two groups, where the ‘low-motion group’ (N=17, µFD=0.11±0.02 mm) and the ‘high-motion group’

(N=13, µFD=0.17±0.03 mm) exhibited statically significant differences in FD (t30=-6.1; CI-95%=-0.009 to 0.04, P= 1.33×10-6). Independent sample t-tests of corrected data between the high- and the low-motion group exhibited no significant differences in FWHM, SNR or absolute metabolite signals (see Table 8.7). Glutamate 102

Table 8.6: Inﬂuence of frequency/phase drift correction on spectral measures. Comparison of striatal spectral measures between uncorrected and frequency/phase drift corrected data in both the patient and control datasets.

Parameter N Uncorrected Corrected CI (95%) Statistic P-Value

FWHM 30 8.83 ± 1.75 8.05 ± 1.45 1.62 to 0.06 t28 = 2.86 0.008 SNR 30 18.17 ± 3.79 17.27 ± 2.80 2.65 to 0.85 t28 = 1.37 0.182 NAA 30 6.14 ± 2.63 7.13 ± 0.75 0.02 to 2.01 t28 = 2 0.054 Cho 30 1.34 ± 0.59 1.56 ± 0.19 0.0 to 0.46 t28 = 2.11 0.044

y Controls Cre 30 5.67 ± 2.43 6.77 ± 0.77 0.15 to 2.04 t28 = 2.25 0.032 Glu 30 5.22 ± 2.28 6.26 ± 0.60 0.16 to 1.91 t28 = 2.51 0.018 Gln 30 2.23 ± 1.26 2.75 ± 0.97 0.07 to 1.11 t28 = 2.55 0.016 Health Glx 30 7.46 ± 3.32 9.01 ± 1.23 0.24 to 2.87 t28 = 2.6 0.015

FWHM 32 9.04 ± 1.82 7.72 ± 1.57 2.18 to 0.46 t30 = 5.08 0.00002 SNR 32 16.31 ± 4.07 17.44 ± 3.38 0.78 to 3.03 t30 =2.57 0.015 NAA 32 5.25 ± 3.17 7.03 ± 0.73 0.61 to 2.94 t30 =3.16 0.0035 Cho 32 1.25 ± 0.77 1.63 ± 0.30 0.09 to 0.68 t30 =2.72 0.011

GTS Cre 32 4.96 ± 2.99 6.61 ± 0.73 0.54 to 2.75 t30 =3.05 0.0047 Glu 32 4.41 ± 2.65 6.09 ± 0.68 0.69 to 2.66 t30 =3.61 0.0011 Gln 32 1.61 ± 1.20 2.16 ± 0.84 0.02 to 1.07 t30 =2.4 0.023 Glx 32 6.02 ± 3.67 8.24 ± 1.19 0.84 to 3.6 t30 =3.28 0.0026 Metabolite concentrations are reported in mmol/L units. Abbreviations: CI =conﬁdence interval; Cho = choline compounds; Cre = (phospoho)creatine; FWHM = full width at half maximum (Hz); Gln = glutamine; Glu = glutamate; Glx = glutamate plus glutamine; NAA = N-acetylaspartate with an additional contribution from N-acetylaspartylglutamate; SNR = signal-to-noise ratio.

Table 8.7: Comparison of frequency/phase corrected spectral data between high and low motion control groups

N Low-motion High-motion CI (95%) Statistic P-value (Low-/High-motion)

-6 FD 17/13 0.11± 0.02 0.17 ± 0.03 0.09 to 0.04 t28 = 6.1 1.34×10 FWHM 17/13 7.68 ± 1.38 8.53 ± 1.38 1.93 to 0.23 t28 = 1.6 0.12 SNR 17/13 17.59 ± 2.64 16.85 ± 2.96 1.43 to 2.91 t28 = 0.7 0.5 NAA 17/13 7.07 ± 0.60 7.22 ± 0.90 0.74 to 0.44 t28 = 0.53 0.6 Cho 17/13 1.57 ± 0.18 1.56 ± 0.20 0.14 to 0.16 t28 = 0.15 0.89 Cre 17/13 6.69 ± 0.70 6.86 ± 0.85 0.77 to 0.43 t28 = 0.60 0.56 Glu 17/13 6.12 ± 0.60 6.44 ± 0.53 0.77 to 0.12 t28 = 1.50 0.15 Gln 17/13 2.71 ± 0.87 2.81 ± 1.08 0.86 to 0.65 t28 = 0.28 0.78 Glx 17/13 8.83 ± 1.33 9.25 ± 1.03 1.38 to 0.52 t28 = 0.93 0.36 Metabolite concentrations are reported in mmol/L units. Abbreviations: CI =conﬁdence interval; Cho = choline compounds; Cre = (phospoho)creatine; FWHM = full width at half maximum (Hz); Gln = glutamine; Glu = glutamate; Glx = glutamate plus glutamine; NAA = N-acetylaspartate with an additional contribution from N-acetylaspartylglutamate; SNR = signal-to-noise ratio. Glutamate 103

8.4.7.3 Inﬂuence of head motion on voxel compartmentation and absolute metabolite quantitation

Given that the GM, WM and CSF compartments exhibit differences in the concentrations of both water and metabolites observed via 1H-MRS [307–309], we further investigated the influence of head motion on the combined metabolic signal arising from different tissue compartments. Our reasoning here is that unless metabolites exhibit the same concentration in all tissue types, subtle shifts in voxel position during 1H-MRS data acquisition would lead to changes in the fraction of within-voxel tissue content and subsequent bias in the combined signal, which differentially arises from GM and WM (no observable spectral signal arises from CSF). To explore the influence of variable partial volumes on the arising metabolic signal, we used the 6-DOF motion parameters extracted from the rs-fMRI data to estimate the extent of voxel shift that could occur over the time-course of 1H-MRS data acquisition. Although the motion indices extracted from the preceding rs-fMRI acquisition will not exactly reflect the time-course of head translations/rotations during the 1H-MRS data acquisition, the assumption we make here is that, on average, the tendency for a subject to move is consistent across sequences and sessions [64]. In this case, we focused on a patient with relatively high motion indices (µFD mm). To estimate within voxel compartmentation during movement, the transformation matrices extracted from the 6-DOF rigid body registration model of the rs-fMRI data were applied to the binary mask of the 1H-MRS aMCC voxel to create a four-dimensional image representing the exhibited voxel shift. An example from a patient examination is shown in Figure 8.10. To consider the different signals arising from the varying amounts of GM and WM, we calculated the proportion of tissue content at each time point and estimated the combined metabolite Glu concentration, [Glu]=f GM

× [Glu]GM + f WM × [Glu] WM; where [Glu] GM and [Glu] WM were taken as 9.4 mmol/L and 4.5 mmol/L, respectively [310–312]. Mean estimates of within voxel tissue content were 0.75±0.006, 0.15±0.005 and 0.09±0.0007 for GM, WM and CSF, respectively. The concentrations of Glu arising from the diﬀerent compartments for the relatively strongly moving patient was 7.8±0.07 mmol/L. This data indicates that subject movement and subsequent variations in within voxel compartmentation did not have a strong inﬂuence on the combined metabolic signal arising from the GM and WM tissue. Glutamate 104

Figure 8.10: The eﬀect of subject movement on within voxel tissue content and metabolite concentration. The top three panels (translations, rotations, FD) illustrate the degree of movement in a relatively strongly moving patient over a scanning period of approximately 5 min. Estimates of within-voxel tissue fractions exhibited stability as illustrated in the tissue fraction and mean voxel location panels. The combined Glu signal arising from GM and WM exhibited only minor variation within a range of 7.70-7.86 mmol/L. Glutamate 105

8.5 Discussion

We hypothesized that glutamatergic signalling is related to pathophysiology of GTS, and aimed at investigating whether GTS patients exhibit alterations in Glu and Gln levels within three CSTC regions using 1H-MRS at 3T. We report significant reductions in striatal concentrations of Gln, Glx and the Gln:Glu ratio and thalamic concentrations of Glx in GTS when compared to healthy controls. In the subset of patients that underwent a four-week pharmacotherapy with aripiprazole, we observed significant increases in striatal Glu and Glx concentrations and trends for increases in striatal Gln and thalamic Glx when compared to baseline GTS measurements. On-treatment patients exhibited no significant differences in detectable metabolites when compared to controls, indicating that the treatment normalized patient metabolite values to control levels. No changes were observed for the aMCC at baseline or after treatment. Multiple regression analysis revealed a significant negative correlation between left striatal Gln levels and tic severity as measured by RVTRS. Correlation with the YGTSS total tic score was not significant possibly due to the fact that YGTSS assesses tic severity over a period of seven days, while RVTRS was used to measure actual tics on the day of the scan. Thalamic Glu levels were negatively correlated with premonitory urges as measured by PUTS, suggesting that urges could be viewed as core element of GTS. Although previous work has indicated that the striatal GABA is implicated in GTS pathophysiology [125–127], we did not observe any changes in GABA concentrations. This potentially owes to the fact that our PRESS parameters were optimized for Glu quantitation and the relative difficulty of reliably measuring GABA without utilizing spectral editing techniques (Figure 8.11) [313–316].

8.5.1 Altered Glutamate-Glutamine Cycling in GTS

A primary diﬀerence between amino acid neurotransmitters (e.g. Glu, GABA) and non-amino acid neurotransmitters (e.g. dopamine) is their relative ubiquity and their participation in numerous biochemical processes. Apart from its role in excitatory neurotransmission, Glu is intimately associated with (a) energy metabolism, serving as a key intermediate linking the metabolism of carbon and nitrogen; and (b) neurotransmitter synthesis, where it may serve as a precursor for other neurotransmitters (e.g. GABA) [285]. Given that neurons lack the enzymes necessary for anaplerosis, they are incapable of de novo synthesis of Glu or GABA. As such, the astrocytic synthesis of Gln from its precursors glucose or Glu plays an instrumental intermediary role in the replenishment of vesicular concentrations of synaptically released Glu and GABA (Figure 8.12). The homeostatic mechanisms involved in the synthesis, degradation, and traﬃcking of Glu, Glutamate 106

Figure 8.11: LCModel individual metabolite ﬁtting. Representative individual 1H-MRS spectrum and metabolite ﬁts achieved using LCModel for the aMCC voxel in a GTS patient.

GABA and Gln among the diﬀerent compartments are instrumental in maintaining the metabolic ﬂux of elements in the GABA-Glu-Gln cycle and the subtle balance between excitatory and inhibitory neurotransmission.

Current models segregate Glu into two time-limited compartments: (a) a neuronal compartment containing a larger Glu pool (≈80%) characterized by a slow turnover rate; Glutamate 107 and (b) a glial compartment containing a smaller Glu pool (≈20%) [317, 318]. The shuttling of metabolites among the different compartments requires the active transport of metabolites via neuronal and astrocytic transport elements against their concentration gradients. 13C-labeled glucose MRS studies have shown that changes in the cycle flux rate are closely matched by changes in the rate of astrocytic glycolysis to support energy-dependent uptake [319–321]. More recent work has provided evidence that local Glu-Gln cycling is necessary to sustain synaptic excitatory transmitter release [322]. As the static concentration of Glu measured by MRS reflects the total Glu pool size which is involved in varied functions (e.g. oxidative metabolism, neurotransmission), caution is often required when attributing changes in Glu or the composite Glx signal to the integrity of glutamatergic neurotransmission [323]. The metabolic cooperation exhibited by the astrocytic-neuronal system to maintain the uptake of Glu for subsequent Gln synthesis, indicates that MRS-visible changes in Gln or the Gln:Glu ratio may represent a more faithful reflection of the GABA-Glu-Gln cycle flux and the integrity of neurotransmission [324, 325]. As such, the observed decreases in striatal concentrations of Gln, Glx and the Gln:Glu ratio indicate that patients with GTS exhibit a dysfunctional astrocytic-neuronal coupling system, thus suggesting an abnormality in GABA-Glu-Gln cycling, which may perturb the subtle balance between excitatory and inhibitory neurotransmission. Given that the neurotransmitter NAAG also exhibits close metabolic links to GABA and Glu [326], it is plausible to suggest that it may exhibit alterations in GTS. However, as we were unable to achieve a reliable separation between NAA and NAAG, our analysis had to be limited to the sum of both metabolite, tNAA, which may be insensitive to changes in NAAG [327].

Our results are buoyed by several gene-based studies that have provided evidence for abnormalities in the cellular trafficking machinery of Glu. Genome scans have provided evidence for linkage to 5p13-q11.2, a region which overlaps with the genomic region encoding the glial Glu transporter gene 1 (EAAT1 or SLCA13) [131, 137, 139]. Other genetic-functional work has indicated that a missense variant of a highly conserved residue (E219D) in the gene that encodes that main astrocytic Glu transporter (SLC1A3), conveys a significant increase in Glu uptake in a patient with GTS [138]. More recent work utilizing a gene-set enrichment analysis centered on functionally compartmentalized genetic clusters (neuronal, astrocyotic, oligodendrocytic and microglial gene-sets) has uncovered a significant association within a single gene-set representing the astrocyte carbohydrate metabolism pathway within the astrocytic cluster [287]. A focused follow-up analysis revealed that association to GTS was the result of an overall, combined effect of a subset of genetic variants encoding elements involved in the Glutamate 108 metabolism of Gln, GABA, pyruvate, lactate, glycogen and tri-carboylic cycle intermedi- ates. Coupled with our results, this data suggest that the genetic alterations in the mechanisms governing glycolytic, glutamatergic, glutaminergic and GABAergic metabolism influence the astrocytic modulation of neuronal function. Ultimately, this may lead to spatially focalized alterations in glutamatergic and GABAergic neurotransmitter ratios in specific striatal subdivisions that would have profound effects on the afferent modulation of dopaminergic discharge and thalamo-cortical output.

8.5.2 The Role of Functionally Selective Modulators in the Adaptive Stabilization of Neurotransmitter Systems in GTS

As an atypical second-generation antipsychotic, aripiprazole presents an example of a functionally selective drug that exhibits adaptive pharmacological efficacy that is dependent on the local levels of the endogenous ligand: dopamine [171]. The concept of functional selectivity postulates that a ligand may produce a mix of effects through the activation or inhibition of a limited number of signal transduction pathways as a result of its ability to induce unique G protein-coupled receptor conformations [329, 330]. In contrast to other antipsychotics, aripiprazole exhibits submaximal intrinsic activity despite its high occupancy leading to differential actions on D2 receptors, where it may act as (a) an antagonist when dopamine concentrations are high and (b) a partial agonist when dopamine concentrations are low [171]. Owing to its unique pharmacological profile it has been suggested that aripiprazole may act differentially on dopaminergic release, suppressing phasic release relatively more than tonic release [331, 332].

While the differential modulation of tonic and phasic dopamine release may provide an explanation for aripiprazole’s therapeutic effects in neuropsychiatric disorders, studies have indicated that the complex and potent effects of aripiprazole may be a result of the influence it is able to exert on extra-dopaminergic systems. Apart from its influence on the serotonergic system, gene expression studies have indicated that aripiprazole also modulates the glutamatergic and GABAergic neurotransmitter systems [171]. Specifi- cally, the chronic use of aripiprazole has been demonstrated to (a) induce the mRNA expression of NMDA receptor subunits enhancing glutamatergic neurotransmission in hippocampal and cortical regions [333]; (b) suppress the mRNA expression of glutamatergic astrocytic transporter genes (EAAT1, EAAT2) as well as neuronal transporter genes in hippocampal (EAAT3) and frontal areas (EAAT4); and (c) enhance the expression of presynaptic vesicular transporters in the hippocampus [334]. Given the significant role exhibited by glutamatergic transporter elements in the astrocytic-neuronal coupling system, decreases in the expression of astrocytic-neuronal transporter elements and increases in the expression of vesicular transporter elements indicate an increase Glutamate 109

Figure 8.12: Astrocytic-neuronal coupling and the homeostasis of glutamatergic and GABAergic neurotransmission. The schematic illustrates key processes involved in the cycling and metabolism of Glu and GABA in neurons and astrocytes. The rapid transamination of the tri-carboylic cycle (TCA) intermediate α- ketoglutarate (α-KG) predominates the formation of Glu in both neurons and astrocytes. As the unbranched penta-carbon skeleton of Glu only differs from GABA by a single carboxyl group, Glu and GABA exhibit close metabolic links; where GABA is directly synthesized from Glu via glutamate decarboxylase (GAD65 or GAD67) and is catabolized into Glu via transamination [285]. Given that neurons lack the enzyme pyruvate dehydrogenase, which is needed to link glycolytic metabolism to the TCA cycle, they are unable to replenish their vesicular concentrations of synaptically released Glu. As such, Gln plays an instrumental role in maintaining the homeostasis of Glu and GABA. Once released into the synaptic cleft, Glu is actively shuttled into astrocytes where it is either converted to glutamine via the glial exclusive enzyme glutamine synthetase (GS), or assimilated into the TCA cycle for energy metabolism (a). As a non-toxic and non-neuroactive equivalent, Gln is released by astrocytes where it is taken up by pre-synpatic neurons to serve as a Glu precursor. On the other hand, the majority of synaptically released GABA is taken up by pre-synpatic neurons where it is able to re-enter the vesicular pool or enter the TCA cycle via the GABA shunt (b). A fraction of synaptically released GABA is also taken up by astrocytes where it eventually generates Gln following successive steps in the GABA shunt and the TCA cycle [328]. Once released by astrocytes, Gln is taken up by pre-synpatic GABAer- gic neurons where it is serves as a precursor for GABA via glutamate. As such, the homeostatic mechanisms involved in the synthesis, degradation, and trafficking of Glu and GABA are instrumental in the maintenance of the subtle balance of excitation and inhibition. The observed decreases in striatal concentrations of Gln, Glx and the Gln:Glu ratio are consistent with the assumption that patients with GTS exhibit a dysfunctional astrocytic-neuronal coupling system. Perturbations in the GABA-Glu-Gln cycle flux may ultimately lead to spatially focalized alterations in glutamatergic and GABAergic neurotransmitter ratios in specific striatal and pallidal subdivisions that would have profound effects on the afferent modulation of dopaminergic discharge and thalamo-cortical output. Glutamate 110 in glutamatergic neurotransmission from hippocampal and frontal cortical regions. On the other hand, aripiprazole has also been demonstrated to influence the mRNA expression of glutamate decarboxylase (GAD67), leading to decreased GABA synthesis in the striatum, prefrontal cortex and somatosensory cortex [335].

Considering (a) the observed decreases of striatal Gln, Glx and the Gln:Glu ratio at baseline; (b) increases in striatal Glu and Glx during treatment; (c) previous work implicating alterations in genes encoding glutamatergic transporters and implying increases in Glu uptake [137–139]; (d) recent data indicating down-regulations in genetic modules encoding glutamatergic synaptic elements [288]; (e) reported decreases of GABAA receptor binding in striatal, pallidal, thalamic and limbic regions [127]; and (f) reported increases of striatal GABA in the supplementary motor area at baseline [130]; it seems plausible to assert that aripiprazole stabilizes the balance between excitatory and inhibitory neurotransmission in GTS, leading to decreases in tonic cortical GABAergic inhibition and increases in excitatory glutamatergic input feeding into the striatum from frontal and hippocampal areas. Along this line of reasoning, it is possible that the recently published data indicating that agents specific to the glutamatergic system (D-serine, riluzole) are not effective in tic suppression [336], since they do not exhibit a functionally selective and adaptive pharmacological profile that targets the multiple systems implicated in GTS.

8.5.3 The Inﬂuence of Subcortical GABA-Glu-Gln Cycling Abnormal- ities on Dopaminergic Signalling and the Phenomenology of Tics

In view of the spatio-temporal synergy exhibited by excitatory, inhibitory and modulatory neurotransmitter systems, we consider the effect of the observed neurochemical perturbations on the local circuit model of subcortical connectivity (Figure 8.13). It has been proposed that tics are generated through the repetitive inappropriate activation of specific striatal sub-territories, leading to the burst-like disinhibition of thalamo-cortical output [76, 78], with topography playing a key role in the manifesting clinical feature [75]. Previous work has demonstrated that striatal neurons code the serial order of syntactic natural behaviour [337], and that the integrity of dopaminergic receptors, transporters and the neurons innervating them are necessary for instinctive sequential stereotypy [338, 339]. While other neurotransmitter systems implicated in GTS (mainly cholinergic, serotonergic, histaminergic) are known to strongly influence dopaminergic signals within the striatum [31, 134], the excitatory and inhibitory dissociable afferent systems influencing tonic/phasic dopaminergic output towards the striatum hold a key role in coordinating cortico-striatal input and output selection [121, 283, 284]. Glutamate 111

Figure 8.13: Local circuit model of subcortical connectivity. The schematic illustrates the interaction of major neurotransmitter systems involved the regulation of cortico-striatal input and thalamo-cortical output. In essence, the basal ganglia op- erate through inhibition, dis-inhibition and facilitation, where the intrinsic quiescence of striatal neurons permits the output nuclei (GPi, SNR) to hold the thalamus under GABA-mediated tonic inhibition to prevent inappropriate movements. Select movements are facilitated upon the activation of striatal MSN populations that reduce the tonic pallido-thalamic output. Coordinated behaviour is thought to ensue as a result of the integrated output of the D1- (direct-pathway, shaded green) and D2- (indirect- pathway, shaded red) receptor mediated MSNs, whose outputs arbitrate the selection of the ultimate program [123, 340]. Dopamine exerts a powerful influence over output by modulating cortico-striatal afferents innervating MSNs as well as tonically active cholinergic interneuron (TAN) populations, which provide another level of control of MSNs via widespread axonal collaterals that regulate the fast-spiking GABAergic interneuron (FSI) populations [122, 341, 342]. Alterations in the dissociable afferent systems influencing tonic/phasic dopamine discharge from the VTA and SNc may have an influence on modulation of cortio-striatal input and thalamo-cortical output. Previous work has indicated that the maintenance of the tonic dopaminergic pulse is also dependent on glutamatergic and GABAergic signals flowing through the hippocampal-ventral tegmental loop (not shown) [119, 120]. On the other hand, phasic dopamine signalling is critically dependent on an NMDA-receptor mediated glutamatergic excitatory drive emanating from prefrontal, subthalamic and peduncolpontine regions [119, 343]. Consequently, chronic perturbations in the flux of metabolites in the GABA-Glu-Gln cycle, could lead to focal alterations in excitatory and inhibitory neurotransmitter ratios and subsequent abnormalities in the archetypical dynamics of tonic/phasic dopamine signalling. This notion is supported by our findings and previous reports indicating reductions and altered distribution of striatal and pallidal FSI and TAN populations. Such perturbations would have a profound influence on the neuro-plastic mechanisms involved in reinforcement learning and habit formation systems that are governed by striatal neurons that code the serial order of syntactic natural behaviour. Glutamate 112

Typically, the tonically driven, extra-synaptic release of dopamine into the striatum provide potent regulatory signals that influence the synaptically-focused phasic release of dopamine by acting on D2 autoreceptors [118]. As such, alterations in the tone of the intrinsic tonic pulse could directly influence the magnitude of the phasic response and subsequent thalamo-cortical output [121]. Previous work has indicated that the maintenance of the tonic dopaminergic pulse is also dependent on glutamatergic and GABAergic signals flowing through the hippocampal-ventral tegmental loop [119, 120]. In this system, glutamatergic output from the hippocampus (ventral subbiculum) towards striatal GABAergic neurons (nucleus accumbens), causes a decrease of the pallidal inhibition of dopaminergic neurons (ventral tegmental area), thus allowing their transition to a tonically active state [119]. On the other hand, as phasic dopamine signalling is critically dependent on an NMDA-receptor mediated glutamatergic excitatory drive, reductions in excitatory input feeding into dopaminergic nuclei would cause alterations in the phasic response. The significant influence of afferent glutamatergic input innervating dopaminergic nuclei was recently demonstrated by Wang et al. [344], who observed habit-learning deficits in mice with dopamine neuron specific NMDA-1 receptor deletions. The notion of extant irregularities in the dissociable afferent systems influencing dopamine release in GTS, is further supported by reported (a) ex-vivo reductions of Glu concentrations within nigral and pallidal areas [124]; (b) ex-vivo reductions and alterations in the distribution of striatal and pallidal inhibitory interneurons [125, 126]; and (c) in-vivo increases in the binding capacity of nigral GABAA receptors [127]. As such, chronic perturbations in the subcortical GABA-Glu-Gln cycle flux could lead to spatially focalized alterations in excitatory and inhibitory neurotransmitter ratios and subsequent abnormalities in the archetypical dynamics of tonic/phasic dopamine signalling. Such perturbations would have a profound influence on the neuro-plastic mechanisms involved in reinforcement learning and habit formation systems, which are governed by striatal neurons that code the serial order of syntactic natural behaviour.

8.5.4 Methodological Limitations and Future Directions

Given that patients with GTS are characterized by movement, head displacements during acquisition may have influenced metabolite quantitation. Although we endeavoured to circumvent the potential bias that could be introduced by motion, we cannot completely assert that the motion bias was entirely removed from the data. A more straightforward approach to avoid such bias would be possible with the integration of prospective motion- correction during data acquisition [345, 346]. Nevertheless, we further investigated the potential influence of within-scan head displacement on spectral measures by performing simulations based on realistic motion parameters and observed no significant effects Glutamate 113 introduced by motion (see Section 8.4.7). Another potential limitation in our study is that we did not perform multiple comparison corrections, which seems to be a relatively common condition in many hypothesis driven MRS studies. We believe that our study warrants replication in a larger independent sample focusing on the quantitation of Glu, Gln and GABA. This would particularly benefit from higher magnetic fields (e.g. 7T) due to improved sensitivity [298]. The assessment of the relationship between spectral measures of Glu, Gln, and GABA in addition to PET measures of dopaminergic signalling, presents a further promising prospect for the elucidation of the nature of GTS pathophysiology.

8.6 Conclusions

In this work, we investigated the neurochemical profile of a well-characterized sample of adult patients with GTS at baseline and following a four-week treatment with the antipsychotic aripiprazole using 1H-MRS at 3T. To obtain spectra of sufficient precision to identify rather subtle metabolic changes, we applied an automated voxel (re-)localization technique, frequency and phase error correction, absolute metabolite quantitation with the consideration of within voxel compartmentation and a careful quality assessment protocol. Our results implicated the flux of metabolites in the GABA-Glu-Gln cycle, thus implying perturbations in subcortical astrocytic-neuronal coupling systems that maintain the subtle balance between excitatory and inhibitory neurotransmission. Such perturbations may ultimately lead to spatially focalized alterations in excitatory, inhibitory and modulatory neurotransmitter ratios in functionally distinct striatal subdivisions, thus leading to the diverse symptomatology associated with GTS. Chapter 9

Elemental investigation of pathophysiology

This chapter presents an investigation of the role of iron in GTS pathophysiology. Quanti- tative Susceptibility Mapping values extracted from subcortical regions served as surrogate measures of iron content. The accurate and non-biased estimation of subcortical susceptibility values was signiﬁcantly improved by the preceding methodological investigation presented in Chapter 7. Blood samples were also acquired for serum ferritin quantiation to assess iron storage capacity. A multi-parametric approach was additionally used in which the relationship between subcortical susceptibility and spectroscopic measures of glutamatergic signalling were investigated. This work underwent a peer review process and was presented at the 2017 annual conference of the International Society of Mag- netic Resonance Imaging in Medicine in Honolulu, Hawaii, USA : Kanaan AS. et al., QSM meets MRS: The inﬂuence of subcortical iron on glutamatergic neurotransmission in a movement disorder population; Proc. Intl. Soc. Mag. Reson. Med. 25, 2017, Program No. 4649 [347].

9.1 Abstract

Gilles de la Tourette Syndrome (GTS) is a neuropsychiatric movement disorder fundamentally characterized by tics and a complex genetic-environmental aetiological basis. Previous work has indicated that the clinical manifestations of GTS are primarily driven by putative abnormalities in dopamine GABA, and glutamate. In view of the crucial role exhibited by the element iron in varied biochemical mechanisms sustaining developmental processes and neurochemical pathways, we postulated that patients with GTS exhibit abnormalities in iron metabolism, which may inﬂuence mechanisms of subcortical neurotransmission. In this work, we used Quantitative Susceptibility Mapping, a recently established Magnetic Resonance technique, to estimate magnetic susceptibility

114 Iron 115 as a proxy measure for iron in 22 healthy controls and 22 patients with GTS for the first time. We additionally utilized previously acquired 1H-MRS data to inspect the relationship between iron and glutamatergic signalling. We demonstrate that patients with GTS exhibit significant reductions of magnetic susceptibility in basal ganglia, brainstem and cerebellar nuclei. Reductions were specific to the striatum, substantia nigra, subthalamic nucleus and the red nucleus and were mirrored by decreases in serum ferritin levels. Significant correlations were observed between striatal – as well as total subcortical – magnetic susceptibility and striatal Gln:Glu. Our results indicate the presence of extant abnormalities in iron metabolism, which may exhibit an influence in GABA-Glu-Gln cycling. Perturbations in mechanisms regulated by iron-containing enzymes, may lead to disruptions in the spatio-temporal dynamics of excitatory, inhibitory and modulatory neurochemical systems, thus driving the manifesting clinical features in GTS.

9.2 Introduction

Iron is a trace element that is essential to the vitality of an organism as it is ideally suited for the catalysis of many biochemical reactions due to its ability to transition between two thermodynamically stable oxidation states [348]. Its flexibility in the trafficking of electrons renders it a crucial component of prosthetic groups (e.g. hemes and iron sulphur clusters) located within diverse proteins involved in oxygen transport, mito- chondrial energy metabolism, protein synthesis, in addition to neurotransmitter synthesis and transport [348]. Mediated by the blood-brain barrier, the acquisition of non-heme iron in the brain occurs in an age- and regionally-dependent manner, where its deposition during a “critical period" is necessary for normal development [349, 350]. The deposition of non-heme iron is sharpest in subcortical grey-matter, where it overlaps with regions that contain dense proportions of the neurotransmitters dopamine, γ-aminobutyric acid (GABA) and glutamate (Glu) [351]. The atypical homeostasis of iron during different periods of development may influence the biochemical mechanisms that sustain typical neurochemical metabolism and myelination [161, 352], providing a biological basis for the acquisition of abnormalities in motor and behavioural functions as exhibited by various developmental neuropsychiatric/movement disorders [151, 153, 353].

Gilles de la Tourette syndrome (GTS) presents an example of a disorder with motor and behavioural deficits as a result of fundamental alterations in the dynamics of cortico- striatal circuitry [28]. In essence, GTS is characterized by the presence of multiple motor and vocal tics and a high incidence of comorbid features, which may include Atten- tion Deficit/Hyperactivity Disorder (ADHD), Obsessive-Compulsive Behavior/Disorder (OCB/D), depression, and anxiety [9, 18]. Though the underlying pathophysiological Iron 116 mechanisms of GTS have not been completely elucidated, current data suggests that acquired abnormalities in habit formation systems [280, 282] may be driven by deficits in subcortical neurochemical signalling [31, 274], thus leading to a burst-like disinhibition of thalamo-cortical output [76, 78]. Methodologically varied work has indicated that patients with GTS exhibit abnormalities in (a) D2 receptor binding; (b) dopamine active transporter density/binding; and (c) phasic dopamine transmission, thus suggesting putative abnormalities in the functional dynamics of tonic and phasic dopaminergic signalling [31]. Given the significant role exhibited by excitatory and inhibitory afferents on the dopaminergic nuclei [118, 121], these effects may be driven or further compounded by alterations in the GABAergic and glutamatergic neurotransmitter systems. Along this line, abnormalities in the GABAergic system in cortical and subcortical regions have been demonstrated via 1H-Magnetic Resonance Spectroscopy (1H-MRS) [130, 286], positron emission tomography (PET) [127] and quantitative post-mortem studies [125, 126]. More recently, our group has demonstrated that GTS patients exhibit abnormalities in glutamatergic metabolism, where reductions of Gln, Glu+Gln (Glx), and the Gln:Glu ratio were observed in the striatum via 1H-MRS [274]. Put together, this data implies the presence of abnormalities in GABA-Glu-Gln cycling, which may lead to alterations in the spatio-temporal dynamics of excitatory, inhibitory and modulatory subcortical neurochemical signalling.

One unifying feature exhibited by excitatory, inhibitory and modulatory neurotransmitters is that the enzymes involved their metabolism and the production of their receptors and transporters require iron for typical function. The effects of iron on the dopaminergic system are well recognized as its deficiency during specific periods of development has been demonstrated to lead to deficits in the synthesis, catabolism, transport and uptake of dopamine [161]. More specifically, iron deficiency has been shown to lead to alterations in the function of (a) D1/D2 receptors; (b) dopamine active transporters; and (c) the dopamine catabolic enzymes tyrosine hydroxylase and monoamine oxidase B [161]. Although work linking iron to Glu/GABA metabolism is less extensive, studies have indicated that iron plays an important functional role in both excitatory and inhibitory neurotransmitter metabolism and transport. Specifically, dietary iron deficiency at different periods of prenatal and postnatal development has been demonstrated to lead to (a) downregulations in the key GABA-Glu-Gln cycle enzymes Glu decarboxylase, Glu dehydrogenase, GABA transaminase and isocitrate dehydrogenase [354–356]; (b) reductions of GABA content in pallidal, striatal and hippocampal regions [357, 358]; (c) reductions in the binding of Glu to vesicular membranes [355, 359]; and (d) the suppression of glutamatergic neurotransmission in striatal and hippocampal regions [360, 361].

Given this data, it seems plausible to hypothesize that patients with GTS may exhibit an abnormality in the cerebral homeostasis of iron. In general, a link between iron Iron 117 deficiency and an increased risk of contracting neurodevelopmental psychiatric disorders (e.g. ADHD, autism spectrum disorder) is well established [362, 363]. The notion of disturbed iron homeostasis in GTS is supported by preliminary work indicating that the patients exhibit reductions in serum ferritin [163, 316, 364, 365]. Ferritin is a stable and reliable indicator of total iron content in the body, as it is the main intracellular iron sequestering protein [366]. However, because serum ferritin levels do not necessarily reflect iron concentrations in specific organs, estimates of regional iron content by non- invasive and in-vivo magnetic resonance imaging (MRI) techniques are of paramount interest.

In essence, non-heme iron sequestered in ferritin exhibits paramagnetic properties that lead to the induction of magnetic moments when applied to an external magnetic field. The dominant magnetic susceptibility of subcortical iron rich structures induces field perturbations that can be non-invasively measured via magnetic resonance (MR) phase imaging techniques. Quantitative Susceptibility Mapping (QSM) is a recently established MRI technique that estimates the (relative) intrinsic magnetic susceptibility of tissue using the MR signal phase following (a) the estimation of the magnetic field distribution, (b) the elimination of background field contributions, and (c) solving the inverse problem from magnetic field perturbation to susceptibility [192, 367]. Considering that the paramagnetic properties of sequestered iron render it as the most dominant contributor to magnetic susceptibility in deep grey matter, a strong linear relationship between subcortical iron and magnetic susceptibility has been demonstrated and validated via multiple techniques [225, 227, 368, 369].

Consequently, the primary aim of this work was to investigate whether patients with GTS exhibit reductions in cerebral iron content as indexed by lower magnetic susceptibility. We focused on estimating magnetic susceptibility in a set of subcortical nuclei that have been implicated in GTS pathophysiology and relating these susceptibility measures to serum ferritin levels. Given the important role exhibited by iron containing enzymes that are integral to the tri-carboxylic acid cycle, we additionally postulated that iron will exhibit an association with excitatory signalling, which, to the best of our knowledge, is a relationship that has not been investigated simultaneously in vivo. To this end, we use the combination of QSM and 1H-MRS to investigate whether subcortical iron levels exhibit an association with glutamatergic neurotransmission as indexed by the Gln:Glu ratio. Iron 118

9.3 Materials and Methods

9.3.1 Population Sampling

The study was approved by the local ethics committees and all participants gave written informed for their participation. A total of 43 right-handed adult patients with GTS and 40 age/gender matched healthy controls were recruited as part of a larger study that included the acquisition of structural, functional, and spectroscopic Magnetic Res- onance (MR) data (see Chapter 8 [274]). Given that the pole-artifact in the vendor provided phase maps discussed in Chapter 7 was only observed after the beginning of data acquisition, the susceptibility weighted gradient-echo sequence was optimized to output multi-channel data for optimal coil-combination off-line. As such, the combination of T1-weighted, susceptibility-weighted (multi-channel), and spectroscopic MR data were acquired from a subsample that included 28 patients (4 female, 18-65 years) and 22 controls (5 female, 18-65 years). Patients using any psychoactive substances underwent a four-week washout period before participation and were deemed ineligible if they exhibited severe tics to the head and face, a history of other neurological disorders, current abuse of drugs and alcohol and MRI contraindications. All participants were diagnosed based on DSM-5 criteria and underwent a thorough clinical assessment battery as described in Chapter 5. Age- and gender-matched healthy control subjects without a history of neurological, psychiatric and tic disorders were recruited and assessed in a similar manner as the patients. All subjects were instructed to (a) not drink coffee or tea and to abstain from smoking for at least 2h before the examination and (b) adhere to a regular sleeping cycle the night before the scan. To minimize the variability that could arise from circadian physiological effects [290], the time of day of the MR exam was matched between patients and controls with the majority of acquisition conducted between 10AM and 4PM.

9.3.2 Measurement of serum Ferritin

A 10ml blood sample was collected from the majority of the subjects for the in-vitro quantitative determination of serum ferritin levels as a representative measure of the body’s iron reserves. The sample was ﬁrst centrifuged at 24,000 rpm for a period of 10 minutes to separate hematocrit from plasma, which was subsequently stored in 1000ml aliquots at -70◦C. Serum ferritin levels were quantiﬁed based on the electochemilumi- nescence immunoassay, in which a voltage applied to a sample containing tagged ferritin molecules induces chemiluminescent emissions that are measured by a photomultiplier (Elecsys 2010, Roche Diagnostics GmbH, Mannheim, Germany). Iron 119

9.3.3 Magnetic Resonance Imaging and Spectroscopy

Magnetic resonance measurements were performed on a 3T MAGNETOM Verio (Siemens Healthcare, Erlangen, Germany) equipped with a 32-channel head coil. The patients were instructed to remain still without actively suppressing their tics and thinking of nothing in particular. Magnetization-Prepared 2 Rapid Gradient Echo (MP2RAGE) and 1H- MRS data were acquired as described in Chapter8[274]. High quality spectra localized to the left-striatal voxel were obtained using a careful acquisition and processing scheme that incorporated (a) an automated voxel (re-)localization technique; (b) the removal of motion corrupted outlier signals; (c) frequency and phase drift correction in the time domain; (d) absolute metabolite quantitation with the consideration of within voxel compartmentation; and (e) a semi-automated quality assessment protocol. Susceptibility weighted data were acquired using a 3D-ﬂow compensated, spoiled, gradient recalled echo ◦ sequence with the parameters: TR=30s; TE=17ms; 256×256 matrix; ﬂip-angle=13 ; 0.8mm isotropic nominal resolution.

9.3.4 Quantitative Susceptibility Mapping

High-quality phase maps were reconstructed from the multi-channel complex signals using an automated, data-driven coil combination method. Specifically, a conjugate virtual-body-coil map was reconstructed by computing the singular value decomposition (SVD) across the channels, and then taking the dominant singular vector as the body coil reference. SVD compressed data were sorted in decreasing order of virtual channel eigen- values, and input into ESPIRiT for the estimation of coil sensitivities [262, 270]. QSM was computed using the superfast dipole inversion approach which employs (a) sophisticated harmonic artifact reduction for phase (SHARP) to eliminate background field contributions and threshold k-space division (TKD) for calculating magnetic susceptibility [273]. All QSM data were referenced to median cerebrospinal fluid susceptibility, which was computed within a subject specific mask of the lateral ventricles [219]. Motivated by previous morphometric, functional, spectroscopic, genetic and post-mortem work, carefully delineated masks of deep grey matter nuclei were generated as described the following section. Target regions of interest included the striatum (caudate-putamen), pallidum, thalamus, substantia nigra, subthalamic nucleus, red nucleus, dentate nucleus. An overview the data processing scheme is illustrated in Figure 9.1. Iron 120

Figure 9.1: Processing and analysis framework utilized to obtain high- quality quantitative susceptibility maps and 1H-MR spectra.

9.3.5 Masking of Subcortical Matter Nuclei

Masks of the striatum (caudate-putamen), globus pallidus and thalamus were obtained via the FSL-FIRST Bayesian model-based subcortical segmentation algorithm [297], which was applied on optimized hybrid-contrast MP2RAGE-QSM images [370]. Robust co-registration between skull-stripped MP2RAGE and Fast Low-Angle SHot (FLASH) data was achieved using rigid-body linear transformation of the T1-weighted data onto Iron 121

N4-bias field corrected FLASH magnitude data. Given the difficulty of segmenting brain stem and cerebellar nuclei on T1-weighted data due to lack of contrast, in addition to the infeasibility of performing manual segmentation of multiple nuclei in many subjects, we utilized an atlas-based registration approach to achieve accurate delineations of brainstem/cerebellar nuclei. Specifically, the diffeomorphic Greedy-SyN ANTS non- linear transformation model (https://github.com/stnava/ANTs) was used to compute a non-linear transformation warp between MP2RAGE and MNI space, which was used to map each subjects QSM data into standard space for subsequent calculation of a population-specific average image. The standardized QSM template exhibited high contrast in brainstem/cerebellar regions and was used to carefully delineate masks of the subthalamic nucleus, substantia nigra, the red nucleus and the dentate nucleus. All masks were delineated by the same operator and were subsequently warped back into native QSM space. The same atlas-based registration procedure was applied to obtain subject specific masks of the lateral ventricles, which were used for referencing the QSM data to CSF. All masks were thresholded at 0.5 to ensure maximal inclusion of grey- matter tissue while limiting partial volume effects. Following visual inspection of all the masks for quality, median susceptibility values from all regions of interest were computed for further analysis.

9.3.6 Quality Control

Given that patients with GTS are ultimately characterized by movement, head motion during MR data acquisition may influence voxel intensities and bias group comparisons. Consequently, we used a step-wise, multivariate outlier detection approach to remove low-quality data based on (a) structural image quality indices calculated on the magnitude image and (b) susceptibility values extracted from subcortical nuclei. Specifi- cally, a multivariate robust squared Mahalanobis distance framework was implemented to detect outlier datasets based on the Shannon entropy focus criterion (EFC), which is an index for image ghosting and blurring [371]; the Quality Index 1 (QI1), which is an index for image degradation resulting from bulk motion, residual magnetization, incomplete spoiling and ghosting [372], and the smoothness of voxels calculated as the full-width half maximum (FWHM) of the spatial distribution of image intensity values in voxel units (https://github.com/preprocessed-connectomes-project). This step was implemented on the whole sample and identified one severely affected dataset which was marked for removal (Figure 9.2A). To ensure that the remaining datasets did not contain further outliers, multivariate robust squared Mahalanobis distance outlier detection was additionally performed on vectors of median susceptibility values extracted from the subcortical masks for each sample separately. This procedure identified four outlier datasets Iron 122

Table 9.1: Statistical comparisons of magnitude image data quality metrics

Controls GTS CI (95%) Statistic P-Value

SNR -0.03±0.47 0.0±0.41 -0.31 to 0.24 U42=229 0.385 CNR 0.18±0.14 0.22±0.16 -0.14 to 0.05 U42=209 0.223 FBER 2.25±0.84 2.14±0.88 -0.42 to 0.65 U42=215 0.267 EFC 0.45±0.07 0.46±0.05 -0.04 to 0.03 U42=217 0.283 QI-1 0.08±0.05 0.1±0.05 -0.05 to 0.01 U42=200 0.165 FWHM 2.1±0.13 2.13±0.18 -0.13 to 0.07 U42=211 0.237 Abbreviations: SNR = Signal-to-noise ratio; CNR = Contrast-to-noise ratio; FBER = Foreground-to-background ratio; EFC = Entropy Focus Criterion; QI-1 = Quality Index 1; FWHM = Voxel Smoothness

within the patient sample which were marked for removal (Figure 9.2B). Following quality control, group comparison of magnitude image quality metrics (signal-to-noise ratio, contrast-to-noise ratio, voxel smoothness EFC, QI1) revealed no signiﬁcant diﬀerences between patients and controls (Table 9.1).

9.3.7 Statistical Analysis

Statistical analysis was performed in the Python programming language (Scipy v.0.15.1 and Statsmodels v.0.6.1). The normality of distribution and homogeneity of variance were assessed via the Kolomgrov-Simrnov and Levene’s tests, respectively. Group differences in serum ferritin values were assessed using Welche’s test given the inhomogeneous variance. Susceptibility data tended to exhibit non-parametric distributions, and as such, group differences of median susceptibility values for each region-of-interest were assessed using Mann-Whitney-Wilcoxon rank sum tests. The significance threshold for initial exploratory analyses of susceptibility values from combined regions of interest (basal ganglia, brainstem, all nuclei) was set to P <0.05 uncorrected. Group comparisons of magnetic susceptibility data from distinct nuclei were corrected using False-discovery- rate multiple comparison correction. A multiple linear regression model accounting for age, gender and two indices of image quality (EFC, QI1) was used to examine the relationship exhibited by susceptibility with (a) ferritin, (b) Gln:Glu ratio and (b) clinical scores. The variance inflation factor was used to assess multi-collinearity between predictor variables. Iron 123

Figure 9.2: QQ plots of the multivariate outlier detection technique implemented via squared Mahalanobis distance. (A) Multivariate outlier detection implemented on the magnitude image quality metrics of the whole sample. (B) Multi- variate outlier detection implemented on the susceptibility values of the patient sample. Iron 124

9.4 Results

Twenty-two healthy control subjects and 28 patients with GTS took part in the study. Multi-channel phase data acquired using a FLASH sequence at 3T were combined using a data driven method to generate high-quality phase maps that were used to reconstruct the QSM images. Median values of relative magnetic susceptibility were estimated from a set of carefully delineated subcortical masks to index iron in the brain. Collection of susceptibility weighted data from one patient was not completed due to claustrophobia. All data underwent rigorous quality control to ensure the reliability of the data input into the statistical models (see section 9.3.6). The remaining samples, which included 22 healthy controls (18-65 years, 5 female) and 22 patients with GTS (18-65 years, 4 female), were comparable in terms of age (t42=0.85, P =0.40), gender (odds ratio= 0.76 P =1.0) and handedness (all right handed). Demographic and clinical characteristics of the study sample included in the ﬁnal analysis are summarized in Table 9.2. The quality of data from single patient and the population average is illustrated in Figure 9.3

Table 9.2: Demographic and clinical characteristics of the QSM study sample included in the ﬁnal analysis

Controls GTS N 22 22 Age (Years) 38.59 ± 11.75 35.59 ± 11.01 Gender (M/F) 17/5 18/4 YGTSS-TTS — 19.59 ± 8.89 YGTSS-GS — 40.23 ± 17.13 RVTRS — 8.33 ± 4.24 PUTS — 18.55 ± 5.4 YBOCS — 1.86 ± 4.67 BDI-II 1.9 ± 3.5 11.59 ± 10.69 CAARS 39.05 ± 6.6 48.95 ± 11.35 Gln:Glu 0.47 ± 0.16 0.36 ± 0.13 Ferritin (ng/ml) 196.2 ± 151.26 88.93 ± 48.8 All the recruited subjects were right handed. Abbreviations: BDI-II = Beck Depression Inventory; CAARS = Conners’ Adult ADHD Rating Scales; OCD = Obsessive-Compulsive Disorder; OCI-R = Obsessive-Compulsive Inventory-Revised; QOL = Quality of Life scale; Y-BOCS = Yale- Brown Obsessive-Compulsive Scale; YGTSS-GS = Yale Global Tic Severity Scale Global Score; YGTSS-TTS = YGTSS Total- Tic Score.

9.4.1 Group diﬀerences in surrogate measures of iron

Group comparison of serum ferritin levels in the control (196 ± 151 ng/ml) and patient

(89 ± 49 ng/ml) samples revealed a statistically signiﬁcant reduction in GTS (t35=2.8, P =0.01). The reduction of body iron reserves in GTS was mirrored by statistically sig-

niﬁcant reductions of median magnetic susceptibility of bilateral brainstem (U42=111,

P =0.001), basal ganglia (U42=167, P =0.04) and all subcortical nuclei (U42=106, P =0.0007) Iron 125

Figure 9.3: QSM Data quality. Representative images from one GTS patient illustrating the quality of coil-combined magnitude (a) and phase (B) maps used reconstruct the QSM data (C). The red-overlays illustrate the quality of nucleus segmentation of basal ganglia (C), brainstem (E, F) and cerebellar nuclei (G). Data quality from the whole sample is represented as a standardized average QSM image as illustrated in (D, H, I, J).

(Table 9.3). These eﬀects were mainly driven by signiﬁcant reductions of susceptibility in the bilateral subthalamic nucleus (U42=94, P =0.0002), substantia nigra (U42=157,

P =0.024), striatum (U42=155, P =0.021), and red nucleus (U42=160, P =0.028) (Fig. 2),

and trends for reduction in the dentate nucleus (U42=175, P =0.056) (Table 9.4). Further inspection of these eﬀects using False-Discovery Rate (FDR) multiple-comparison correction revealed that the subthalamic nucleus, substantia nigra, striatum and red nucleus

passed FDR correction (P FDR<0.05). Cohen’s D effect sizes for these nuclei varied between 0.6-1.2 indicating practical significance. Post-hoc comparison of each hemisphere separately revealed that these effects were mainly driven by more pronounced reductions in left-hemisphere as illustrated in the split violin plots in Figure 9.4. Iron 126 δχ ) reduc- GTS patients exhibited significant susceptibility ( <0.05. FDR P denotes Nucleus segmentation quality and group comparison statistics. 9.4: Figure tions , passing false-discoveryeffects rate were multiple mainly comparison driven correction by in reductions the in red left nucleus, hemisphere substantia nuclei nigra, as subthalamic illustrated nucleus and by striatum. the split These violin plots. Reported P-values are uncorrected. * Iron 127

Table 9.3: Statistical comparison of magnetic susceptibility within general regions of interest

Controls Patients Cohen’s D CI (95 %) Statistic P-Value

Subcortical Grey Matter 44.7±9.1 33.9±9.6 1.13 0.01 to 0.02 U42=106 0.00074* Brainstem 114.6±23.8 90.5±22.2 1.02 0.01 to 0.04 U42=111 0.00109* Basal Ganglia 19.6±9.4 11.2±11.3 0.79 0.00 to 0.01 U42=167 0.0403* Bi-lateral median susceptibility values are reported in parts-per-billion (ppb) units. Brainstem includes the substantia nigra, subthalamic nucleus and the red nucleus. Basal Ganglia includes the caudate, putamen and pallidum. Subcortical grey matter includes the thalamus, dentate nucleus, brainstem and basal ganglia. The signiﬁcance levels was set at P <0.05 uncorrected.

Table 9.4: Statistical comparison of magnetic susceptibility within distinct subcortical nuclei

Controls Patients Cohen’s D CI (95%) Statistic P-Value P-FDR

Subthalamic Nucleus 80.3±24.2 49.1±27.6 1.18 0.02 to 0.05 U42=94 0.00027 0.00189* Striatum 8.7±13.1 -0.3±10.4 0.75 0.0 to 0.02 U42=155 0.0212 0.0488* Substantia Nigra 132.7±23.3 117.0±28.7 0.59 0.0 to 0.03 U42=157 0.0237 0.0488* Red Nucleus 130.7±45.1 105.3±34.5 0.62 0.0 to 0.05 U42=160 0.0279 0.0488* Dentate Nucleus 70.3±34.2 50.8±44.6 0.48 -0.01 to 0.04 U42=175 0.0593 0.083 Globus Pallidus 82.8±17.4 75.5±19.1 — 0.0 to 0.02 U42=197 0.148 0.173 Thalamus -37.1±11.8 -37.9±14.1 — -0.01 to 0.01 U42=224 0.341 0.341 Bi-lateral median susceptibility values are reported in parts-per-billion (ppb) units. The signiﬁcance levels was set at P-FDR(False Discovery Rate)<0.05.

9.4.2 Magnetic susceptibility correlations with ferritin, Gln:Gln and clinical measures

To explore the relationship between serum ferritin and magnetic susceptibility, we employed a multivariate linear regression model accounting for age, gender and two indices of image quality (EFC, QI1). Correlation analyses revealed significant associations between serum ferritin and magnetic susceptibility in the combination of all subcortical (R=0.65, P =0.009) and basal ganglia (R=0.44, P =0.02) nuclei 9.5, and trends for association with brainstem nuclei (R=0.63, P =0.062). To explore the relationship between iron and glutamatergic neurotransmission, we implemented a similar multivariate linear regression model accounting for age, gender, EFC and QI1. Significant associations were observed between striatal Gln:Glu and susceptibility in the striatum (R=0.65, P =0.002) and all subcortical nuclei (R=0.59, p=0.033) (Figure 9.6). All correlations exhibited an approximate variance inflation factor of 1.5 indicating little to no multi-collinearity between predictor variables. For the clinical scores, no significant correlations were observed with susceptibility. Iron 128

Figure 9.5: Ferritin group differences and correlations with susceptibility. (A) Bar-charts illustrating the significant reductions of serum Ferritin observed in GTS patients in comparison to controls. (B,C,D) Correlation plots illustrating the significant correlations observed between serum Ferritin and susceptibility in all subcortical, basal ganglia and brainstem nuclei. The strong correlations were observed in the whole sample and in each of the sub-samples separately. These results indicate the GTS patients exhibit disturbances in the homeostasis of iron. Iron 129

Figure 9.6: Correlational analysis between susceptibility and Gln:Glu (a) An exemplary striatal 1H-MRS spectrum of frequency and phase-drift corrected data collected from one GTS patients. The inset image illustrates the localization of the region of interest on an axial section. Signiﬁcant correlations were observed between striatal Gln:Glu and susceptibility in the same striatal region of interest (B), as well as the combination of all subcortical nuclei (C). These results provide in-vivo support for a relationship between iron metabolism and glutamatergic neurotransmission. Iron 130

9.5 Discussion

Using a combination of QSM and 1H-MRS, we investigated subcortical iron content and its association with glutamatergic signalling indices in GTS. We observed significant reductions of serum ferritin and magnetic susceptibility in the striatum, substantia nigra, subthalamic nucleus and red nucleus. Importantly, we also observed significant associations between striatal susceptibility (and the combination of all subcortical nuclei) with spectroscopic measures of striatal Gln:Glu. Put together, our results suggest that GTS patients exhibit disturbances in iron homeostasis in subcortical regions that typically exhibit the sharpest concentrations of iron and the densest proportions of dopamine, GABA and Glu. As the developmental processes of synaptogenesis, dendritogenesis and myelination are highly dependent on iron containing enzymes, subtle deficiencies of iron content throughout specific developmental epochs may ultimately influence mechanisms of subcortical neurochemical signalling and drive the acquisition of deficits in motor, affective and cognitive behaviours as manifested in GTS.

9.5.1 Disturbed iron homeostasis in GTS

The most consequential discovery of methodologically varied and comprehensive work investigating the role of iron in the developing brain, is that iron deficiency may lead to a specific set of behavioural outcomes that are dependent on its timing, severity and duration [349, 373]. Both human infant and animal model studies have consistently showed that early life iron deficiency leads to long-term abnormalities in motor, cognitive and affective behaviour, that are irreversible with iron repletion at weaning [153, 374, 375]. Particularly, children and adults with perinatal, neonatal and postnatal iron deficiency exhibited (a) cognitive impairments in selective attention, perceptual speed and inhibitory control; in addition to (b) affective impairments that included depressive and anxiety-like symptoms [374]. Within (c) the motor domain, early life iron deficiency in human infants was associated with delays in developmental motor mile- stones, lower global indices of motor function [374]. As the interpretation of human studies exploring early life iron deficiency may be confounded by overlooked differences in nutritional and environmental parameters, animal model studies offered further insight on specific motor abnormalities associated with iron deficiency. Specifically, the acute effects of early life iron deficiency included deficits in motor coordination and locomotor control [374]. On the other hand, long-term effects following periods of iron repletion during weaning periods revealed deficits in forelimb placing, increased hesitancy and most importantly, imperfect acquisition of grooming chains [376]. Iron 131

It is of note to emphasize that some of the symptoms exhibited by early life iron deficiency also manifest in GTS. Particularly these include abnormalities in syntactic motor chains, inhibitory control, selective attention and the presence of depressive/anxiety like behaviour [9, 373]. The correspondence in exhibited behaviour between iron deficiency and GTS may be reflected by similarities in their etiological basis, which may be driven by genetic or environmental factors. Considering that GTS is, in essence, a polygenic disorder with environmental influences [22, 24], (a) variants in genes involved in the regulation of iron metabolism; and (b) environmental factors influencing iron availability have both been reported in GTS. In a microarray study exploring the influence of gestational/lactational iron deficiency on mRNA expression levels, variants in genetic clusters involved in cytoskeletal stability and synaptic function exhibited chronic downregulations following iron repletion therapy at weaning [377]. Similarly, two independent studies have demonstrated associations between GTS and genetic variants in the BTDB9 gene, which contains a protein-protein interaction motif (BTB domain) implicated in cytoskeletal regulation, ion channel gating and transcriptome repression [378, 379]. As the cytoskeleton forms the primary backbone of neuronal architecture, in which the coordinated interplay of cytoskeletal elements is essential for axonal outgrowth, synapse formation and cargo logistics [380], abnormalities in cytoskeletal function driven by disturbances in iron homeostasis may affect synaptic efficacy and underlie some of the long-term behavioural outcomes manifested in GTS. On the other hand, gestational noxious exposures including maternal smoking and prenatal stressors have been linked to increased risk of both iron deficiency and the onset of GTS symptoms [22, 24]. Consequently, this data indicates that deficits in iron homeostasis hold an important role in the onset and course of GTS symptomatology and may be driven by a complex relationship between genetic and environmental influences.

9.5.2 Disruptions in iron homeostasis inﬂuence mechanisms of subcortical neurochemical signaling

In view of the significant influence exhibited by iron on excitatory, inhibitory and modulatory neurotransmitter systems, we consider the effect of the implied disturbances in iron homeostasis on subcortical neurochemical systems associated with GTS. As a primary modulatory neurotransmitter system, dopamine exerts a powerful influence over striatal output by modulating cortico-striatal afferents innervating distinct populations of striato-nigral and striato-pallidal medium-sized spiny neurons. The dynamic spatio- temporal synergy exhibited between tonic and phasic dopaminergic signals emanating from the substania nigra and ventral tegmental area, is a key factor in driving thalamo- cortical output and coordinated behaviour, which is thought to ensue as a result of the Iron 132 integrated output of D1- and D2-recepetor medium spiny neurons [121]. As such, alterations in the mechanisms sustaining the typical spatio-temporal synergy of tonic/phasic dopaminergic signalling may have a profound influence on reinforcement learning and habit formation systems that are governed by striatal neurons that code the serial order of syntactic natural behaviour [337, 339].

Disruptions in varied functional elements sustaining typical tonic/phasic dopaminergic signalling have been demonstrated in both iron deficiency and GTS. Specifically, alterations in the (a) density of D1/D2 receptors, (b) dopamine transporter function and (c) monoamine catabolism have been reported in GTS [31] and also observed in association with iron deficiency [381]. This data implies that iron deficiency holds an important role in abnormal dopaminergic neurotransmission and is supported by the observed reductions in magnetic susceptibility in the substantia nigra [381]. On the other hand, alterations tonic/phasic dopaminergic signalling may also be driven by deficits in the glutamatergic and GABAergic afferent systems regulating nigral dopaminergic output towards the striatum [118, 121]. Along this line, the observed association between striatal magnetic susceptibility and striatal Gln:Glu, indicates that iron might exhibit an influence on GABA-Glu-Gln cycling. This notion is supported by previous work indicating that iron deficiency during gestational and lactational periods in rodents leads to alterations in key enzymes involved in the GABA-Glu-Gln cycle, ultimately leading to reductions in striatal, pallidal and hippocampal GABAergic/glutamatergic neurotransmission [354–358, 360, 361, 382]. As the rapid transamination of tricarboxylic acid (TCA) cycle intermediate alpha-Ketoglutarte predominates the formation of Glu and subsequent GABA, abnormalities in GABA-Glu-Gln cycling may be in part driven by abnormalities in the TCA cycle enzymes aconitase and succinate dehydrogenase, which depend on incorporated iron-sulfur clusters for typical function. The observed reductions in serum ferritin, subcortical magnetic susceptibility, as well as the association between striatal magnetic susceptibility and the Gln:Glu ratio imply the presence of putative abnormalities in the mechanisms sustaining the typical spatio-temporal dynamics of excitatory, inhibitory and modulatory neurochemical signalling.

In summary, disruptions in iron regulatory mechanisms during specific epochs of brain development — which may be driven by a complex relationship between genetic and environmental influences — may lead to pervasive abnormalities in mechanisms regulated by iron-containing enzymes. Abnormalities in mRNA transcription, cytoskeletal regulation, synapse formation, axonal outgrowth, ion-channel gating, iron-sulfur cluster biogenesis mechanisms during specific developmental epochs, may lead to alterations in the spatio- temporal dynamics of excitatory, inhibitory and modulatory subcortical neurochemical signalling, thus driving the manifesting clinical features in GTS. Iron 133

9.5.3 Limitations and future directions

We first note that although iron sequestered in ferritin is considered as the dominant source of the observed susceptibility signal, other elemental sources may contribute to perturbations in the magnetic field. These may include (a) paramgnetic transition elements (copper(II), manganese, zinc) and (b) diamagnetic alkaline earth metals (calcium, magnesium) which may reduce the specificity of magnetic susceptibility to iron effects. Although it is possible that abnormalities in the homeostasis of these elements contribute to GTS pathophysiology — given that important roles they play in neurochemical signalling — it may be plausible to suggest these elements have no major effects on the magnetic susceptibility signal given their relative scarcity in contrast to iron [383]. More- over, although previous work has demonstrated strong relationships between iron and GABA/dopamine metabolism, one limitation of our study is that we made indirect links between iron and GABA/dopamine without their direct measurement. Another limitation of our study is related to the sample size. Although the observed significant reduction in magnetic susceptibility passed FDR multiple comparison correction, we believe that our study warrants replication in a larger independent sample. The assessment of the relationships between (a) spectral measures of Glu, Gln and GABA, (b) PET measures of dopamine, and (c) susceptibility measures of iron presents a promising prospect for the elucidation nature of GTS pathophysiology. Particularly, we believe that the assessment MR measures of iron, Glu, Gln, GABA in developing populations to be most critical. The in-vivo measurement of elemental and neurochemical MR indices may prove crucial in the identification of GTS specific biomarkers, potentially allowing early intervention using pharmacological agents and iron-supplementation therapy. Chapter 10

Investigation of the inﬂuence of aripiprazole on clinical status

This chapter presents a comprehensive analysis of the clinical characteristics of the entire patient study sample at baseline and following treatment with the antipsychotic aripiprazole. The presented work details clinical data related to the objectives of this thesis and reﬂects partial material published in peer review-journal article: Gerasch S., Kanaan A.S., Jakubovski E. and Müller-Vahl K.R. Aripiprazole improves associated comorbid Conditions in addition to Tics in adult Patients with Gilles de la Tourette Syndrome. Frontiers in Neuroscience 2016; 10: 416 [384].

10.1 Abstract

Gilles de la Tourette Syndrome (GTS) is characterized by motor and vocal tics, as well as associated comorbid conditions that include obsessive-compulsive disorder (OCD), attention deficit/hyperactivity disorder (ADHD), depression, and anxiety. Although randomized controlled trials including a large number of patients are lacking, aripiprazole is currently considered as a first choice drug for the treatment of tics. The aim of this study was to further investigate efficacy and safety of aripiprazole in a group of drug- free, adult patients. Specifically, we investigated the influence of aripiprazole on tic severity, comorbidities, premonitory urges, and quality of life (QOL). Moreover, we were interested in the factors that influence a patient’s decision in electing for-or against- pharmacological treatment. Utilizing a prospective, uncontrolled open-label longitudinal study design, 44 patients were recruited and assessed via a number of rating scales to assess tic severity, premonitory urges, comorbidities, and QOL at baseline and during treatment with aripiprazole. Eighteen out of 44 patients elected for undergoing treatment for their tics with aripiprazole and completed follow-up assessments after 4-6 weeks. Our 134 Aripiprazole 135 major findings were (a) aripiprazole resulted in significant reduction of tics, but did not affect premonitory urges; (b) aripiprazole significantly improved OCD and showed a trend toward improvement of other comorbidities including depression, anxiety, and ADHD; (c) neither severity of tics, nor premonitory urges or QOL influenced patients’ decisions for or against treatment of tics with aripiprazole; instead patients with comorbid OCD tended to decide in favor of, while patients with comorbid ADHD tended to decide against tic treatment; (d) most frequently reported adverse effects were sleeping problems; (e) the patients QOL was mostly impaired by comorbid depression. Our results suggest that aripiprazole may improve associated comorbid conditions in addition to tics in patients with GTS. It can be hypothesized that these beneficial effects are related to aripiprazole’s adaptive pharmacological profile, which exhibits an influence on the dopaminergic as well as a number of other neurotransmitter systems. For the first time, our data provide evidence that a patients’ decision in undergoing treatment is also influenced by factors other than tic severity and QOL.

10.2 Introduction

Although different therapeutic strategies are currently being used to treat patients with GTS (e.g. behavioral therapy, pharmacotherapy, and surgical interventions), no currently used treatment is able to target the multiple symptoms associated with GTS. With respect to pharmacological treatment, recommendations are currently based only on a small number of controlled or uncontrolled studies, as large sample and longitudinal randomized controlled trials (RCT) have not been conducted yet [6]. Nonetheless, there is general agreement that antipsychotic agents targeting the dopaminergic system are most effective in the treatment of tics. Currently, haloperidol remains as the only antipsychotic that is formally approved for the treatment of GTS in most European countries. However, haloperidol is no longer being prescribed in Germany and most European countries, since it is associated with the emergence of significant adverse effects (e.g. extrapyramidal symptoms, sedation and weight gain) [385]. Owing to their more favorable therapeutic profile, second generation antipsychotics (e.g. risperidone, aripiprazole, sulpiride, and tiapride) have become the most common off-label substances used for the treatment of tics [6]. Specifically, aripiprazole has become the favorite neuroleptic in many centers for treating tics given its remarkable efficacy and favorable side effect profile [91].

Currently, aripiprazole is approved by the American Food and Drug Administration (FDA) and the European Medicines Agency (EMEA) for the treatment of schizophrenia, bipolar I disorder, and manic episodes [386]. Although large RCTs are lacking, aripirazole is used for the treatment of GTS symptoms since: (a) it is better tolerated Aripiprazole 136 and causes less side effects relative to other antipsychotic drugs [170], and (b) is effective in severely affected and otherwise treatment-refractory patients. Therefore, many clini- cians classify aripiprazole as first choice medication in the treatment of tics [6]. To date, varied groups have investigated the efficacy and safety of aripiprazole in the treatment of tics in GTS. However, the majority of these studies focused on child and adolescent patients groups. In one of the largest RCTs, Ghanizadeh and Haghighi [387] investigated the effects aripiprazole on the treatment of tics in 60 children and adolescents with tic disorders. Aripiprazole resulted in a significant reduction of both motor and vocal tics (Yale Global Tic Severity Scale — YGTSS), as well as a significant improvement of quality of life (QOL). In general, aripiprazole was well tolerated, although some adverse effects were reported including drowsiness (25.8%) and increased appetite (25.8%). A systematic review of published literature has revealed that aripiprazole is effective in the treatment of tics in both adults and children with GTS with an adverse effect profile that seems to be safer than in other antipsychotic drugs [388].

Interestingly, studies exploring the clinical effects of aripiprazole on the treatment of psychiatric comorbidities in GTS are few in number. Investigating the influence of aripiprazole on psychiatric comorbidities in GTS is particularly interesting given that it is, in essence, a functionally selective modulator that is able to stabilize multiple neurotransmitters systems, which may exhibit varying abnormalities within the different sub-classifications of GTS [171]. Thus, it may be plausible to suggest that aripiprazole may amerliorate both tics and psychiatric comorbid features including OCD, ADHD, depression, anxiety and rage attacks. Along this line, Wenzel et al. [170] reported that 5 of 100 patients exhibited improvements in depression, anxiety, and aggression following treatment with aripiprazole. Other works investigating the effects of aripiprazole in OCD have demonstrated that it can successfully ameliorate obsessive-compulsive symptoms [389, 390]. For ADHD, results were variable as one study showed a strong influence on reducing attention deficit scores [167], while another study reported a moderate effect on a child and adolescent population [391]. In contrast, Frölich [392] investigated the influence of aripiprazole on OCD and ADHD in GTS patients and reported that aripiprazole did not significantly influence psychiatric comorbidities. In the two published controlled trials investigating the effect of aripiprazole on GTS [387, 393], no data on treatment effects with respect to psychiatric symptoms were reported. Thus, given the lack of published data, the effect of aripiprazole on the treatment of GTS-related comorbidities is not currently clear.

As such, the aim of this study was to investigate the eﬀect of aripiprazole on tics and comorbid conditions in a large group of unmedicated adult patients with GTS. Our primary goal was to investigate the eﬀect of aripiprazole on tics, quality of life, premonitory urges (PU) in addition to psychiatric comorbidities (depression, OCD, anxiety, and Aripiprazole 137

ADHD). As treatment with aripiprazole was voluntary for the sample, we were able to compare clinical data between the sub-samples that elected for and against aripiprazole administration. Our hypothesis was that aripiprazole improves both tics and behavioral problems in adult and otherwise untreated patients with GTS, and that those patients tend to choose treatment with aripiprazole suﬀer from more severe tics and lower quality- of-life.

10.3 Methods

10.3.1 Population Sampling

Forty-four right-handed adult patients with GTS (18-65 years) were recruited from the outpatient Clinic of Psychiatry, Social Psychiatry and Psychotherapy at Hannover Med- ical School according to DSM-5 criteria [19]. Patients using any psychoactive substances underwent a four-week washout period before participation. Exclusion criteria were (a) ages below 18 or above 65 years, (b) inability to lie still in Magnet Resonance Imaging (MRI) system, (c) known MRI contraindications such as metals, tattoos, and claustrophobia as well as pregnancy and breastfeeding. The study was approved by the ethics committees of Hannover Medical School and the University of Leipzig. All participants gave written informed consent before participating the study.

10.3.2 Clinical Assessment

All patients underwent a neuropsychiatric interview and a comprehensive clinical assessment battery including measurements of tics, premonitory urges; QOL, and psychiatric comorbidities (OCD, ADHD, depression, anxiety, and autism) as described in Chapter 5.

10.3.3 Statistical Analysis

Statistical analysis was performed in the Statistical Package for Social Sciences (SPSS, Version 20). Descriptive statistics were computed for all measurements. Treatment effects in terms of pre/post-comparison of the tic severity and comorbidity scores were carried out using the Wilcoxon-Mann-Whitney-Test for paired samples. The baseline characteristics of patients electing for- and against-treatment with aripiprazole were compared using the Wilcoxon-Mann- Whitney-Test for independent samples. All statistical tests were two-sided and the alpha value was set at 0.05. No adjustment for Aripiprazole 138 multiple comparisons was performed due to the exploratory nature of the analysis. To inspect patients’ compliance to drug intake, aripiprazole serum levels were correlated with administered oral dosage via Pearson correlation.

10.4 Results

10.4.1 Patient characteristics at baseline

Clinical data was collected from a total of 44 patients (9 Females, 39.4±12.2 years). The sample exhibited moderate tic severity (YGTSS-TTS =22.2±8.5, scale range = 3–39; RVTRS=9.6±5.0, scale range = 0–18) and a mean comorbidity score of 1.36 (range 0-4) (Tables 10.1 and 10.2).

Table 10.1: Subgroup classiﬁcation of the whole study sample based on comorbidities

Baseline Follow-up Overall Untreated Treated Treated N=15 N=26 N=18 N=18 GTS only 15 9 6 8 GTS+comorbidities 29 17 12 10 GTS+OCD 8 2 6 2 GTS+ADHD 8 8 1 1 GTS+OCD+ADHD 7 4 3 2 Others 6 3 2 5 GTS=Gilles de la Tourette Syndrome; OCD=Obsessive-Compulsive Disorder; ADHD=Attention- Deficit/Hyperactivity Disorder; N=Number of cases; Others=Patients with comorbidities that do not fulfill criteria for one of the other defined subgroups. The others group included N = 6 for depression, N = 6 for anxiety, N = for The comorbidities include OCD, ADHD, depression, and anxiety depression+anxiety Aripiprazole 139

Table 10.2: Clinical characteristics of the whole study sample at baseline and following treatment with aripirazole

Treated (N = 26 Treated (N=18 Baseline Baseline Follow-up Difference YGTSS-TTS 22.7 ±7.9 19.2 ±7.5 -3.5* YGTSS-MT 12.8 ±4.5 13.7 ±3.4 11.8 ±3.4 -1.9* Tics YGTSS-VT 9.4 ±6.2 9.0 ±5.8 7.4 ±5.4 -1.6* YGTSS-GS 42.5 ±17.8 50.9 ±17.2 35.9 ±17.4 -15.0** RVTRS 8.9 ±4.7 10.4 ±5.5 8.0 ±4.2 -2.4* PU PUTS 21.8 ±5.9 19.7 ±6.1 19.8 ±5.5 0.1 Clinical diagnosis N=6 N=9 N=5 -4* M.I.N.I. OCD current N=6 N=9 N=7 -2 Y-BOCS 3.7 ±6.9 5.3 ±6.9 4.2 ±5.5 -1.1 Obsessions 1.2 ±3.6 1.7 ±3.5 1.3 ±3.5 -0.4 Compulsions 2.5 ±4.4 3.7 ±4.9 3.0 ±4.9 -0.7 OCI-R 16.6 ±15.0 15.5 ±11.6 14.6 ±13.8 -0.9 OCD Washing 0.9 ±1.9 0.8 ±1.20 1.6 ±2.85 0.8 Obsessing 3.2 ±3.3 2.7 ±2.6 2.6 ±2.5 -0.1 Hoarding 3.2 ±2.7 2.1 ±2.8 2.1 ±2.9 - Ordering 3.7 ±3.2 4.3 ±3.3 3.4 ±3.1 -0.9 Mental neutralization 1.4 ±2.1 2.0 ±2.5 2.2 ±2.9 0.2 Checking 3.7 ±3.4 3.6 ±3.3 2.8 ±3.3 -0.8 Clinical diagnosis N=8 N=6 N=4 -2 M.I.N.I. MD current N=5 N=3 N=1 -2 Depression BDI-II 12.8 ±12.9 13.7 ±10.8 10.6 ±8.20 -3.1 MADRS 8.7 ±8.27 7.2 ±5.6 8.6 ±4.00 1.4 Clinical diagnosis N=9 N=6 N=4 -2 M.I.N.I. panic current N=3 N=1 N=1 — agoraphobia N=4 N=3 N=3 — Anxiety social phobia N=1 N=0 N=0 — GAD N=2 N=1 N=1 — BAI 13.8 ±13.8 9.4 ±8.7 8.6 ±7.0 -0.8 Autism AQ 18.2 ±8.4 19.3 ±7.7 19.7 ±7.2 0.4 Clinical diagnosis N=12 N=4 N=3 -1 CAARS Inattention 50.4 ±13.2 50.6 ±10.4 47.9 ±7.3 -2.7 Hyperactivity- 51.6 ±9.9 51.9 ±11.3 51.4 ±9.9 -0.5 Restlessness Impulsivity 50.8 ±10.2 52.5 ±11.9 50.5 ±11.5 -2 Self concept 48.8 ±8.9 47.6 ±8.9 48.5 ±8.30 0.9 ADHD Inattention 53.1 ±15.6 52.8 ±15.7 54.7 ±13.9 1.9 Hyperactive-Impulsive 51.4 ±13.3 51.1 ±14.4 51.5 ±9.4 0.4 ADHD total 53.1 ±14.8 52.3 ±14.9 54.2 ±12.3 1.9 ADHD index 53.1 ±11.2 53.8 ±11.5 53.8 ±9.6 — WURS-k 35.3 ±15.5 29.2 ±12.8 — — DSM-IV Attention 4.3 ±2.9 4.3 ±3.2 — — Hyperactivity 3.5 ±3.1 2.7 ±2.4 — — GTS-QoL 33.4 ±23.8 26.7 ±16.1 24.5 ±17.1 -2.2 QoL GTS-QoL-VAS 61.5 ±27.3 60.6 ±20.5 67.1 ±19.4 6.5 YGTSS, Yale Global Tic Severity Scale; TTS, Total Tic Score; MT, Motor Tic Score; VT, Vocal Tic Score; GS, Global Score; MRVS, Modified Rush Video-Based Tic Scale; PU, Premonitory Urge; PUTS, Premonitory Urge for Tics Scale; OCD, Obsessive-Compulsive Disorder; M.I.N.I., International Neuropsychiatric Interview; Y-BOCS, Yale-Brown Obsessive Compulsive Scale; OCI-R, Obsessive-Compulsive Inventory Revised; MD, Major Depression; BDI, Beck Depression In- ventory; MADRS, Montgomery Asberg Depression; GAD, Generalized Anxiety Disorder; BAI, Beck Anxiety Inventory, AQ, Autism-Spectrum-Quotient; ADHD, Attention-Deficit/Hyperactivity Disorder; CAARS, Conners Adult ADHD Rating Scale (for patients treated with aripiprazole, CAARS scores are given for N = 15/17 due to missing data), WURS-k, Wender Utah Rating Scale; DSM, Diagnostic and Statistical Manual; QoL, Quality of Life (the higher the sum, the lower the QoL); VAS, Visual Analogue Scale (the higher the sum, the higher the satisfaction). Clinical diagnosis, Diagnoses for comorbidities as defined above; N, Number of cases. *p < 0.05, **p < 0.01. Aripiprazole 140

10.4.2 Patient characteristics during treatment with aripiprazole

Of the 44 recruited patients, 18 patients elected for undergoing treatment of their tics with aripiprazole. At follow-up, mean dosage of aripiprazole was 12.2 mg (median=10 mg, range, 2.5-30 mg). All patients reported that they had reached their individual target dosage at the follow-up visit.

10.4.2.1 Tics and premonitory urges

Treatment with aripiprazole resulted in a significant tic reduction according to YGTSS (YGTSS-TTS: difference: -3.5, P =0.027, YGTSS-MT: difference = -1.9, P = 0.037, YGTSS-VT: difference= -1.6, P = 0.045, YGTSS-GS: difference= -15.0, P=0.002) and RVTRS (difference = -2.4, P = 0.022) (Figure 10.1 and Table 10.2). On the other hand, aripiprazole did not result in a significant improvement of premonitory urges as assessed by PUTS (P = 0.917).

Figure 10.1: Distribution of YGTSS-TTS scores at baseline and followup in the diﬀerent subgroups .

10.4.2.2 Psychiatric comorbidities and quality of life

Related to associated psychiatric comorbidities, aripiprazole led a significant impact on the patients’ clinical characteristics. Specifically, the classification of patients in the Aripiprazole 141

GTS-plus, GTS-OCD, GTS-OCD-ADHD subgroups decreased by 2, 4 and 1 respectively. However, the GTS-ADHD subgroup remained unchanged (Table10.1 and Figure 10.2). Focusing on the different comorbidities separately, improvements were observed in OCD, depression and anxiety diagnoses following treatment (Table 10.2). Although changes in OCD scores were not significant (Y-BOCS (P = 0.445), M.I.N.I. OCD current (P = 0.317), OCI-R (P = 0.585)), OCD diagnosis decreased by approximately half (50% to 27.8%) and no patients developed OCD symptoms following treatment. Similarly, the number of patients diagnosed with depression and anxiety decreased by 4 patients (33.33% to 22.2%) and 2 patients, respectively, although comparison of the clinical parameters off- and on-treatment were not significant. Notably, none of the patients developed OCD, depression or anxiety disorder during treatment with aripiprazole. On the other hand, the GTS-ADHD subgroup only decreased by one patient (22.2% to 16.7%). With respect to autism, two treated patients exhibited autistic traits at baseline, only one of which was above the AQ cutoff following treatment. As for QOL and satisfaction- with-life, treatment with aripiprazole did not lead to any significant improvements as assessed by both GTS-QOL (difference: -2.2, P =0.760) and GTS-QOL-VAS (difference: +6.5, P =0.106).

Figure 10.2: Psychiatric comorbidity subclassiﬁcattion at baseline and following treatment with aripiprazole. Aripiprazole 142

Figure 10.3: Prevalence of comorbidities in patients that elected for- and against-treatment with aripiprazole.

10.4.3 Comparison of clinical characteristics of patients electing for- and against-treatment

Of 44 patients included in this study, 18 patients elected for undergoing treatment of their tics with aripiprazole (4 Females, age=38.5±13.7 years) while 26 patients (5 Females, age=40±11.4 years) elected for no treatment. A number of reasons factored into the patient’s decision including: (a) no interest in undergoing medical treatment; (b) lack of disabling impairments by their tics; (c) worries about possible adverse events related to aripiprazole. Since enrollment into the treatment study was voluntary, based solely on the patient’s decision and not on tic severity or advice of the treating physician, we compared the clinical characteristics between both groups at baseline in order to identify factors that may influence their decision for undergoing medical treatment for their tics. Interestingly, tic severity (YGTSS and RVTRS), premonitory urges (PUTS), QOL (GTS-QOL and GTS-QOL-VAS) were not significantly different between both groups (Figure 10.3). On the other hand, we observed that the diagnosis of comorbid OCD at baseline tended to be significantly more common in patients who elected for- than against- treatment (50% to 23.1%, P =0.067). With respect to comorbid ADHD, the opposite case was observed as the diagnosis of ADHD was more common in patients that elected against treatment (Figure 10.3). Aripiprazole 143

10.4.4 Adverse Eﬀects and continuation of treatment

Twelve out of 18 patients (66.7%) reported AEs (Figure 10.4). However, the AEs experienced by the patients were not severe, as no medical intervention was necessary for any patient. After the end of the study, 14 patients (77.77%) decided to continue treatment with aripiprazole. Four patients stopped medication, among them N=3 due to AEs (drowsiness, internal unrest (akathisia), sleep disturbance, restlessness of legs) and N=1 due to no tic improvement.

Figure 10.4: Adverse eﬀects reported by the patients following the administration of aripiprazole.

10.4.5 Serum levels of aripiprazole

Serum levels of aripiprazole were measured in 14/18 patients ranging from 7.5-269 µg/L (125.3 +79.8) (therapeutic range: 150-250 µg/l). Serum levels correlated signiﬁcantly with administered oral doses (2.5-30 mg/day) (R=0.7, P =0.003) indicating successful adherence to treatment (Figure 10.5).

10.5 Discussion

In this work, we investigated the eﬀects of aripiprazole on tic severity and psychiatric comorbidities in an open-label clinical trial of adult patients with GTS. Our results Aripiprazole 144

Figure 10.5: Prevalence of comorbidities in patients that elected for- and against-treatment with aripiprazole. indicated that: (a) aripiprazole is an effective and safe treatment for the treatment of tics as well as for comorbid OCD and possibly other comorbidities such as ADHD, depression and anxiety; (b) aripiprazole has no influence on premonitory urges; (c) patients with comorbid OCD are more likely to vote for medical treatment to reduce their of their tics when compared to those with comorbid ADHD; and (d) tic severity, premonitory urges; QOL do not influence the patients’ decision making process.

10.5.1 Eﬃcacy of aripiprazole on tics and premonitory urges

Our findings are in line with preliminary data demonstrating that aripiprazole results in a significant improvement of tics (according to YGTSS-TTS and RVTRS) in the majority of adult patients with GTS [168, 170, 388, 394–397]. On the other hand, aripiprazole did not result in a significant improvement of PU. To the best of our knowledge, this is the first study, investigating the effect of aripiprazole on premonitory urges in adult patients with GTS. Thus, from our results it is suggested that effective treatment of tics is possible without improvement of PU. These findings support recent clinical studies Aripiprazole 145 suggesting that premonitory urges is not as closely related to tic severity as previously assumed [398, 399].

10.5.2 GTS subgroup classiﬁcation

For a thorough clinical characterization of our sample, a number of clinical assessment tools were utilized to diagnose and quantify each of the commonly associated psychiatric comorbidities. Specifically, three rating scales were used for each of obsessive-compulsive, inattention/hyperactivity and depressive symptoms, while two rating scales were used for anxiety. Nevertheless, it is well known that the use of different assessments — although developed for the measurement of the same symptoms — may reflect different sides of the same disorder and therefore, may lead to inconsistent findings. In addition, it is generally accepted that patients’ self-perception may differ from professional evaluation [400] resulting in discrepant results when using self-ratings compared to examiner rating scales [401]. For example, the subtle differences of commonly used depression rating scales were pointed out by Uher et al. [239], who showed that HAMD-17, MADRS and BDI- II, while all being valid and reliable scales, reflect “internally consistent but mutually distinct estimates of depression severity”. While the MADRS is closer to the core of depression - observable from the outside - the BDI represents a “cognitive” dimension - which is more of an internal experience. Uher et al. [239], therefore, recommended to use these scales in a complementary fashion. Comparably, we found contrary results when using the BDI-II as a self-rating compared to the MADRS as an expert rating: while treatment with aripiprazole resulted in a decrease in mean BDI-II values, mean MADRS scores increased. A closer investigation showed that this obvious inconsistency was due to the high incidence of “sleep disturbances” (44.4%) and “internal unrest” (16.7%), which were the most frequently reported AEs during treatment with aripiprazole. While having little impact on the BDI-II score, the presence of these symptoms had a considerably high impact on the MADRS scores leading erroneously to high scores for depression.

10.5.3 Eﬃcacy of aripiprazole on associated comorbid conditions

According to defined diagnoses, treatment with aripiprazole resulted in a significant improvement of OCD, although respective assessments for OCD demonstrated no significant changes at follow-up compared to baseline. Due to the relatively small number of patients, results in other psychiatric comorbidities did not reach statistical significance. However, treatment with aripiprazole, resulted in the remission of comorbid depression in 4 of 6 patients, of comorbid anxiety in 4 of 6 patients, and of comorbid ADHD in 1 of 4 patients. Accordingly, the mean comorbidity score decreased from 1.38 at baseline to Aripiprazole 146

1.16 at follow-up. Furthermore, pathologically autistic traits remitted in 1 of 2 patients and absolute scores (according to AQ) improved in 8 patients.

Our findings corroborate available case reports in patients with GTS reporting about beneficial effects of aripiprazole in the treatment of depression [167, 170], OCD [167, 389, 390], anxiety, self-injurious behavior [170], inattention [167], and ADHD [391], but in contrast to preliminary results by Frölich [392] who found no improvement of ADHD and OCD in children with GTS. However, in none of these studies the effect of aripiprazole has been investigated specifically for the treatment of psychiatric comorbidities, comorbidities were not assessed by using a variety of self- and examiner-ratings, and, at least in part, data were collected retrospectively from patient records.

Furthermore, our findings in GTS are completely in line with data in patients with pure OCD, where aripiprazole has been found to be effective not only in the treatment of uncomplicated OCD [402], but even in patients resistant to treatment with selective serotonin reuptake inhibitors [403–406]. Aripiprazole has also been found helpful in the treatment of depression [407] and anxiety disorders [408, 409]. It is even one of the most often prescribed medication in patients with anxiety and mood disorders [410]. In ADHD — although often prescribed [410] — its efficacy has not been demonstrated [411].

The effective influence of aripiprazole on a number of psychiatric conditions has been suggested to be a result of its unique pharmacological profile. Specifically, aripiprazole is a functionally selective drug that exhibits an adaptive pharmacological profile that is dependent on the local levels of the endogenous ligands. Aripiprazole is a partial dopamine D2 agonist, a partial serotonin 5-HT1A agonist, and a 5-HT2A antagonist. Apart from its recognized influence on the dopaminergic and serotonergic systems, aripiprazole has also been shown to modulate the glutamatergic and GABAergic neurotransmitter systems [171]. Therefore, it can be speculated that beneficial effects of aripiprazole on OCD, depression and anxiety in patients with GTS may be the result of its ability to selectively and adaptively stabilize multiple neurotransmitter systems.

10.5.4 Inﬂuence of aripiprazole on quality of life

Comparable to previous reports by Müller-Vahl et al. [412] and Jalenques et al. [413], we found that in adult patients with GTS both QOL and satisfaction-with-life (as assessed by GTS-QOL and GTS-QOL-VAS) are mainly impaired by depression. This was the case at baseline and also during treatment with aripiprazole. Although aripiprazole resulted in a significant improvement of tics, we only found a trend towards an improvement in patients’ QOL. This is in line with the finding that depression influences patients’ QOL more than the tics. Interestingly, we found a significant correlation between premonitory Aripiprazole 147 urges and QOL at baseline. Assuming that premonitory urges is a kind of an OCB as suggested recently [414] and against the background that it is well-known that OCD significantly impairs QOL in adult patients with GTS [412], this correlation can possibly be explained by the negative influence of OCD on patients’ QOL. The complex interplay between tics, comorbidities, and QOL is also expressed in changes in the YGTSS-GS: this “global score” of the YGTSS is a measurement for both tic severity and overall impairment. Completely in line with all above mentioned results, after treatment with aripiprazole we found a much greater reduction of the YGTSS-GS (p=0.002) compared to the tic score of the YGTSS (YGTSS-TTS, p=0.027).

10.5.5 Adverse eﬀects of aripiprazole

Although AEs were reported by a substantial number of patients (66.7%), most AEs were mild and/or tolerable corroborating recent data that in most adult patients with GTS aripiprazole is well tolerated [170, 396]. No serious AEs occurred. 3 of 18 patients (16.7%) decided to stop treatment with aripiprazole due to AEs such as drowsiness, internal unrest, sleep disturbance, and restlessness of legs. Comparable to our data, Wenzel et al. [170] also reported about the occurrence of AEs in nearly 2/3 (59%) of their patients. However, in contrast to our results, they found drowsiness (20%) to be the most common side effect, while sleep disturbances were quite rare (9%). In this study, sleep disturbances (44.4%) followed by internal unrest (16.7%) were the most often reported AEs, while drowsiness occurred in only 11.1% of our patients. This difference might be explained by different study designs and different treatment durations (4-6 weeks in our study vs. 1-60 months in Wenzel et al. [170]. Hence, it can be assumed that in the context of a longer-term treatment, patients may tolerate drowsiness rather than internal unrest and sleep disturbances. Nonetheless, our data further support the clinical practice to start treatment with aripiprazole once daily in the morning, and to postpone intake to the evening, if significant drowsiness occurs.

10.5.6 Decision factors for treatment with aripiprazole

Our study design provided us with the possibility of investigating the factors that influence patient’s decision in electing for- or against-treatment. This choice was solely based on each patient’s own preference. We found no differences between both groups with respect to age, gender, and comorbidity score. Most interestingly, neither tic severity (according to YGTSS and RVTRS), nor premonitory urges (according to PUTS) nor QOL (as assessed by GTS-QOL and GTS-QOL-VAS) was significantly different between both groups. However, we found a trend with respect to comorbid OCD and ADHD. Aripiprazole 148

While OCD was more common in those patients who decided for treatment with aripiprazole, ADHD was more common in those who decided against. Thus, although not reaching the significance threshold, our data seems to indicate that there are aspects influencing patients’ decisions for or against medical treatment for tics beyond tic severity. Since comorbid OCD has a strong negative impact on patients’ quality of life, it can be speculated that this might be a driving force that also influences patients’ treatment decision in favor of treatment. However, it can also be possible that patients with comorbid OCD differ in their assessment of impairment caused by their tics as compared to patients without OCD, possibly due to a larger extent of ruminating and worrying caused by their compulsions. Finally, it can be hypothesized that patients with comorbid ADHD are less impaired by their tics and therefore tend to decide against treatment. This is particularly noteworthy, since it has been demonstrated that patients with comorbid ADHD are less able to suppress their tics [414] and effective tic suppression has a positive impact on patients’ QOL [415].

10.5.7 Characteristics of the Sample and serum levels of aripiprazole

With respect to tic severity, comorbidities, and distribution of gender, in this open-label study a representative clinic sample of adult patients with GTS was included. Although, all patients participating in this study, in addition, participated in an MRI study - and therefore patients also had to fulfill inclusion criteria for that study - our group of patients was characterized by moderate tics (mean tic severity=22.2 according to YGTSS-TTS, N=44). Usually, a threshold of YGTSS-TTS > 14 indicates clinically significant tics that justify treatment [174, 416]. In contrast to most other studies investigating the efficacy of aripiprazole in patients with GTS, we included only adults > age of 18 years. It is noteworthy that all patients who received treatment with aripiprazole were otherwise free of any other psychoactive drug for at least 4 weeks before entering the study. Thus, interactions with other psychoactive substances or augmentation effects are not of concern. Measurements of serum levels of aripiprazole demonstrated patient adherence. For the first time, we were able to show positive correlation between oral dosage and serum levels of aripiprazole in this group of patients.

10.5.8 Limitations

A number of limitations of this study that are worth mentioning are: (a) no placebo patient or control group was included; (b) the study sample exhibited varying degrees of severity and history of drug use; (c) the number of patients undergoing treatment was relatively low; (d) given that treatment duration was relatively short (only 4-6 weeks) Aripiprazole 149 and that aripiprazole’s long half-life of approximately 72 hours, we cannot rule out that aripiprazole levels were still increasing at the follow-up assessment; (d) we cannot exclude that patients decided for participation in the study at baseline and/or follow-up due to monetary compensation, travel restrictions or reluctance to participate in further MRI investigation; (e) we cannot exclude that other factors than comorbid OCD and ADHD may have inﬂuenced a patients decisions in undergoing treatment.

10.6 Conclusions

To the best of our knowledge, this is one of the largest prospective open-label clinical trial examining untreated adult patients with GTS patients at baseline and during 4-6 weeks monotherapy with aripiprazole. Our results indicated that aripiprazole leads to significant reductions of tics in adult patients with GTS without affecting premonitory urges. Related to non-motor symptoms, the sample exhibited reductions in some of the associated conditions. OCD diagnosis decreased by approximately 50% and no patients developed OCD symptoms following treatment. Similarly, the number of patients diagnosed with depression and anxiety decreased by about 10%. On the other hand, ADHD classification remained unchanged in the patients that underwent treatment. While patients with comorbid ADHD tended to elect against undergoing treatment, it is interesting to report that patients with comorbid OCD tended to elect for undergoing treatment. In general, no patients reported any severe side effects. Our results indicated that aripiprazole is a relatively safe and reliable treatment. Part V

CONCLUSIONS

150 Chapter 11

Key ﬁndings and signiﬁcance

This chapter summarizes the key ﬁndings of methodological and pathophysiological investigations conducted for the purposes of this thesis. Detailed discussion of the results of each experiment is presented in Chapters 6-10.

Despite the fact that the range of treatments used to manage the symptomatology of GTS has been expanding over the last decade, current treatment strategies are often unsatisfactory and many are associated with severe side eﬀects. As such, there is currently an urgent need in the ﬁeld to identify new therapeutic approaches that can ameliorate the manifesting motor and non-motor clinical features. This goal has been hampered by the relative dearth of investigations on the neurobiology of GTS and the subsequent lack of a comprehensive pathophysiological model.

This thesis describes a series of experiments that were conducted to further elucidate the nature of GTS pathophysiology using multi-parametric, quantitative neuroimaging approaches at multiple scales. A well characterized and relatively large sample of patients was recruited to partake in investigations at the elemental and neurochemical levels. To achieve these aims, two methodological investigations were conducted to obtain quantitative neurochemical and elemental measurements of suﬃcient precision to identify rather subtle changes. A longitudinal study design was additionally employed to investigate the characteristics of the sample at baseline and during treatment with the commonly used antipsychotic aripiprazole. The main results of this work are summarized in the sections below.

151 Key ﬁndings and signiﬁcance 152

11.1 Pathological glutamatergic neurotransmission in GTS

Given the spatio-temporal and metabolic interdependence exhibited by neurotransmitter glutamte (Glu) with the putatively abnormal dopamine and γ-Aminobutyric acid (GABA), respectively, we utilized in vivo 1H Magnetic Resonance spectroscopy (1H- MRS) to investigate the role of glutamatergic signalling in GTS at baseline and during treatment with aripiprazole.

To ensure the reproducibility of our neurochemical measurements, we ﬁrst examined the test-retest reliability of absolute metabolite quantitation with the consideration of the variabilities of within voxel tissue proportions across subjects and sessions (Chapter 6). We observed that sophisticated and commonly used segmentation algorithms yield diﬀer- ent regional estimates of within voxel tissue fractions in a sample of healthy controls. One tissue segmentation algorithm (SPM) yielded a high consistency in the estimation of between session tissue proportions relative to others. Considering the algorithm’s estimates of within voxel tissue proportions for the correction of absolute metabolite quantities, we observed a decrease in the variance of absolute estimates of key metabolites that include Glu and glutamine (Gln). This methodological investigation led to the conclusion that the careful consideration and utility of reliable segmentation algorithms is essential for the interrogation of changes in relevant metabolite concentrations in psychiatric and neurological disorders such as GTS. In conclusion, this methodological investigation better informed our examination of the role of Glu in GTS pathophysiology.

Given the diﬃculty of acquiring high quality spectra in subcortical regions, especially in a patient population characterized by movement, we employed a careful acquisition and processing scheme that incorporated (a) an automated voxel (re-)localization technique; (b) the removal of motion corrupted outlier signals; (c) frequency and phase drift correction in the time domain; (d) absolute metabolite quantitation with the consideration of within voxel compartmentation; and (e) a semi-automated quality assessment protocol (Chapter 8). By utilizing these techniques, we observed a high reliability of voxel relocalization and metabolite quantitation across sessions in the healthy control sample.

Comparing the healthy control group with the GTS sample at baseline, we observed significant reductions in striatal concentrations of Gln, Glu + Gln (Glx) and the Gln:Glu ratio, and thalamic concentrations of Glx in GTS. ON-treatment patients exhibited no significant metabolite differences when compared to controls but significant increases in striatal Glu and Glx, and trends for increases in striatal Gln and thalamic Glx compared to baseline measurements. Multiple regression analysis revealed a significant negative correlation between (a) striatal Gln and actual tic severity; and (b) thalamic Glu and premonitory urges. Key findings and significance 153

Given the interdependent metabolic relationship exhibited between Glu and GABA via the non-neuroactive intermediate Gln, our results indicate that patients with GTS exhibit an abnormality in the ﬂux of metabolites in the GABA-Glu-Gln cycle. In general, this implies the presence of perturbations in subcortical astrocytic-neuronal coupling systems that maintain the subtle balance between excitatory and inhibitory neurotransmission. Such perturbations may ultimately lead to spatially focalized alterations in excitatory, inhibitory and modulatory neurotransmitter ratios in functionally distinct striatal subdivisions, which would have a profound inﬂuence on the neuroplastic mechanisms involved in reinforcement learning and habit formation systems that are governed by striatal neurons that code the serial order of syntactic natural behaviour.

11.2 Subcortical iron reductions associated with glutamatergic neurotransmission in GTS

In our next line of questioning, we focused on the element iron in view of the crucial role it exhibits in varied biochemical mechanisms sustaining developmental processes and neurochemical pathways. Based on the results achieved in the 1H-MRS study and preliminary work indicating reductions in the iron sequestering protein ferritin in both children and adults, we postulated that patients with GTS exhibit abnormalities in iron metabolism, which may inﬂuence mechanisms of subcortical neurotransmission. With the recent establishment of Quantitative Susceptibility Mapping (QSM) as a technique that can serve as a surrogate measure of iron concentrations in deep grey matter nuclei, we employed QSM to estimate iron levels in basal ganglia, brainstem and cerebellar nuclei.

In general, QSM images are reconstructed from tissue phase maps that are typically generated by combining the signals recorded by multiple-channel coil element. However, the coil-combination of multiple-channel data is often degraded and may contain pole artifacts as exhibited by the widely used Adaptive Combine technique. Therefore, to faithfully interrogate subtle changes in iron content in the deep brain, we first investigated the difference in susceptibility values achieved via the the adaptive coil combination technique and a novel method (EPIRiT-SVD) that does not produce coil artifacts. Inter- estingly, difference images between the two algorithms demonstrated the presence of the pole artifacts along subcortical regions implicated in GTS pathophysiology (e.g. striatum). Considering susceptibility differences between various regions of interest including the basal ganglia, brainstem and cerebellar nuclei as well cortical regions, significant differences were observed for the caudate, red nucleus, and the cingulum between the two methods. Notably, the ESPIRiT-SVD coil combination method consistently yielded Key findings and significance 154 lower standard deviations. In conclusion, this methodological investigation indicated that pole artifacts significantly corrupt QSM data. As such, the utility of coil combination methods without pole artifacts are necessary for accurate measurement and comparison of susceptibility values which serve as a surrogate index for iron levels in deep grey matter nuclei.

By utilizing the ESPIRiT-SVD coil combination algorithm, we were able to achieve tissue phase maps without any pole artifacts for QSM reconstruction in the healthy control and GTS samples. We focused on estimating susceptibility values in subcortical nuclei that are implicated in GTS pathophysiology and also exhibit high concentrations of iron and the neurotransmitters dopamine, Glu and GABA. Our results indicated that GTS patients exhibit significant reductions of magnetic susceptibility in basal ganglia, brainstem and cerebellar nuclei. These reductions were specific to the striatum, substantia nigra, subthalamic nucleus and the red nucleus and were mirrored by decreases in serum ferritin levels. Serum ferritin levels were positively correlated with susceptibility values in basal ganglia and brainstem nuclei. These correlations were significant for the control, GTS, as well the entire sample. Using the combination of the QSM and 1H-MRS data, we found significant correlations between striatal – as well as total subcortical – magnetic susceptibility and striatal Gln:Glu. Put together, these results suggest that patients with GTS exhibit disturbances in iron regulatory systems in subcortical regions that typically exhibit the sharpest concentrations of iron and the densest proportions of dopamine, GABA and Glu. The correlation between striatal susceptibility and the Gln:Glu ratio provides support for an in-vivo relationship between iron metbaolism and the Glu:Gln cycle. These abnormalities may thus exhibit an influence on mechanisms of excitatory and inhibitory neurotransmission. As the developmental processes of synaptogenesis, dendritogenesis and myelination are highly dependent on iron containing enzymes, subtle deficiencies of iron content throughout specific developmental epochs may ultimately influence mechanisms of subcortical neurochemical signalling and drive the acquisition of deficits in motor, affective and cognitive behaviours as manifested in GTS.

11.3 Aripiprazole improves associated comorbid conditions in addition to tics in GTS

Considering that the drug discovery and development process and is a long and arduous one that may take up to an average of 12 years, it is thus critical to carefully examine the influence of currently available drugs on the clinical characteristics of the patients. In this work, we chose to investigate the influence of aripiprazole on the clinical characteristics of the patient sample given its efficacy and its favorable side effect profile [170]. Key findings and significance 155

Although randomized controlled trials including large patient samples are currently lacking, aripiprazole is currently considered as a ﬁrst choice drug for the treatment of tics by the European and German societies [6].

Currently, the majority of clinical investigations on aripiprazole have been conducted on child and adolescent populations. Additionally, most studies only explored the influence of aripiprazole on motor features of GTS, with a small number of reports investigating its influence on comorbid conditions. In this regard, we aimed at investigating the influence of aripiprazole on both motor and non-motor features exhibited by GTS, and additionally aimed at further exploring its efficacy and safety in a group of drug-free adult patients. Given that the patients recruited at baseline were presented with the option of partaking in the longitudinal study design, in which about half elected to undergo treatment, we were also able to investigate the factors that influence a patients’ decision in electing for or against treatment.

Similar to previous work, our results indicated that aripiprazole leads to significant reductions of tics in adult patients with GTS. Interestingly, however, we did not find any effect on premonitory urges which precede tics, indicating that tics and premonitory urges are not necessarily related events as suggested by other studies [398, 399]. We also found that tic severity, premonitory urges and quality of life were not significant factors in a patient’s decision in undergoing treatment.

In relation to psychiatric comorbidities, the sample exhibited reductions in some of the associated conditions. Obsessive compulsive disorder (OCD) diagnosis decreased by approximately 50% and no patients developed OCD symptoms following treatment. Simi- larly, the number of patients diagnosed with depression and anxiety decreased by about 10%. Notably, none of the patients developed OCD, depression or anxiety disorder during treatment with aripiprazole. On the other hand, attention deficit/hyperactivity disorder (ADHD) classification remained unchanged in the patients that underwent treatment. While patients with comorbid ADHD tended to elect against undergoing treatment, it is interesting to report that patients with comorbid OCD tended to elect for undergoing treatment. In general, no patients reported any severe side effects. Similar to previous work, our results indicated that aripiprazole is a safe and reliable treatment.

As an atypical second-generation antipsychotic, aripiprazole presents an example of a functionally selective drug that exhibits an adaptive pharmacological profile that produces a mix of effects through the activation or inhibition of a limited number of signal transduction pathways as a result of its ability to induce unique G protein-coupled receptor conformations. While it mainly targets the dopaminergic system, recent studies [171] have indicated it also exhibit potent effects on other systems including the serotonergic, GABAergic and glutamatergic system. In this regard, it may be plausible to suggest Key findings and significance 156 that efficacy of aripiprazole in treating motor as well as non-motor features are of GTS, are a result of its influence on the multiple affected systems in GTS. The treatment of motor and non-motor features associated with GTS may stand to be more effective with the design of functionally selective modulators with an adaptive pharmacological profile that targets multiple systems.

11.4 Signiﬁcance

Our results indicate that patients with GTS exhibit an abnormality in the flux of metabolites in the GABA-glutamate-glutamine cycle, thus implying perturbations in astrocytic- neuronal coupling systems that maintain the subtle balance between excitatory and inhibitory neurotransmission within subcortical nuclei. These abnormalities may be driven or further compounded by abnormalities in iron metabolism. Chronic perturbations in the subcortical GABA-Glu-Gln cycle flux could lead to spatially focalized alterations in excitatory, inhibitory and modulatory subcortical neurochemical ratios that would have a profound influence on the neuroplastic mechanisms involved in reinforcement learning and habit formation systems. This work sheds a new light on the neurobiological basis of GTS and provides novel clues that may prove critical in the future development of functionally selective pharmacological modulators that target multiple neurochemical systems.

*** Part VI

Bibliography

157 Bibliography

[1] Gilles de la Tourette. Étude sur une aﬀection nerveuse caractérisée par de l’incoordination motrice accompagnée d’écholalie et de coprolalie. Arch. Neurol, 9(c):19–42, 1885.

[2] Hugh Rickards, Ian Woolf, and Andrea Eugenio Cavanna. "Trousseau’s disease:" a description of the Gilles de la Tourette syndrome 12 years before 1885. Movement disorders: oﬃcial journal of the Movement Disorder Society, 25(14):2285–9, oct 2010.

[3] Sheryl Geisler. Une Lecon Clinique a la Salpetriere (A Clinical Lesson at the Salpetriere), Andre Brouillet (1887). The Journal of Physician Assistant Education, 22(3), 2011.

[4] Arthur K Shapiro and E Shapiro. Treatment of Gilles de la Tourette’s Syndrome with Haloperidol. The British Journal of Psychiatry, 114(508):345–350, mar 1968.

[5] Danielle C Cath, Tammy Hedderly, Andrea G Ludolph, Jeremy S Stern, Tara Murphy, Andreas Hartmann, Virginie Czernecki, Mary May Robertson, Davide Martino, A Munchau, and R Rizzo. European clinical guidelines for Tourette Syndrome and other tic disorders. Part I: assessment. European Child & Adolescent Psychiatry, 20(4):155–171, 2011.

[6] Veit Roessner, Kerstin J Plessen, Aribert Rothenberger, Andrea G Ludolph, Renata Rizzo, Liselotte Skov, Gerd Strand, Jeremy S Stern, Cristiano Termine, and Pieter J Hoekstra. Eu- ropean clinical guidelines for Tourette syndrome and other tic disorders. Part II: pharmacological treatment. European Child & Adolescent Psychiatry, 20(4):173–96, apr 2011.

[7] Cara Verdellen, Jolande van de Griendt, Andreas Hartmann, and Tara Murphy. European clinical guidelines for Tourette syndrome and other tic disorders. Part III: behavioural and psychosocial interventions. European Child & Adolescent Psychiatry, 20:197–207, 2011.

[8] Kirsten R Müller-Vahl, Danielle C Cath, Andrea E Cavanna, Sandra Dehning, Mauro Porta, Mary M Robertson, and Veerle Visser-Vandewalle. European clinical guidelines for Tourette syndrome and other tic disorders. Part IV: deep brain stimulation. European Child & Adolescent Psychiatry, 20(4):209–17, apr 2011.

[9] James F Leckman. Tourette’s syndrome. Lancet, 360(9345):1577–86, nov 2002.

[10] Tristan Knight, Thomas Steeves, Lundy Day, Mark Lowerison, Nathalie Jette, and Tamara Pring- sheim. Prevalence of tic disorders: a systematic review and meta-analysis. Pediatric neurology, 47(2):77–90, aug 2012.

[11] Joseph Jankovic. Tourette syndrome. Phenomenology and classiﬁcation of tics. Neurologic Clinics, 15:267–275, 1997.

[12] Michael H Bloch and James F Leckman. Clinical course of Tourette syndrome. Journal of Psy- chosomatic Research, 67:497–501, 2009.

[13] James F. Leckman, Heping Zhang, Amy Vitale, Fatima Lahnin, Kimberly Lynch, Colin Bondi, Young-Shin Kim, and Bradley S. Peterson. Course of tic severity in Tourette syndrome: the ﬁrst two decades. Pediatrics, 102:14–19, 1998.

[14] Carolyn Kwak, Kevin Dat Vuong, and Joseph Jankovic. Premonitory sensory phenomenon in Tourette’s syndrome. Movement disorders oﬃcial journal of the Movement Disorder Society, 18:1530–1533, 2003.

158 Bibliography 159

[15] Stephanie C Cohen, James F Leckman, and Michael H Bloch. Neuroscience and Biobehavioral Re- views Clinical assessment of Tourette syndrome and tic disorders. Neuroscience and Biobehavioral Reviews, 37(6):997–1007, 2013.

[16] Matthew E Hirschtritt, Paul C Lee, David L Pauls, Yves Dion, Marco A Grados, Cornelia Illmann, Robert A King, Paul Sandor, William M McMahon, Gholson J Lyon, Danielle C Cath, Roger Kurlan, Mary M. Robertson, Lisa Osiecki, Jeremiah M Scharf, and Carol A Mathews. Lifetime prevalence, age of risk, and genetic relationships of comorbid psychiatric disorders in Tourette syndrome. JAMA psychiatry, 72(4):325–33, 2015.

[17] Nanette M Debes, Helle Hjalgrim, and Liselotte Skov. The presence of attention-deﬁcit hyperactivity disorder (ADHD) and obsessive-compulsive disorder worsen psychosocial and educational problems in Tourette syndrome. Journal of child neurology, 25:171–181, 2010.

[18] Mary M Robertson. Tourette syndrome, associated conditions and the complexities of treatment. Brain: a journal of neurology, 123(3):425–462, mar 2000.

[19] American Psychiatric Association. Diagnostic and statistical manual of mental disorders: DSM-5. American Psychiatric Association, Washington, D.C., 2013.

[20] Kevin SP McNaught and Jonathan W Mink. Advances in understanding and treatment of Tourette syndrome. Nature reviews. Neurology, 7(12):667–76, dec 2011.

[21] Avshalom Caspi and Terrie E Moﬃtt. Gene-environment interactions in psychiatry: joining forces with neuroscience. Nature reviews. Neuroscience, 7(7):583–90, jul 2006.

[22] Hao Deng, Kai Gao, and Joseph Jankovic. The genetics of Tourette syndrome. Nature reviews. Neurology, 8(4):203–13, apr 2012.

[23] David L. Pauls, Thomas V. Fernandez, Carol a. Mathews, Matthew W. State, and Jeremiah M. Scharf. The inheritance of Tourette Disorder: A review. Journal of Obsessive-Compulsive and Related Disorders, 3:380–385, 2014.

[24] Pieter J Hoekstra, Andrea Dietrich, Mark J Edwards, Ishraga Elamin, and Davide Martino. En- vironmental factors in Tourette syndrome. Neuroscience and biobehavioral reviews, 37(6):1040–9, jul 2013.

[25] Carol A Mathews, Brianne Bimson, Thomas L Lowe, Luis Diego Herrera, Cathy L Budman, Gerald Erenberg, Allen Naarden, Ruth D Bruun, Nelson B Freimer, and Victor I Reus. Association Between Maternal Smoking and Increased Symptom Severity in Tourette’s Syndrome. American Journal of Psychiatry, 163(June):1066–1073, 2006.

[26] Douglas L Leslie, Laura Kozma, Andrés Martin, Angeli Landeros, Liliya Katsovich, Robert a King, and James F Leckman. Neuropsychiatric disorders associated with streptococcal infection: a case-control study among privately insured children. Journal of the American Academy of Child and Adolescent Psychiatry, 47:1166–1172, 2008.

[27] Loren K Mell, Robert L Davis, and David Owens. Association between streptococcal infection and obsessive-compulsive disorder, Tourette’s syndrome, and tic disorder. Pediatrics, 116(1):56–60, 2005.

[28] Ryan J Felling and Harvey S Singer. Neurobiology of Tourette Syndrome: Current Status and Need for Further Investigation. The Journal of neuroscience: the oﬃcial journal of the Society for Neuroscience, 31(35):12387–12395, aug 2011.

[29] Christos Ganos, Veit Roessner, and Alexander Münchau. The functional anatomy of Gilles de la Tourette syndrome. Neuroscience and biobehavioral reviews, 37(6):1050–62, jul 2013.

[30] Deanna J Greene, Kevin J Black, and Bradley L Schlaggar. Neurobiology and functional anatomy of tic disorders. In D. Martino and J Leckman, editors, Tourette Syndrome, pages 238–275. Oxford University Press, 2013.

[31] Harvey Singer. The Neurochemistry of Tourette Syndrome. In Davide Martino and James F. Leckman, editors, Tourette Syndrome, pages 276–297. Oxford University Press, 2013.

[32] Bogdan Draganski, Davide Martino, Andrea E Cavanna, Chloe Hutton, Michael Orth, Mary M Bibliography 160

Robertson, Hugo D Critchley, and Richard S Frackowiak. Multispectral brain morphometry in Tourette syndrome persisting into adulthood. Brain: a journal of neurology, 133(Pt 12):3661–75, dec 2010.

[33] Bradley Peterson, Malcolm Riddle, D. J. Cohen, L. D. Katz, J. C. Smith, M. T. Hardin, and J F Leckman. Reduced basal ganglia volumes in Tourette’s syndrome using three-dimensional reconstruction techniques from magnetic resonance images. Neurology, 43(5):941–941, may 1993.

[34] Bradley S Peterson, Prakash Thomas, Michael J Kane, Lawrence Scahill, Heping Zhang, Richard Bronen, Robert a King, James F Leckman, and Lawrence Staib. Basal Ganglia volumes in patients with Gilles de la Tourette syndrome. Archives of general psychiatry, 60(4):415–24, apr 2003.

[35] Cherine Fahim, Uicheul Yoon, Samir Das, Oliver Lyttelton, John Chen, Rozie Arnaoutelis, Guy Rouleau, Paul Sandor, Kirk Frey, Catherine Brandner, and Alan C Evans. Somatosensory-motor bodily representation cortical thinning in Tourette: eﬀects of tic severity, age and gender. Cortex; a journal devoted to the study of the nervous system and behavior, 46(6):750–60, jun 2010.

[36] Elizabeth R Sowell, Eric Kan, June Yoshii, Paul M Thompson, Ravi Bansal, Dongrong Xu, Arthur W Toga, and Bradley S Peterson. Thinning of sensorimotor cortices in children with Tourette syndrome. Nature neuroscience, 11(6):637–9, jun 2008.

[37] Yulia Worbe, Emilie Gerardin, Andreas Hartmann, Romain Valabrégue, Marie Chupin, Léon Tremblay, Marie Vidailhet, Olivier Colliot, and Stéphane Lehéricy. Distinct structural changes underpin clinical phenotypes in patients with Gilles de la Tourette syndrome. Brain: a journal of neurology, 133(Pt 12):3649–60, dec 2010.

[38] James F Leckman, Michael H Bloch, Megan E Smith, Daouia Larabi, and Michelle Hampson. Neurobiological substrates of Tourette’s disorder. Journal of Child and Adolescent Psychophar- macology, 20(4):237–47, aug 2010.

[39] Michelle Hampson, Fuyuze Tokoglu, Robert A King, R Todd Constable, and James F Leckman. Brain Areas Coactivating with Motor Cortex During Chronic Motor Tics and Intentional Move- ments. Biological Psychiatry, 65(7):594–599, feb 2009.

[40] Zhishun Wang, TV Maia, and Rachel Marsh. The neural circuits that generate tics in Tourette’s syndrome. American Journal of . . . , pages 1326–1337, 2011.

[41] Bradley S Peterson, Pawel Skudlarski, Adam W Anderson, Heping Zhang, Chris Gatenby, Cheryl M Lacadie, James F Leckman, and John C Gore. A functional magnetic resonance imaging study of tic suppression in Tourette syndrome. Archives of general psychiatry, 55(4):326–33, apr 1998.

[42] Stephan Bohlhalter, A Goldﬁne, S Matteson, G Garraux, T Hanakawa, K Kansaku, R Wurz- man, and M Hallett. Neural correlates of tic generation in Tourette syndrome: an event-related functional MRI study. Brain: a journal of neurology, 129(Pt 8):2029–37, aug 2006.

[43] Luigi Mazzone, Shan Yu, Clancy Blair, Benjamin C Gunter, Zhishun Wang, Rachel Marsh, and Bradley S Peterson. An FMRI study of frontostriatal circuits during the inhibition of eye blinking in persons with Tourette syndrome. The American journal of psychiatry, 167(3):341–9, mar 2010.

[44] Bharat Biswal, John L Ulmer, Robert L Krippendorf, Harold H Harsch, David L Daniels, James S Hyde, and Victor M Haughton. Abnormal cerebral activation associated with a motor task in Tourette syndrome. AJNR. American journal of neuroradiology, 19(8):1509–12, sep 1998.

[45] Nanette M Debes, Adam Hansen, Liselotte Skov, and Henrik Larsson. A functional magnetic resonance imaging study of a large clinical cohort of children with Tourette syndrome. Journal of child neurology, 26(5):560–569, 2011.

[46] Robert A Bornstein, G B Baker, T Bazylewich, and A B Douglass. Tourette syndrome and neuropsychological performance. Acta Psychiatrica Scandinavica, 84:212–216, 1991.

[47] Shelley Channon, Polly Pratt, and Mary M Robertson. Executive function, memory, and learning in Tourette’s syndrome. Neuropsychology, 17:247–254, 2003.

[48] Laura H Watkins, Barbara J Sahakian, Mary M Robertson, David M Veale, Robert D Rogers, Kathryn M Pickard, Michael RF Aitken, and Trevor W Robbins. Executive function in Tourette’s Bibliography 161

syndrome and obsessive-compulsive disorder. Psychological Medicine, 35:571–582, 2005.

[49] Sally Ozonoﬀ, David L Strayer, William M McMahon, and Francis Filloux. Inhibitory deﬁcits in Tourette syndrome: a function of comorbidity and symptom severity. Journal of child psychology and psychiatry, and allied disciplines, 39(8):1109–18, 1998.

[50] Georgina M Jackson, S C Mueller, K Hambleton, and C P Hollis. Enhanced cognitive control in Tourette Syndrome during task uncertainty. Experimental Brain Research, 182:357–364, 2007.

[51] Sven C Mueller, Georgina M Jackson, Ranu Dhalla, Sophia Datsopoulos, and Chris P Hollis. Enhanced cognitive control in young people with Tourette’s syndrome. Current Biology, 16:570– 573, 2006.

[52] Rachel Marsh, Hongtu Zhu, Ahishun Wang, Pawel Skudlarski, and Bradley Peterson. A developmental fMRI study of Seld-Regulatory Control in Tourette’s Syndrome. American Journal of Psychiatry, 164(6):955–966, 2007.

[53] Nico Brand, Rinie Geenen, Milo Oudenhoven, Bastiaan Lindenborn, Annette Van Der Ree, Peggy Cohen-Kettenis, and Jan K Buitelaar. Brief report: cognitive functioning in children with Tourette’s syndrome with and without comorbid ADHD. Journal of Pediatric Psychology, 27:203– 208, 2002.

[54] Tamara Hershey, Kevin J Black, Johanna M Hartlein, Deanna M Barch, Todd S Braver, Juanita L Carl, and Joel S Perlmutter. Cognitive-pharmacologic functional magnetic resonance imaging in tourette syndrome: a pilot study. Biological Psychiatry, 55:916–925, 2004.

[55] Amir Raz, Hongtu Zhu, Shan Yu, Ravi Bansal, Zhishun Wang, Gerianne M Alexander, Jason Royal, and Bradley S Peterson. Neural substrates of self-regulatory control in children and adults with Tourette syndrome. Canadian journal of psychiatry. Revue canadienne de psychiatrie, 54(9):579–588, 2009.

[56] Aaron Alexander-Bloch, Liv Clasen, Michael Stockman, Lisa Ronan, Francois Lalonde, Jay Giedd, and Armin Raznahan. Subtle in-scanner motion biases automated measurement of brain anatomy from in vivo MRI. Human Brain Mapping, 37(7):2385–2397, 2016.

[57] Ludovica Griﬀanti, Gholamreza Salimi-Khorshidi, Christian F Beckmann, Edward J Auerbach, Gwenaëlle Douaud, Claire E Sexton, Eniko Zsoldos, Klaus P Ebmeier, Nicola Filippini, Clare E Mackay, Steen Moeller, Junqian Xu, Essa Yacoub, Giuseppe Baselli, Kamil Ugurbil, Karla L Miller, and Stephen M Smith. ICA-based artefact removal and accelerated fMRI acquisition for improved resting state network imaging. NeuroImage, 95:232–47, jul 2014.

[58] John Muschelli, Mary Beth Nebel, Brian S Caﬀo, Anita D Barber, James J Pekar, and Stew- art H Mostofsky. Reduction of motion-related artifacts in resting state fMRI using aCompCor. NeuroImage, 96:22–35, aug 2014.

[59] Ameera X Patel, Prantik Kundu, Mikail Rubinov, P Simon Jones, Petra E Vértes, Karen D Ersche, John Suckling, and Edward T Bullmore. A wavelet method for modeling and despiking motion artifacts from resting-state fMRI time series. NeuroImage, 95:287–304, jul 2014.

[60] Jonathan D Power, Anish Mitra, Timothy O Laumann, Abraham Z Snyder, Bradley L Schlaggar, and Steven E Petersen. Methods to detect, characterize, and remove motion artifact in resting state fMRI. NeuroImage, 84:320–341, aug 2013.

[61] Raimon HR Pruim, Maarten Mennes, Daan van Rooij, Alberto Llera, Jan K Buitelaar, and Christian F Beckmann. ICA-AROMA: A robust ICA-based strategy for removing motion artifacts from fMRI data. NeuroImage, 112:267–277, 2015.

[62] Raimon H R Pruim, Maarten Mennes, Jan K Buitelaar, and Christian F Beckmann. Evaluation of ICA-AROMA and alternative strategies for motion artifact removal in resting state fMRI. NeuroImage, 112:278–287, 2015.

[63] Theodore D Satterthwaite, Mark a Elliott, Raphael T Gerraty, Kosha Ruparel, James Loughead, Monica E Calkins, Simon B Eickhoﬀ, Hakon Hakonarson, Ruben C Gur, Raquel E Gur, and Daniel H Wolf. An improved framework for confound regression and ﬁltering for control of motion artifact in the preprocessing of resting-state functional connectivity data. NeuroImage, 64:240–56, jan 2013. Bibliography 162

[64] Ling-Li Zeng, Danhong Wang, Michael D Fox, Mert Sabuncu, Dewen Hu, Manling Ge, Randy L Buckner, and Hesheng Liu. Neurobiological basis of head motion in brain imaging. Proceedings of the National Academy of Sciences of the United States of America, 111(16):6058–62, apr 2014.

[65] Jonathan D Power, Kelly a Barnes, Abraham Z Snyder, Bradley L Schlaggar, and Steven E Petersen. Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. NeuroImage, 59(3):2142–54, feb 2012.

[66] Jonathan D Power, Bradley L Schlaggar, and Steven E Petersen. Recent progress and outstanding issues in motion correction in resting state fMRI, oct 2015.

[67] Martin Reuter, M Dylan Tisdall, Abid Qureshi, Randy L Buckner, André J W van der Kouwe, and Bruce Fischl. Head motion during MRI acquisition reduces gray matter volume and thickness estimates. NeuroImage, 107:107–115, 2015.

[68] Malcolm B Carpenter. Core text of neuroanatomy. Williams & Wilkins, Baltimore, fourth edition, 1991.

[69] Garrett E. Alexander and Michael D. Crutcher. Functional architecture of basal ganglia circuits: neural substrates of parallel processing. Trends in Neurosciences, 13(7):266–271, 1990.

[70] Mahlon DeLong and Thomas Wichmann. Update on models of basal ganglia function and dysfunction. Parkinsonism & related disorders, 15 Suppl 3:S237–40, dec 2009.

[71] Janet M Kemp and T P S Powell. The Connexions of the Striatum and Globus Pallidus: Synthesis and Speculation. Philosophical Transactions of the Royal Society of London Series B, Biological Sciences, 262(845):441–457, sep 1971.

[72] Garret E Alexander, Mahlon R DeLong, and Peter L Strick. Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annual review of neuroscience, 9:357–381, 1986.

[73] Paul Krack, Marwan I Hariz, Christelle Baunez, Jorge Guridi, and Jose a Obeso. Deep brain stimulation: from neurology to psychiatry? Trends in neurosciences, 33(10):474–84, oct 2010.

[74] Mahlon R DeLong and Thomas Wichmann. Circuits and circuit disorders of the basal ganglia. Archives of neurology, 64(1):20–24, jan 2007.

[75] Léon Tremblay, Yulia Worbe, Stéphane Thobois, Véronique Sgambato-Faure, and Jean Féger. Selective dysfunction of basal ganglia subterritories: From movement to behavioral disorders. Movement disorders: oﬃcial journal of the Movement Disorder Society, 30(9):1155–70, 2015.

[76] Roger L Albin and Jonathan W Mink. Recent advances in Tourette syndrome research. Trends in Neurosciences, 29(3):175–182, mar 2006.

[77] Kendra Harris and Harvey S Singer. Tic disorders: Neural circuits, neurochemistry, and neuroim- munology. Journal of Child Neurology, 21(8):678–89, 2006.

[78] Jonathan W Mink. Basal ganglia dysfunction in Tourette’s syndrome: A new hypothesis. Pediatric Neurology, 25(3):190–198, 2001.

[79] Jonathan W. Mink. The basal ganglia: Focused selection and inhibition of competing motor programs, 1996.

[80] Yulia Worbe, Véronique Sgambato-Faure, Justine Epinat, Marion Chaigneau, Dominique Tandé, Chantal Francois, Jean Féger, and Léon Tremblay. Towards a primate model of Gilles de la Tourette syndrome: Anatomo-behavioural correlation of disorders induced by striatal dysfunction. Cortex, 49(4):1126–1140, 2013.

[81] Yulia Worbe, Nicolas Baup, David Grabli, Marion Chaigneau, Stéphanie Mounayar, Kevin Mc- Cairn, Jean Féger, and Léon Tremblay. Behavioral and movement disorders induced by local inhibitory dysfunction in primate striatum. Cerebral Cortex, 19(8):1844–1856, 2009.

[82] Ann M Graybiel. The basal ganglia and chunking of action repertoires. Neurobiology of learning and memory, 70(1-2):119–36, 1998. Bibliography 163

[83] Jill R Crittenden and Ann M Graybiel. Basal Ganglia disorders associated with imbalances in the striatal striosome and matrix compartments. Frontiers in neuroanatomy, 5(59), sep 2011.

[84] Henry H Yin, Barbara J Knowlton, and Bernard W Balleine. Lesions of dorsolateral striatum preserve outcome expectancy but disrupt habit formation in instrumental learning. European Journal of Neuroscience, 19:181–189, 2004.

[85] Carlos Cepeda, N a Buchwald, and M S Levine. Neuromodulatory actions of dopamine in the neostriatum are dependent upon the excitatory amino acid receptor subtypes activated. Proceed- ings of the National Academy of Sciences of the United States of America, 90(20):9576–80, oct 1993.

[86] James Surmeier, Jun Ding, Michelle Day, Zhongfeng Wang, and Weixing Shen. D1 and D2 dopamine-receptor modulation of striatal glutamatergic signaling in striatal medium spiny neurons. Trends in neurosciences, 30(5):228–35, may 2007.

[87] Anthony R West and Anthony a Grace. Opposite inﬂuences of endogenous dopamine D1 and D2 receptor activation on activity states and electrophysiological properties of striatal neurons: studies combining in vivo intracellular recordings and reverse microdialysis. The Journal of neuroscience: the oﬃcial journal of the Society for Neuroscience, 22(1):294–304, jan 2002.

[88] Carlos Cepeda and Michael S Levine. Dopamine and N-methyl-D-aspartate receptor interactions in the neostriatum. Dev Neurosci, 20:1–18, 1998.

[89] Harvey S Singer, I J Butler, L E Tune, W E Seifert, and J T Coyle. Dopaminergic dsyfunction in Tourette syndrome. Annals of neurology, 12(0364-5134 (Print)):361–366, 1982.

[90] Donald L Gilbert and Joseph Jankovic. Pharmacological treatment of Tourette syndrome. Journal of Obsessive-Compulsive and Related Disorders, 3(4):407–414, 2015.

[91] Andreas Hartmann and Yulia Worbe. Pharmacological treatment of Gilles de la Tourette syndrome. Neuroscience and biobehavioral reviews, 37(6):1157–61, jul 2013.

[92] Harvey S. Singer. Treatment of tics and tourette syndrome. Current Treatment Options in Neu- rology, 12(6):539–561, 2010.

[93] Mauro Porta, Marco Sassi, Mario Cavallazzi, Maurizio Fornari, Arianna Brambilla, and Domenico Servello. Tourette’s Syndrome and Role of Tetrabenazine. Clinical Drug Investigation, 28(7):443– 459, 2008.

[94] Sweet RD, R Bruun, E Shapiro, and Shapiro AK. Presynaptic catecholamine antagonists as treatment for tourette syndrome: Eﬀects of alpha methyl para tyrosine and tetrabenazine. Archives of General Psychiatry, 31(6):857–861, dec 1974.

[95] Keun-Ah Cheon, Young-Hoon Ryu, Kee Namkoong, Chan-Hyung Kim, Jae-Jin Kim, and Jong Doo Lee. Dopamine transporter density of the basal ganglia assessed with [123I]IPT SPECT in drug- naive children with Tourette’s disorder. Psychiatry research, 130(1):85–95, 2004.

[96] Donald L Gilbert, BT Christian, MJ Gelfand, B Shi, J Mantil, and FR Sallee. Altered mesolim- bocortical and thalamic dopamine in Tourette syndrome. Neurology, 67(9):1695–1697, 2006.

[97] Cheng Chun Lee, I. Ching Chou, Chang Hai Tsai, Tso Ren Wang, Tsai Chung Li, and Fuu Jen Tsai. Dopamine receptor D2 gene polymorphisms are associated in Taiwanese children with Tourette syndrome. Pediatric Neurology, 33(4):272–276, 2005.

[98] R T Malison, C J McDougle, C H van Dyck, L Scahill, R M Baldwin, J P Seibyl, L H Price, J F Leckman, and R B Innis. [123I]beta-CIT SPECT imaging of striatal dopamine transporter binding in Tourette’s disorder. The American journal of psychiatry, 152(9):1359–1361, 1995.

[99] I. Mena, M. Miranda, M. Hernandez, and M. Fruns. Dopamine transporter imaging in Tourette syndrome: Evaluation by NeuroSPECT of Trodat 1-Tc99m. In Movement Disorders Society, volume 19, pages S437 – P1282, 2004.

[100] Karen Minzer, Olivia Lee, John J Hong, and Harvey S Singer. Increased prefrontal D2 protein in Tourette syndrome: A postmortem analysis of frontal cortex and striatum. Journal of the Neurological Sciences, 219(1-2):55–61, 2004. Bibliography 164

[101] Kirsten R Müller-Vahl, G Berding, T Brücke, H Kolbe, G J Meyer, H Hundeshagen, R Dengler, W H Knapp, and H M Emrich. Dopamine transporter binding in Gilles de la Tourette syndrome. Journal of neurology, 247:514–520, 2000.

[102] Kirsten R Müller-Vahl, G Berding, H Kolbe, G J Meyer, H Hundeshagen, R Dengler, W H Knapp, and H M Emrich. Dopamine D2 receptor imaging in Gilles de la Tourette syndrome. Acta neurologica Scandinavica, 101(3):165–71, mar 2000.

[103] Thomas D L Steeves, Ji Hyun Ko, David M Kideckel, Pablo Rusjan, Sylvain Houle, Paul Sandor, Anthony E Lang, and Antonio P Strafella. Extrastriatal Dopaminergic Dysfunction in Tourette Syndrome. Annals of neurology, 67(2):170–181, feb 2010.

[104] Jordi Serra-Mestres, H A Ring, D C Costa, S Gacinovic, Z Walker, A J Lees, M M Robertson, and M R Trimble. Dopamine transporter binding in Gilles de la Tourette syndrome: a [123I]FP- CIT/SPECT study. Acta Psychiatr Scand, 109(2):140–146, 2004.

[105] Nora Turjanski, G V Sawle, E D Playford, R Weeks, A A Lammerstma, A J Lees, and D J Brooks. PET studies of the presynaptic and postsynaptic dopaminergic system in Tourette’s syndrome. Journal of Neurology, Neurosurgery & Psychiatry, 57:688–692, 1994.

[106] Steven S Wolf, Douglas W Jones, Michael B Knable, Julia G Gorey, Kan Sam Lee, Thomas M Hyde, Richard Coppola, and Daniel R Weinberger. Tourette syndrome: prediction of phenotypic variation in monozygotic twins by caudate nucleus D2 receptor binding. Science, 273(5279):1225– 1227, aug 1996.

[107] Dean F Wong, Harvey S Singer, Jason Brandt, Elias Shaya, Catherine Chen, Jan Brown, a W Kimball, Albert Gjedde, Robert F Dannals, Hayden T Ravert, P David Wilson, and Henry N Wagner. D2-like dopamine receptor density in Tourette syndrome measured by PET. Journal of nuclear medicine: oﬃcial publication, Society of Nuclear Medicine, 38(8):1243–7, aug 1997.

[108] Dean F. Wong, James R. Brašić, Harvey S. Singer, David J. Schretlen, Hiroto Kuwabara, Yun Zhou, Ayon Nandi, Marika A. Maris, Mohab Alexander, Weiguo Ye, Olivier Rousset, Anil Kumar, Zsolt Szabo, Albert Gjedde, and Anthony A. Grace. Abnormalities of dopamine and serotonin neuroreceptors with pet in Tourette syndrome. In Society for Neuroscience, page 1012.15, 2005.

[109] Dean F Wong, J R Brasic, Harvey S Singer, David J Schretlen, Hiroto Kuwabara, Yun Zhou, Ayon Nandi, Marika a Maris, Mohab Alexander, Weiguo Ye, Olivier Rousset, Anil Kumar, Zsolt Szabo, Albert Gjedde, and Anthony a Grace. Mechanisms of dopaminergic and serotonergic neurotransmission in Tourette syndrome: Clues from an in vivo neurochemistry study with PET. Neuropsychopharmacology, 33(6):1239–1251, may 2008.

[110] Dustin Y Yoon, Colin D Gause, James F Leckman, and Harvey S Singer. Frontal dopaminergic abnormality in Tourette syndrome: a postmortem analysis. Journal of the neurological sciences, 255(1-2):50–56, 2007.

[111] Ian J Butler, Stephan H Koslow, William E Seifert, Richard M Caprioli, and Harvey S Singer. Biogenic amine metabolism in Tourette syndrome. Annals of neurology, 6(1):37–39, 1979.

[112] Harvey S Singer, I H Hahn, E Krowiak, E Nelson, and T Moran. Tourette’s syndrome: a neurochemical analysis of postmortem cortical brain tissue. Annals of Neurology, 27:443–446, 1990.

[113] Michael J Beckstead, D K Grandy, K Wickman, and J T Williams. Vesicular dopamine release elicits an inhibitory postsynaptic current in midbrain dopamine neurons. Neuron, 42(6):939–946, 2004.

[114] Michael J. Beckstead, Christopher P. Ford, Paul E M Phillips, and John T. Williams. Presy- naptic regulation of dendrodendritic dopamine transmission. European Journal of Neuroscience, 26(6):1479–1488, 2007.

[115] Stephanie J Cragg and Sussan a Greenfield. Differential autoreceptor control of somatodendritic and axon terminal dopamine release in substantia nigra, ventral tegmental area, and striatum. The Journal of neuroscience: the official journal of the Society for Neuroscience, 17(15):5738– 5746, 1997.

[116] Christopher P Ford, Stephanie C Gantz, Paul EM Phillips, and John T Williams. Control of Bibliography 165

Extracellular Dopamine at Dendrite and Axon Terminals. Journal of Neuroscience, 30(20):6975– 6983, 2010.

[117] MG Lacey, NB Mercuri, and RA North. Dopamine acts on D2 receptors to increase potassium conductance in neurones of the rat substantia nigra zona compacta. The Journal of physiology, 392:397–416, 1987.

[118] Margaret E Rice, Jyoti C Patel, and Stephanie J Cragg. Dopamine release in the basal ganglia. Neuroscience, 198:112–137, 2011.

[119] Stan B Floresco, Christopher L Todd, and Anthony A Grace. Glutamatergic aﬀerents from the hippocampus to the nucleus accumbens regulate activity of ventral tegmental area dopamine neurons. The Journal of neuroscience: the oﬃcial journal of the Society for Neuroscience, 21(13):4915–4922, 2001.

[120] Stan B Floresco, Anthony R West, Brian Ash, Holly Moore, and Anthony A Grace. Afferent modulation of dopamine neuron firing differentially regulates tonic and phasic dopamine transmission. Nature Neuroscience, 6(9):968–973, 2003.

[121] Anthony A Grace, Stan B. Floresco, Yukiori Goto, and Daniel J. Lodge. Regulation of ﬁring of dopaminergic neurons and control of goal-directed behaviors. Trends in Neurosciences, 30(5):220– 227, 2007.

[122] James M. Tepper and J. Paul Bolam. Functional diversity and speciﬁcity of neostriatal interneurons. Current Opinion in Neurobiology, 14(6):685–692, 2004.

[123] Paolo Calabresi, Barbara Picconi, Alessandro Tozzi, Veronica Ghiglieri, and Massimiliano Di Filippo. Direct and indirect pathways of basal ganglia: a critical reappraisal. Nature Neuroscience, 17(8):1022–1030, 2014.

[124] George M Anderson, Eleanor S Pollak, Diptendu Chatterjee, James F Leckman, Mark A Riddle, and Donald J Cohen. Postmortem analysis of subcortical monoamines and amino acids in Tourette syndrome. Advances in neurology, 58:123–133, 1992.

[125] Paul S Kalanithi, Wei Zheng, Yuko Kataoka, Marian DiFiglia, Heidi Grantz, Cliﬀord B Saper, Michael L Schwartz, James F Leckman, and Flora M Vaccarino. Altered parvalbumin-positive neuron distribution in basal ganglia of individuals with Tourette syndrome. Proceedings of the National Academy of Sciences of the United States of America, 102(37):13307–13312, 2005.

[126] Yuko Kataoka, P. S a Kalanithi, Heidi Grantz, Michael L. Schwartz, Cliﬀord Saper, J F Leckman, and Flora M. Vaccarino. Decreased number of parvalbumin and cholinergic interneurons in the striatum of individuals with tourette syndrome. Journal of Comparative Neurology, 518(3):277– 291, 2010.

[127] Alicja Lerner, Anto Bagic, Janine M Simmons, Zoltan Mari, Omer Bonne, Ben Xu, Diane Kazuba, Peter Herscovitch, Richard E Carson, Dennis L Murphy, Wayne C Drevets, and Mark Hallett. Widespread abnormality of the γ-aminobutyric acid-ergic system in Tourette syndrome. Brain: a journal of neurology, 135(6):1926–1936, may 2012.

[128] Nicolaas AJ Puts, Ashley D Harris, Deana Crocetti, Carrie Nettles, Harvey S Singer, Mark Tom- merdahl, Richard Ae Edden, and Stewart H Mostofsky. Reduced GABAergic inhibition and abnormal sensory processing in children with Tourette Syndrome. Journal of Neurophysiology, page jn.00060.2015, 2015.

[129] Ashley D Harris, HS Singer, A Horska, T Kline, M Ryan, RAE Edden, and EM Mahone. GABA and glutamate in children with primary complex motor stereotypies: An 1H-MRS study at 7T. American Journal of Neuroradiology, 37(3):552–557, 2016.

[130] Amelia Draper, Mary C Stephenson, Georgina M Jackson, Sophia Pépés, Paul S Morgan, Peter G Morris, and Stephen R Jackson. Increased GABA Contributes to Enhanced Control over Motor Excitability in Tourette Syndrome. Current biology, 24(19):2343–7, oct 2014.

[131] Harvey S Singer, Christina Morris, and Marco Grados. Glutamatergic modulatory therapy for Tourette syndrome. Medical hypotheses, 74(5):862–867, may 2010.

[132] David Sulzer, Myra P Joyce, Ling Lin, Daron Geldwert, Suzanne N. Haber, Toshiaki Hattori, Bibliography 166

and Stephen Rayport. Dopamine neurons make glutamatergic synapses in vitro. The Journal of neuroscience: the oﬃcial journal of the Society for Neuroscience, 18:4588–4602, 1998.

[133] Alberto Del Arco and Francisco Mora. Neurotransmitters and prefrontal cortex-limbic system interactions: implications for plasticity and psychiatric disorders. Journal of neural transmission, 116(8):941–52, aug 2009.

[134] Arvid Carlsson, Nicholas Waters, and Maria L Carlsson. Neurotransmitter Interactions in Schizophrenia - Therapeutic Implications. Biological Psychiatry, 46(10):1388–1395, 1999.

[135] Lawrence S. Kegeles, Anissa Abi-Dargham, Yolanda Zea-Ponce, Janine Rodenhiser-Hill, J. John Mann, Ronald L. Van Heertum, Thomas B. Cooper, Arvid Carlsson, and Marc Laruelle. Modula- tion of amphetamine-induced striatal dopamine release by ketamine in humans: Implications for schizophrenia. Biological Psychiatry, 48(7):627–640, 2000.

[136] Harvey S Singer, AL Reiss, JE Brown, EH Aylward, B Shih, E Chee, EL Harris, MJ Reader, GA Chase, RN Bryan, and MB Denckla. Volumetric MRI changes in basal ganglia of children with Tourette’s syndrome. Neurology, 43(5):950–950, may 1993.

[137] Cathy L Barr, Karen G Wigg, Andrew J Pakstis, Roger Kurlan, David Pauls, Kenneth K Kidd, Lap-Chee Tsui, and Paul Sandor. Genome scan for linkage to Gilles de la Tourette syndrome. American journal of medical genetics, 88(4):437–445, aug 1999.

[138] Abby Adamczyk, Colin D Gause, Rita Sattler, Svetlana Vidensky, Jeﬀery D Rothstein, Harvey Singer, and Tao Wang. Genetic and functional studies of a missense variant in a glutamate transporter, SLC1A3, in Tourette syndrome. Psychiatric genetics, 21(2):90–97, 2011.

[139] TSAICG. Genome scan for Tourette disorder in aﬀected-sibling-pair and multigenerational families. American journal of human genetics, 80(2):265–272, feb 2007.

[140] Paul D Arnold, David R Rosenberg, Emanuela Mundo, Subi Tharmalingam, James L Kennedy, and Margaret A Richter. Association of a glutamate (NMDA) subunit receptor gene (GRIN2B) with obsessive-compulsive disorder: a preliminary study. Psychopharmacology, 174:530–538, 2004.

[141] Paul D Arnold, Tricia Sicard, Eliza Burroughs, Margaret A Richter, and James L Kennedy. Gluta- mate transporter gene SLC1A1 associated with obsessive-compulsive disorder. Archives of general psychiatry, 63:769–776, 2006.

[142] Paul D Arnold, Frank P Macmaster, Margaret A Richter, Gregory L Hanna, Tricia Sicard, Eliza Burroughs, Yousha Mirza, Phillip C Easter, Michelle Rose, James L Kennedy, and David R Rosen- berg. Glutamate receptor gene (GRIN2B) associated with reduced anterior cingulate glutamatergic concentration in pediatric obsessive-compulsive disorder. Psychiatry research, 172:136–139, 2009.

[143] Timothy J DeVito, Dick J Drost, William Pavlosky, Richard W J Neufeld, Nagalingam Rajaku- mar, B Duncan McKinlay, Peter C Williamson, and Rob Nicolson. Brain magnetic resonance spectroscopy in Tourette’s disorder. Journal of the American Academy of Child and Adolescent Psychiatry, 44(12):1301–8, dec 2005.

[144] Meiyu Xu, Andrew Kobets, Jung-Chieh Du, Jessica Lennington, Lina Li, Mounira Banasr, Ronald S Duman, Flora M Vaccarino, Ralph J DiLeone, and Christopher Pittenger. Targeted ablation of cholinergic interneurons in the dorsolateral striatum produces behavioral manifestations of Tourette syndrome. Proceedings of the National Academy of Sciences of the United States of America, 112(3):893–8, 2015.

[145] D J Cohen, B A Shaywitz, B Caparulo, J G Young, and M B Bowers. Chronic, multiple tics of Gilles de la Tourette’s disease. CSF acid monoamine metabolites after probenecid administration. Archives of general psychiatry, 35:245–250, 1978.

[146] James F Leckman, Wayne K Goodman, George M Anderson, Mark A Riddle, Phillip B Chap- pell, Maureen T McSwiggan-Hardin, Christopher J McDougle, Lawrence D Scahill, Sharon I Ort, David L Pauls, Donald J Cohen, and Lawrence H Price. Cerebrospinal ﬂuid biogenic amines in obsessive compulsive disorder, tourette’s syndrome, and healthy controls. Neuropsychopharmacology, 12(1):73–86, 1995.

[147] Kirsten R Müller-Vahl, Geerd J Meyer, Wolfram H Knapp, Hinderk M Emrich, Peter Gielow, Bibliography 167

Thomas Brücke, and Georg Berding. Serotonin transporter binding in Tourette Syndrome. Neu- roscience letters, 385:120–125, 2005.

[148] A Heinz, M B Knable, S S Wolf, D W Jones, J G Gorey, T M Hyde, and D R Weinberger. Tourette’s syndrome: [I-123]beta-CIT SPECT correlates of vocal tic severity. Neurology, 51:1069–1074, 1998.

[149] Steven Haugbøl, Lars H Pinborg, Lisbeth Regeur, Elsebet S Hansen, Tom G Bolwig, Finn Å Nielsen, Claus Svarer, Lene T Skovgaard, and Gitte M Knudsen. Cerebral 5-HT(2A) receptor binding is increased in patients with Tourette’s syndrome. International Journal of Neuropsy- chopharmacology, 10(2):245 LP – 252, apr 2007.

[150] Lissandra Castellan Baldan, KyleA Williams, Jean Dominique Gallezot, Vladimir Pogorelov, Max- imiliano Rapanelli, Michael Crowley, GeorgeM Anderson, Erin Loring, Roxanne Gorczyca, Eileen Billingslea, Suzanne Wasylink, KaitlynE Panza, a. Gulhan Ercan-Sencicek, Kuakarun Krusong, BennettL Leventhal, Hiroshi Ohtsu, MichaelH Bloch, ZoëA Hughes, JohnH Krystal, Linda Mayes, Ivan DeAraujo, Yu Shin Ding, MatthewW State, and Christopher Pittenger. Histidine Decar- boxylase Deﬁciency Causes Tourette Syndrome: Parallel Findings in Humans and Mice. Neuron, 81(1):77–90, 2014.

[151] James Stankiewicz, S. Scott Panter, Mohit Neema, Ashish Arora, Courtney E. Batt, and Rohit Bakshi. Iron in Chronic Brain Disorders: Imaging and Neurotherapeutic Implications. Neurother- apeutics, 4(3):371–386, 2007.

[152] Domingo J Pinero and James R Connor. Iron in the Brain: An Important Contributor in Normal and Diseased States. The Neuroscientist, 6(6):435–453, dec 2000.

[153] John L Beard and James R Connor. Iron status and neural functioning. Annual review of nutrition, 23:41–58, jan 2003.

[154] Hossein SM Sadrzadeh and Yasi Saﬀari. Iron and Brain Disorders. Pathology Patterns Reviews, 121(1):65–70, jan 2004.

[155] James R Connor and Sharon L Menzies. Relationship of iron to oligodendrocytes and myelination. Glia, 17:83–93, 1996.

[156] D Ben-Shachar, R Ashkenazi, and M B Youdim. Long-term consequence of early iron-deﬁciency on dopaminergic neurotransmission in rats. International journal of developmental neuroscience: the oﬃcial journal of the International Society for Developmental Neuroscience, 4(1):81–8, jan 1986.

[157] John L Beard, Qing Chen, James Connor, and Byron C Jones. Altered monamine metabolism in caudate-putamen of iron-deﬁcient rats. Pharmacology Biochemistry and Behavior, 48(3):621–624, 1994.

[158] Christopher Nelson, Keith Erikson, Domingo J Piñero, and John L Beard. In vivo dopamine metabolism is altered in iron-deﬁcient anemic rats. The Journal of nutrition, 127(12):2282–8, 1997.

[159] K M Erikson, B C Jones, and J L Beard. Iron deﬁciency alters dopamine transporter functioning in rat striatum. The Journal of nutrition, 130(11):2831–7, nov 2000.

[160] Erica L Unger, Jason A Wiesinger, Lei Hao, and John L Beard. Dopamine D2 receptor expression is altered by changes in cellular iron levels in PC12 cells and rat brain tissue. The Journal of Nutrition, 138(12):2487–2494, 2008.

[161] Laura Bianco, Erica Unger, and John Beard. Iron Deﬁciency and Neuropharmacology, pages 141–158. Humana Press, Totowa, NJ, 2010.

[162] Daniel A Gorman, Hongtu Zhu, George M Anderson, Mark Davies, Bradley S Peterson, D Ph, George M Anderson, Mark Davies, and Bradley S Peterson. Ferritin Levels and Their Association With Regional Brain Volumes in Tourette ’ s Syndrome. The American journal of psychiatry, 163(7):1264–1272, jul 2006.

[163] Ling Liu, Zhi Gui Jiang, Wei Li, Hui Bing Liang, and Yan Lin. Epidemiological investigation of tic disorders among pupils in the Shunde Longjiang area, and their relationship to trace elements. Chinese Journal of Contemporary Pediatrics, 15(8):657–660, 2013. Bibliography 168

[164] RR Silva, DM Muñoz, W Daniel, J Barickman, and AJ Friedhoﬀ. Causes of haloperidol discontin- uation in patients with Tourette’s disorder: management and alternatives. The Journal of clinical psychiatry, 57:129–135, 1996.

[165] Yavuz S Silay and Joseph Jankovic. Emerging drugs in Tourette syndrome. Expert opinion on emerging drugs, 10(2):365–80, may 2005.

[166] Kevin D Burris, Thaddeus F Molski, Cen Xu, Elaine Ryan, Katsura Tottori, Tetsuro Kikuchi, Frank D Yocca, and Perry B Molinoﬀ. Aripiprazole, a novel antipsychotic, is a high-aﬃnity partial agonist at human dopamine D2 receptors. The Journal of pharmacology and experimental therapeutics, 302:381–389, 2002.

[167] Tanya K Murphy, P Jane Mutch, Jeannette M Reid, Paula J Edge, Eric A Storch, Michael Bengtson, and Mark Yang. Open label aripiprazole in the treatment of youth with tic disorders. Journal of Child and Adolescent Psychopharmacology, 19(4):441–447, 2009.

[168] Prasad R Padala, S Faiz Qadri, and Vishal Madaan. Aripiprazole for the treatment of Tourette’s disorder. Prim Care Companion J Clin Psychiatry, 2(6):297–300, 2005.

[169] Wan Seok Seo, Hyung-Mo Sung, Hyun Seok Sea, and Dai Seg Bai. Aripiprazole treatment of children and adolescents with Tourette disorder or chronic tic disorder. Journal of Child and Adolescent Psychopharmacology, 18(2):197–205, 2008.

[170] Claudia Wenzel, Alexandra Kleimann, Stefanie Bokemeyer, and Kirsten R Müller-Vahl. Arip- iprazole for the treatment of tourette syndrome: a case series of 100 patients. Journal of clinical psychopharmacology, 32:548–50, 2012.

[171] Andrea De Bartolomeis, Carmine Tomasetti, and Felice Iasevoli. Update on the Mechanism of Action of Aripiprazole: Translational Insights into Antipsychotic Strategies beyond Dopamine Receptor Antagonism. CNS Drugs, 29(9):773–799, 2015.

[172] Natalie J Forde, Ahmad S Kanaan, Joanna Widomska, Shanmukha S Padmanabhuni, Ester Ne- spoli, John Alexander, Juan I Rodriguez Arranz, Siyan Fan, Rayan Houssari, Muhammad S. Nawaz, Francesca Rizzo, Luca Pagliaroli, Nuno R. Zilhäo, Tamas Aranyi, Csaba Barta, Tobias M. Boeckers, Dorret I. Boomsma, Wim R. Buisman, Jan K Buitelaar, Danielle Cath, Andrea Diet- rich, Nicole Driessen, Petros Drineas, Michelle Dunlap, Sarah Gerasch, Jeﬀrey Glennon, Bastian Hengerer, Odile A van den Heuvel, Cathrine Jespersgaard, Harald E Möller, Kirsten R Müller- Vahl, Thaïra J. C. Openneer, Geert Poelmans, Petra J. W. Pouwels, Jeremiah M Scharf, Hreinn Stefansson, Zeynep Tümer, Dick J Veltman, Ysbrand D van der Werf, Pieter J Hoekstra, Andrea Ludolph, and Peristera Paschou. TS-EUROTRAIN: A European-Wide Investigation and Training Network on the Etiology and Pathophysiology of Gilles de la Tourette Syndrome. Frontiers in Neuroscience, 10(316978):1–9, 2016.

[173] David Sheehan, J Janavs, R Baker, K Harnett-Sheehan, E Knapp, M Sheehan, Y Lecrubier, E Weiller, T Hergueta, P Amorim, LI Bonora, and JP Lépine. M.I.N.I. International Neuropsy- chiatric Interview Version 5.0. Paris, 2006.

[174] James F Leckman, Mark A Riddle, Maureen T Hardin, Sharon I Ort, Karen L Swartz, John Stevenson, and Donald J Cohen. The Yale Global Tic Severity Scale: initial testing of a clinician- rated scale of tic severity. Journal of the American Academy of Child and Adolescent Psychiatry, 28(4):566–573, 1989.

[175] Christopher G Goetz, Eric J Pappert, Elan D Louis, Rema Raman, and Sue Leurgans. Advantages of a modiﬁed scoring method for the Rush video-based tic rating scale. Movement disorders: oﬃcial journal of the Movement Disorder Society, 14(3):502–506, 1999.

[176] Douglas W Woods, John Piacentini, Michael B Himle, and Susanna Chang. Premonitory Urge for Tics Scale (PUTS): initial psychometric results and examination of the premonitory urge phenomenon in youths with Tic disorders. Journal of developmental and behavioral pediatrics : JDBP, 26:397–403, 2005.

[177] Andrea E Cavanna, A Schrag, D Morley, M Orth, M M Robertson, E Joyce, H D Critchley, and C Selai. The Gilles de la Tourette syndrome-quality of life scale (GTS-QOL): development and validation. Neurology, 71:1410–1416, 2008. Bibliography 169

[178] Wayne K Goodman, Lawrence H Price, Steven A Rasmussen, Carolyn Mazure, Roberta L Fleis- chmann, Candy L Hill, George R Heninger, and Dennis S Charney. The Yale-Brown Obsessive Compulsive Scale. I. Development, use, and reliability., 1989.

[179] Stuart A Montgomery and Marie Asberg. A new depression scale designed to be sensitive to change. The British journal of psychiatry: the journal of mental science, 134:382–389, 1979.

[180] C. Keith Conners, Drew Erhardt, and Elizabeth Sparrow. Conners’ adult ADHD rating scales: technical manual. Multi-Health Systems, Park Ave., Toronto, ON, 1990.

[181] Aaron T Beck. Beck Depression Inventory. Depression, 2006:2–4, 1961.

[182] AaronT Beck, R A Steer, R Ball, and W Ranieri. Comparison of Beck Depression Inventories -IA and -II in psychiatric outpatients. Journal of personality assessment, 67(3):588–597, 1996.

[183] Edna B Foa, Jonathan D Huppert, Susanne Leiberg, Robert Langner, Rafael Kichic, Greg Hajcak, and Paul M Salkovskis. The Obsessive-Compulsive Inventory: development and validation of a short version. Psychological assessment, 14:485–496, 2002.

[184] Simon Baron-Cohen, Sally Wheelwright, Richard Skinner, Joanne Martin, and Emma Clubley. The autism-spectrum quotient (AQ): evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. Journal of autism and developmental disorders, 31(1):5–17, feb 2001.

[185] Donald McRobbie, Elizabeth A Moore, Martin J Graves, and Martin R Prince. MRI: From Picture to Proton. Cambridge University Press, New York, 2003.

[186] K Ranga Rama Krishnan and P Murali Doraiswamy. Brain imaging in clinical psychiatry. Marcel Dekker, New York, 1997.

[187] Thomas P Naidich, Mauricio Castillo, Soonmee Cha, and James G Smirniotopoulos. Imaging of the Brain: Expert Radiology Series. Expert Radiology. Elsevier Health Sciences, 2012.

[188] Charlotte Stagg and Douglas L Rothman. Magnetic Resonance Spectroscopy: Tools for Neuro- science Research and Emerging Clinical Applications. Academic Press, London, 2014.

[189] Peter Barker, Alberto Bizzi, Nicola Stefanp, Rao Gullapalli, and Doris D. M. Lin. Clinical MR Spectroscopy. Cambridge University Press, New York, 2010.

[190] Jamie Near. Spectral Quantiﬁcation and Pitfalls in Interpreting Magnetic Resonance Spectro- scopic Data: What To Look Out. In C Stagg and DL Rothman, editors, Magnetic Resonance Spectroscopy: Tools for Neuroscience Research and Emerging Clinical Applications, pages 49–67. Academic Press, San Diego, 2014.

[191] Stefan Blüml and Ashok Panigrahy. Magnetic Resonance Spectroscopy: Basics. In S. Blüml and A. Panigrahy, editors, MR Spectroscopy of Pediatric Brain Disorders, pages 11–23. Springer Science+Business Media, LLC, 2013.

[192] Andreas Deistung, Ferdinand Schweser, and Jürgen R Reichenbach. Overview of quantitative susceptibility mapping. NMR in Biomedicine, doi: 10.10, 2016.

[193] E Mark Haacke, S Mittal, Z Wu, Jaladhar Neelavalli, and Y-CN Cheng. Susceptibility-weighted imaging: technical aspects and clinical applications, part 1. American Journal of Neuroradiology, 30(1):19–30, 2009.

[194] Jürgen R Reichenbach and E Mark Haacke. Introduction to Susceptibility Weighted Imaging. In Susceptibility Weighted Imaging in MRI, pages 1–16. John Wiley & Sons, Inc., 2011.

[195] Chunlei Liu, Wei Li, Karen A. Tong, Kristen W. Yeom, and Samuel Kuzminski. Susceptibility- weighted imaging and quantitative susceptibility mapping in the brain. Journal of Magnetic Resonance Imaging, 42(1):23–41, 2015.

[196] Stefan K. Piechnik, J. Evans, L. H. Bary, R. G. Wise, and P. Jezzard. Functional changes in CSF volume estimated using measurement of water T2 relaxation. Magnetic Resonance in Medicine, 61(3):579–586, 2009.

[197] Greg J Stanisz, Ewa E Odrobina, Joseph Pun, Michael Escaravage, Simon J Graham, Michael J Bibliography 170

Bronskill, and R Mark Henkelman. T1, T2 relaxation and magnetization transfer in tissue at 3T. Magnetic Resonance in Medicine, 54(3):507–512, sep 2005.

[198] Robin A de Graaf. In Vivo NMR Spectroscopy: Principles and Techniques. Wiley, West Sussex, England, 2008.

[199] S W Provencher. Estimation of metabolite concentrations from localized in vivo proton NMR spectra. Magnetic Resonance in Medicine, 30(6):672–679, dec 1993.

[200] M L Simmons, C G Frondoza, and J T Coyle. Immunocytochemical localization of N-acetylaspartate with monoclonal antibodies. Neuroscience, 45(1):37–45, 1991.

[201] Kaja Nordengen, Christoph Heuser, Johanne E gge Rinholm, Reuben Matalon, and Vidar Gun- dersen. Localisation of N-acetylaspartate in oligodendrocytes/myelin. Brain structure & function, 220(2):899–917, 2015.

[202] John R Moﬀett, Prasanth Ariyannur, Peethambaran Arun, and Aryan M A Namboodiri. N- Acetylaspartate and N-Acetylaspartylglutamate in Central Nervous System Health and Disease. In Douglas B T Magnetic Resonance Spectroscopy Rothman, editor, Magnetic Resonance Spec- troscopy: Tools for Neuroscience Research and Emerging Clinical Applications, pages 71–90. Aca- demic Press, San Diego, 2014.

[203] Clare E Turner and Nicholas Gant. The Biochemistry of Creatine. In Douglas B T Magnetic Res- onance Spectroscopy Rothman, editor, Magnetic Resonance Spectroscopy: Tools for Neuroscience Research and Emerging Clinical Applications, pages 91–103. Academic Press, San Diego, 2014.

[204] Joanne C Lin and Nicholas Gant. The Biochemistry of Choline. In Douglas B T Magnetic Res- onance Spectroscopy Rothman, editor, Magnetic Resonance Spectroscopy: Tools for Neuroscience Research and Emerging Clinical Applications, pages 104–110. Academic Press, San Diego, 2014.

[205] Jonathan G Best, Charlotte J Stagg, and Andrea Dennis. Other Signiﬁcant Metabolites: Myo- Inositol, GABA, Glutamine, and Lactate. pages 122–138. Academic Press, San Diego, 2014.

[206] Jun Shen. Glutamate. In Douglas B T Magnetic Resonance Spectroscopy Rothman, editor, Magnetic Resonance Spectroscopy: Tools for Neuroscience Research and Emerging Clinical Appli- cations, pages 111–121. Academic Press, San Diego, 2014.

[207] E M Haacke and J R Reichenbach. Susceptibility Weighted Imaging in MRI: Basic Concepts and Clinical Applications. Wiley, 2011.

[208] Peter Jezzard and Stuart Clare. Sources of Distortions in Functional MRI Data. Human Brain Mapping, 8(May):80–85, 1999.

[209] Jürgen R Reichenbach, R Venkatesan, D J Schillinger, D K Kido, and E M Haacke. Small vessels in the human brain: MR venography with deoxyhemoglobin as an intrinsic contrast agent. Radiology, 204(1):272–277, jul 1997.

[210] HáKon Gudbjartsson and Samuel Patz. The rician distribution of noisy mri data. Magnetic Resonance in Medicine, 34(6):910–914, 1995.

[211] Alexander Rauscher, E Mark Haacke, Jaladhar Neelavalli, and Jürgen R Reichenbach. Phase and its Relationship to Imaging Parameters and Susceptibility. In Susceptibility Weighted Imaging in MRI, pages 47–71. John Wiley & Sons, Inc., 2011.

[212] Jürgen R Reichenbach and E Mark Haacke. Gradient Echo Imaging. In Susceptibility Weighted Imaging in MRI, pages 33–46. John Wiley & Sons, Inc., 2011.

[213] S Mittal, Z Wu, J Neelavalli, and E M Haacke. Susceptibility-weighted imaging: technical aspects and clinical applications, part 1. AJNR. American journal of neuroradiology, 30(2):232–52, feb 2009.

[214] Hongfu Sun and Alan H Wilman. Background ﬁeld removal using spherical mean value ﬁltering and Tikhonov regularization. Magnetic Resonance in Medicine, 71(3):1151–1157, 2014.

[215] Ferdinand Schweser, Andreas Deistung, Berengar Wendel Lehr, and Jürgen Rainer Reichenbach. Quantitative imaging of intrinsic magnetic tissue properties using MRI signal phase: An approach to in vivo brain iron metabolism? NeuroImage, 54(4):2789–2807, 2011. Bibliography 171

[216] Wei Li, Bing Wu, and Chunlei Liu. Quantitative susceptibility mapping of human brain reﬂects spatial variation in tissue composition. NeuroImage, 55(4):1645–1656, apr 2011.

[217] Tian Liu, Ildar Khalidov, Ludovic de Rochefort, Pascal Spincemaille, Jing Liu, A John Tsiouris, and Yi Wang. A novel background ﬁeld removal method for MRI using projection onto dipole ﬁelds (PDF). NMR in biomedicine, 24(9):1129–1136, nov 2011.

[218] Wei Li, Alexandru V Avram, Bing Wu, Xue Xiao, and Chunlei Liu. Integrated Laplacian-based phase unwrapping and background phase removal for quantitative susceptibility mapping. NMR in biomedicine, 27(2):219–227, feb 2014.

[219] Sina Straub, Till M. Schneider, Julian Emmerich, Martin T. Freitag, Christian H. Ziener, Heinz Pe- ter Schlemmer, Mark E. Ladd, and Frederik B. Laun. Suitable reference tissues for quantitative susceptibility mapping of the brain. Magnetic Resonance in Medicine, doi: 10.10, 2016.

[220] Chunlei Liu, Hongjiang Wei, Nan-jie Gong, Matthew Cronin, Russel Dibb, and Kyle Decker. Quantitative Susceptibility Mapping: Contrast Mechanisms and Clinical Applications. Tomogra- phy, 1(1):3–17, 2015.

[221] Chunlei Liu, Wei Li, G Allan Johnson, and Bing Wu. High-ﬁeld (9.4 T) MRI of brain dysmyeli- nation by quantitative mapping of magnetic susceptibility. NeuroImage, 56(3):930–938, jun 2011.

[222] E Mark Haacke, Norman Y C Cheng, Michael J House, Qiang Liu, Jaladhar Neelavalli, Robert J Ogg, Asadullah Khan, Muhammad Ayaz, Wolﬀ Kirsch, and Andre Obenaus. Imaging iron stores in the brain using magnetic resonance imaging. Magnetic Resonance Imaging, 23(1):1–25, jan 2005.

[223] John F. Schenck. The role of magnetic susceptibility in magnetic resonance imaging: MRI magnetic compatibility of the ﬁrst and second kinds. Medical Physics, 23(6):815, 1996.

[224] B Hallgren and P Sourander. The eﬀect of age on the non-haemin iron in the human Brain. Journal of Neurochemistry, 3(1):41–51, oct 1958.

[225] Christian Langkammer, Ferdinand Schweser, Nikolaus Krebs, Andreas Deistung, Walter Goessler, Eva Scheurer, Karsten Sommer, Gernot Reishofer, Kathrin Yen, Franz Fazekas, Stefan Ropele, and Jürgen R Reichenbach. Quantitative susceptibility mapping (QSM) as a means to measure brain iron? A post mortem validation study. NeuroImage, 62(3):1593–9, sep 2012.

[226] John F Schenck, C L Dumoulin, R W Redington, H Y Kressel, R T Elliott, and I L McDougall. Hu- man exposure to 4.0-Tesla magnetic ﬁelds in a whole-body scanner. Medical Physics, 19(4):1089– 1098, jul 1992.

[227] Hongfu Sun, Andrew J. Walsh, R. Marc Lebel, Gregg Blevins, Ingrid Catz, Jian Qiang Lu, Ed- ward S. Johnson, Derek J. Emery, Kenneth G. Warren, and Alan H. Wilman. Validation of quantitative susceptibility mapping with Perls’ iron staining for subcortical gray matter. NeuroImage, 2015.

[228] Weili Zheng, Helen Nichol, Saifeng Liu, Yu-Chung Norman Cheng, and E Mark Haacke. Measur- ing iron in the brain using quantitative susceptibility mapping and X-ray ﬂuorescence imaging. NeuroImage, 78:68–74, 2013.

[229] Carsten Stüber, Markus Morawski, Andreas Schäfer, Christian Labadie, Miriam Wähnert, Christoph Leuze, Markus Streicher, Nirav Barapatre, Katja Reimann, Stefan Geyer, Daniel Spe- mann, and Robert Turner. Myelin and iron concentration in the human brain: A quantitative study of MRI contrast. NeuroImage, 93(P1):95–106, 2014.

[230] Burton Drayer, Peter Burger, Robert Darwin, Stephen Riederer, Robert Herfkens, and G Al- lan Johnson. Magnetic Resonance Imaging of Brain Iron. American Journal of Neuroradiology, 7(3):373–380, may 1986.

[231] A. C. Smyth and S. T. Meier. Evaluating the Psychometric Properties of the Conners Adult ADHD Rating Scales. Journal of Attention Disorders, doi: 10.11, 2016.

[232] Abigail Taylor, Shoumitro Deb, and Gemma Unwin. Scales for the identiﬁcation of adults with attention deﬁcit hyperactivity disorder (ADHD): A systematic review. Research in Developmental Disabilities, 32(3):924–938, 2011. Bibliography 172

[233] M. Rösler, W. Retz, P. Retz-Junginger, J. Thome, T. Supprian, T. Nissen, R. D. Stieglitz, D. Blocher, G. Hengesch, and G. E. Trott. Instrumente zur diagnostik der aufmerksamkeitsdeﬁzit- / hyperaktivitätsstörung (ADHS) im erwachsenenalter. Selbstbeurteilungsskala (ADHS-SB) und diagnosecheckliste (ADHS-DC). Nervenarzt, 75(9):888–895, 2004.

[234] P Retz-Junginger, W Retz, D Blocher, R.-D. Stieglitz, T Georg, T Supprian, P H Wender, and M Rösler. Reliabilität und Validität der Wender-Utah-Rating-Scale-Kurzform. Der Nervenarzt, 74(11):987–993, 2003.

[235] Mark F Ward, Paul H Wender, and Fred W Reimherr. The Wender Utah Rating Scale: an aid in the retrospective diagnosis of childhood attention deﬁcit hyperactivity disorder. American Journal of Psychiatry, 150(6):885–890, jun 1993.

[236] C. Kühner, C. Bürger, F. Keller, and M. Hautzinger. Reliabilität und validität des revidierten Beck- Depressionsinventars (BDI-II). Befunde aus deutschsprachigen stichproben. Nervenarzt, 78(6):651–656, 2007.

[237] R. Van Der Lem, N. J A Van Der Wee, T. Van Veen, and F. G. Zitman. Generaliseerbaarheid van depressietrials naar de dagelijkse praktijk. Tijdschrift voor Psychiatrie, 57(8):579–587, 2015.

[238] Martijn S. Van Noorden, Esther M. Van Fenema, Nic J A Van Der Wee, Frans G. Zitman, and Erik J. Giltay. Predicting outcome of depression using the depressive symptom proﬁle: The Leiden routine outcome monitoring study. Depression and Anxiety, 29(6):523–530, 2012.

[239] R Uher, a Farmer, W Maier, M Rietschel, J Hauser, a Marusic, O Mors, a Elkin, R J Williamson, C Schmael, N Henigsberg, J Perez, J Mendlewicz, J G E Janzing, a Zobel, M Skibinska, D Kozel, a S Stamp, M Bajs, a Placentino, M Barreto, P McGuﬃn, and K J Aitchison. Measuring depression: comparison and integration of three scales in the GENDEP study. Psychological medicine, 38(2):289–300, 2008.

[240] Aaron T Beck, Gary Brown, Norman Epstein, and Robert A Steer. An Inventory for Measur- ing Clinical Anxiety: Psychometric Properties. Journal of Consulting and Clinical Psychology, 56(6):893–897, 1988.

[241] J. Margraf and A. Ehlers. Beck Angst-Inventar. Harcourt Test Services GmbH, Frankfurt am Main, 2007.

[242] Tina Phan, Owen Carter, Claire Adams, Grant Waterer, Li Ping Chung, Maxine Hawkins, Co- bie Rudd, Mel Ziman, and Natalie Strobel. Discriminant validity of the Hospital Anxiety and Depression Scale, Beck Depression Inventory (II) and Beck Anxiety Inventory to conﬁrmed clinical diagnosis of depression and anxiety in patients with chronic obstructive pulmonary disease. Chronic respiratory disease, 13(3):220–8, 2016.

[243] Timm Wetzel, Marc Tittgemeyer, Alfred Anwander, and Harald E. Möller. Investigation of white- matter aging by quantitative MRI and MRS. In Proceedings of the 13th Annual Meeting of ISMRM, page 1164, 2005.

[244] AS Kanaan, Andre Pampel, Kirsten R. Müller-Vahl, and Harald E Möller. Test-Retest reliability of absolute metabolite concentrations with partial volume correction using diﬀerent segmentation methods. In Proceedings of the 23rd Annual Meeting of ISMRM, Toronto, ON, Canada, 2015.

[245] Thomas Ernst, Roland Kreis, and BD Ross. Absolute Quantitation of Water and Metabolites in the Human Brain. I. Compartments and Water. Journal of Magnetic Resonance, Series B, 102(1):1–8, 1993.

[246] Vladimir Mlynárik, Stephan Gruber, and Ewald Moser. Proton T1 and T2 relaxation times of human brain metabolites at 3 Tesla. NMR in biomedicine, 14(5):325–331, 2001.

[247] Frederick Klauschen, Aaron Goldman, Vincent Barra, Andreas Meyer-Lindenberg, and Arvid Lundervold. Evaluation of automated brain MR image segmentation and volumetry methods. Human brain mapping, 30(4):1310–27, apr 2009.

[248] Charles Gasparovic, Tao Song, Deidre Devier, H Jeremy Bockholt, Arvind Caprihan, Paul G Mullins, Stefan Posse, Rex E Jung, and Leslie a Morrison. Use of tissue water as a concentration reference for proton spectroscopic imaging. Magnetic Resonance in Medicine, 55(6):1219–26, jul 2006. Bibliography 173

[249] Weiqiang Dou, Oliver Speck, Thomas Benner, Jörn Kaufmann, Meng Li, Kai Zhong, and Martin Walter. Automatic voxel positioning for MRS at 7 T. Magnetic Resonance Materials in Physics, Biology and Medicine, 28:259–270, 2015.

[250] Hedok Lee, Elisabeth Caparelli, Haifang Li, Amit Mandal, S David Smith, Shaonan Zhang, Thomas V Bilﬁnger, and Helene Benveniste. Computerized MRS voxel registration and partial volume eﬀects in single voxel 1H-MRS. Magnetic resonance imaging, 31(7):1197–1205, sep 2013.

[251] Richard AEE Edden, Nicolaas AJJ Puts, Ashley D Harris, Peter B Barker, and C. John Evans. Gannet: A batch-processing tool for the quantitative analysis of gamma-aminobutyric acid-edited MR spectroscopy spectra. Journal of Magnetic Resonance Imaging, 40(6):1445–1452, nov 2014.

[252] Lee R . Dice. Measures of the Amount of Ecologic Association Between Species. Ecology, 26(3):297– 302, 1945.

[253] Thomas Sørensen. A method of establishing groups of equal amplitude in plant sociology based on similarity of species and its application to analyses of the vegetation on Danish commons. Biol. Skr., 5:1–34, 1948.

[254] Alexander Gussew, Marko Erdtel, Patrick Hiepe, Reinhard Rzanny, and Jürgen R Reichenbach. Absolute quantitation of brain metabolites with respect to heterogeneous tissue compositions in 1H-MR spectroscopic volumes. Magnetic Resonance Materials in Physics, Biology and Medicine, 25(5):321–333, oct 2012.

[255] Martin H J Busch, Wolfgang Vollmann, Serban Mateiescu, Manuel Stolze, Martin Deli, Mari- etta Garmer, and Dietrich H W Grönemeyer. Reproducibility of brain metabolite concentration measurements in lesion free white matter at 1.5T. BMC Medical Imaging, 15:1–13, 2015.

[256] Riccardo Metere, Ahmad Seif Kanaan, Berkin Bilgic, Torsten Schlumm, and Harald E. Möller. Eﬀects of coil combination algorithms on Quantitative Susceptibility Mapping. In Proceedings of the 25th Annual Meeting of ISMRM, page 2433, 2017.

[257] Christian Langkammer, Kristian Bredies, Benedikt A Poser, Markus Barth, Gernot Reishofer, Audrey Peiwen Fan, Berkin Bilgic, Franz Fazekas, Caterina Mainero, and Stefan Ropele. Fast quantitative susceptibility mapping using 3D EPI and total generalized variation. NeuroImage, 111:622–630, may 2015.

[258] Andreas Schäfer, Sam Wharton, Penny Gowland, and Richard Bowtell. Using magnetic ﬁeld simulation to study susceptibility-related phase contrast in gradient echo MRI. NeuroImage, 48(1):126–137, oct 2009.

[259] Peter B Roemer, WA Edelstein, CE Hayes, SP Souza, and OM Mueller. The NMR phased array. Magnetic Resonance in Medicine, 16(2):192–225, nov 1990.

[260] Klaas P Pruessmann, Markus Weiger, Markus B Scheidegger, and Peter Boesiger. SENSE: Sensi- tivity encoding for fast MRI. Magnetic Resonance in Medicine, 42(5):952–962, nov 1999.

[261] Mark A Griswold, Peter M Jakob, Robin M Heidemann, Mathias Nittka, Vladimir Jellus, Jianmin Wang, Berthold Kiefer, and Axel Haase. Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magnetic Resonance in Medicine, 47(6):1202–1210, jun 2002.

[262] Martin Uecker, Peng Lai, Mark J. Murphy, Patrick Virtue, Michael Elad, John M. Pauly, Shreyas S. Vasanawala, and Michael Lustig. ESPIRiT - An eigenvalue approach to autocalibrating parallel MRI: Where SENSE meets GRAPPA. Magnetic Resonance in Medicine, 71(3):990–1001, 2014.

[263] David O Walsh, Arthur F Gmitro, and Michael W Marcellin. Adaptive reconstruction of phased array MR imagery. Magnetic Resonance in Medicine, 43(5):682–690, may 2000.

[264] Derya Gol Gungor and Lee C Potter. A subspace-based coil combination method for phased-array magnetic resonance imaging. Magnetic Resonance in Medicine, 75(2):762–774, feb 2016.

[265] Dennis L Parker, Allison Payne, Nick Todd, and J Rock Hadley. Phase reconstruction from multiple coil data using a virtual reference coil. Magnetic Resonance in Medicine, 72(2):563–569, aug 2014.

[266] Simon Robinson, Günther Grabner, Stephan Witoszynskyj, and Siegfried Trattnig. Combining Bibliography 174

phase images from multi-channel {RF} coils using 3D phase oﬀset maps derived from a dual-echo scan. Magnetic Resonance in Medicine, 65(6):1638–1648, jun 2011.

[267] Simon Daniel Robinson, Kristian Bredies, Diana Khabipova, Barbara Dymerska, José P Marques, and Ferdinand Schweser. An illustrated comparison of processing methods for MR phase imaging and QSM: combining array coil signals and phase unwrapping. NMR in Biomedicine, 10.1002/nb, jan 2016.

[268] E Mark Haacke, Saifeng Liu, Sagar Buch, Weili Zheng, Dongmei Wu, and Yongquan Ye. Quanti- tative susceptibility mapping: current status and future directions. Magnetic Resonance Imaging, 33(1):1–25, jan 2015.

[269] Martin Uecker and Michael Lustig. Estimating absolute-phase maps using ESPIRiT and virtual conjugate coils. Magnetic Resonance in Medicine, 00, 2016.

[270] Berkin Bilgic, Jonathan R Polimeni, Lawrence L Wald, Kawin Setsompop, and United States. Automated tissue phase and QSM estimation from multichannel data. Proc Intl Soc Mag Reson Med 24 (2016), (2849), 2016.

[271] José P Marques, Tobias Kober, Gunnar Krueger, Wietske van der Zwaag, Pierre-François Van de Moortele, and Rolf Gruetter. MP2RAGE, a self bias-ﬁeld corrected sequence for improved segmentation and T1-mapping at high ﬁeld. NeuroImage, 49(2):1271–81, jan 2010.

[272] Martin Uecker, Frank Ong, Jonathan I Tamir, Dara Bahri, Patrick Virtue, Joseph Y Cheng, Tao Zhang, and Michael Lustig. Berkeley Advanced Reconstruction Toolbox. In Proceedings of the {ISMRM} 23rd {Annual} {Meeting} & {Exhibition}, Toronto, Canada, may 2015.

[273] Ferdinand Schweser, Andreas Deistung, Karsten Sommer, and Jürgen Rainer Reichenbach. Toward online reconstruction of quantitative susceptibility maps: Superfast dipole inversion. Magnetic Resonance in Medicine, 69(6):1582–1594, 2013.

[274] Ahmad Seif Kanaan, Sarah Gerasch, Isabel García-García, Leonie Lampe, André Pampel, Alfred Anwander, Jamie Near, Harald E Möller, and Kirsten Müller-Vahl. Pathological glutamatergic neurotransmission in Gilles de la Tourette syndrome. Brain: a journal of neurology, 140(1):218– 234, 2017.

[275] R L Albin, A B Young, and J B Penney. The functional anatomy of basal ganglia disorders. Trends in neurosciences, 12(10):366–375, 1989.

[276] Marjan Jahanshahi, Ignacio Obeso, John C. Rothwell, and José A. Obeso. A fronto-striato- subthalamic-pallidal network for goal-directed and habitual inhibition. Nature reviews. Neuro- science, 16(12):719–32, 2015.

[277] Wolfram Schultz, Peter Dayan, and P Read Montague. A Neural Substrate of Prediction and Reward. Science, 275(5306):1593–1599, mar 1997.

[278] Kent C. Berridge and Terry E. Robinson. What is the role of dopamine in reward: Hedonic impact, reward learning, or incentive salience? Brain Research Reviews, 28(3):309–369, 1998.

[279] Xin Jin and Rui M Costa. Start/stop signals emerge in nigrostriatal circuits during sequence learning. Nature, 466(7305):457–462, 2010.

[280] Cécile Delorme, Alexandre Salvador, Romain Valabrègue, Emmanuel Roze, Stefano Palminteri, Marie Vidailhet, Sanne de Wit, Trevor Robbins, Andreas Hartmann, and Yulia Worbe. Enhanced habit formation in Gilles de la Tourette syndrome. Brain: a journal of neurology, 139(2):605–615, jan 2016.

[281] Stefano Palminteri, Maël Lebreton, Yulia Worbe, Andreas Hartmann, Stéphane Lehéricy, Marie Vidailhet, David Grabli, and Mathias Pessiglione. Dopamine-dependent reinforcement of motor skill learning: Evidence from Gilles de la Tourette syndrome. Brain: a journal of neurology, 134(8):2287–2301, 2011.

[282] Yulia Worbe, Stefano Palminteri, Andreas Hartmann, Marie Vidailhet, Stéphane Lehéricy, and Mathias Pessiglione. Reinforcement Learning and Gilles de la Tourette Syndrome. Archives of General Psychiatry, 68(12):1257–1266, 2011. Bibliography 175

[283] Anthony A Grace. Phasic versus tonic dopamine release and the modulation of dopamine system responsivity: A hypothesis for the etiology of schizophrenia. Neuroscience, 41(1):1–24, 1991.

[284] Anthony A Grace. Cortical regulation of subcortical dopamine systems and its possible relevance to schizophrenia. Journal of Neural Transmission, 91(2-3):111–134, 1993.

[285] Helle Sønderby Waagepetersen, U Sonnewald, and Arne Schousboe. Glutamine, Glutamate, and GABA: Metabolic Aspects. In Handbook of neurochemistry and molecular neurobiology, pages 1–21. Springer US, New York, 2007.

[286] Sule Tinaz, Beth A. Belluscio, Patrick Malone, Jan Willem van der Veen, Mark Hallett, and Silvina G. Horovitz. Role of the Sensorimotor Cortex in Tourette Syndrome using Multimodal Imaging. Human Brain Mapping, 35(12):5834–5846, 2014.

[287] Christiaan de Leeuw, Andrea Goudriaan, August B Smit, Dongmei Yu, Carol a Mathews, Jeremiah M Scharf, J M Scharf, D L Pauls, D Yu, C Illmann, L Osiecki, B M Neale, C a Mathews, V I Reus, T L Lowe, N B Freimer, N J Cox, L K Davis, G a Rouleau, S Chouinard, Y Dion, S Gi- rard, D C Cath, Danielle Posthuma, J H Smit, P Heutink, R a King, T Fernandez, J F Leckman, P Sandor, C L Barr, W McMahon, G Lyon, M Leppert, J Morgan, R Weiss, M a Grados, H Singer, J Jankovic, J a Tischﬁeld, G a Heiman, Mark H G Verheijen, and Danielle Posthuma. Involvement of astrocyte metabolic coupling in Tourette syndrome pathogenesis. European Journal of Human Genetics, 23(August 2014):1–4, 2015.

[288] Jessica B Lennington, Gianﬁlippo Coppola, Yuko Kataoka-Sasaki, Thomas V Fernandez, Dean Palejev, Yifan Li, Anita Huttner, Mihovil Pletikos, Nenad Sestan, James F Leckman, and Flora M Vaccarino. Transcriptome Analysis of the Human Striatum in Tourette Syndrome. Biological Psychiatry, 79(5):372–382, mar 2016.

[289] Sabine Krabbe, Johanna Duda, Julia Schiemann, Christina Poetschke, Gaby Schneider, Eric R Kandel, Birgit Liss, Jochen Roeper, and Eleanor H Simpson. Increased dopamine D2 receptor activity in the striatum alters the ﬁring pattern of dopamine neurons in the ventral tegmental area. Proceedings of the National Academy of Sciences of the United States of America, 112(12):E1498– 506, 2015.

[290] Noam Soreni, Michael D Noseworthy, Toni Cormier, Wendy K Oakden, Sonya Bells, and Russell Schachar. Intraindividual variability of striatal 1H-MRS brain metabolite measurements at 3 T. Magnetic resonance imaging, 24(2):187–194, feb 2006.

[291] Daniel-Paolo Streitbürger, André Pampel, Gunnar Krueger, Jöran Lepsien, Matthias L Schroeter, Karsten Mueller, and Harald E Möller. Impact of image acquisition on voxel-based-morphometry investigations of age-related structural brain changes. NeuroImage, 87C:170–182, nov 2013.

[292] Paul A Bottomley. Selective volume method for performing localized NMR spectroscopy. US patent 4,480,228, 1984.

[293] Rolf Gruetter and Ivan Tkác. Field mapping without reference scan using asymmetric echo-planar techniques. Magnetic Resonance in Medicine, 43(2):319–323, 2000.

[294] Rolf Gruetter. Automatic, localized in Vivo adjustment of all ﬁrst-and second-order shim coils. Magnetic Resonance in Medicine, 29:804–811, 1993.

[295] Jamie Near, Richard Edden, C. John Evans, Raphael Paquin, Ashley Harris, and Peter Jezzard. Frequency and phase drift correction of magnetic resonance spectroscopy data by spectral registration in the time domain. Magnetic Resonance in Medicine, 73(1):44–50, jan 2015.

[296] Robin Simpson, Gabriel A. Devenyi, Peter Jezzard, T. Jay Hennessy, and Jamie Near. Advanced Processing and Simulation of MRS Data Using the FID Appliance (FID-A)âĂŤAn Open Source, MATLAB-Based Toolkit. Magnetic Resonance in Medicine, 2016.

[297] Brian Patenaude, Stephen M. Smith, David N. Kennedy, and Mark Jenkinson. A Bayesian model of shape and appearance for subcortical brain segmentation. NeuroImage, 56(3):907–922, 2011.

[298] Ivan Tkác, Gülin Oz, Gregor Adriany, Kamil UÄ§urbil, and Rolf Gruetter. In vivo 1H NMR spectroscopy of the human brain at high magnetic ﬁelds: metabolite quantiﬁcation at 4T vs. 7T. Magnetic Resonance in Medicine, 62(4):868–79, oct 2009. Bibliography 176

[299] Roland Kreis. The trouble with quality ﬁltering based on relative Cramér-Rao lower bounds. Magnetic Resonance in Medicine, 75:15–18, 2016.

[300] Chen Lin, Matt Bernstein, John Huston, and Sean Fain. Measurements of T1 relaxation times at 3.0 T: implications for clinical MRA. In Proc. Intl. Soc. Mag. Reson. Med 9, 1391, 2001.

[301] Changho Choi, Nicholas J Coupland, Paramjit P Bhardwaj, Sanjay Kalra, Colin a Casault, Kim Reid, and Peter S Allen. T2 measurement and quantiﬁcation of glutamate in human brain in vivo. Magnetic Resonance in Medicine, 56(5):971–977, nov 2006.

[302] William T Norton, Shirley E Poduslo, and Kunihiko Suzuki. Subacute sclerosing leukoencephalitis. II. Chemical studies including abnormal myelin and an abnormal ganglioside pattern. Journal of Neuropathology & Experimental Neurology, 25(4):582–597, oct 1966.

[303] Travis E Oliphant. SciPy: Open source scientiﬁc tools for Python. Computing in Science and Engineering, 9:10–20, 2007.

[304] Skipper Seabold and Josef Perktold. Statsmodels: econometric and statistical modeling with python. In Proceedings of the 9th Python in Science Conference, number Scipy, pages 57–61, 2010.

[305] Mark Jenkinson, Peter Bannister, Michael Brady, and Stephen Smith. Improved optimization for the robust and accurate linear registration and motion correction of brain images. NeuroImage, 17(2):825–841, 2002.

[306] Gunther Helms and Andreas Piringer. Restoration of motion-related signal loss and line-shape deterioration of proton MR spectra using the residual water as intrinsic reference. Magnetic Resonance in Medicine, 46(2):395–400, 2001.

[307] Hoby P Hetherington, Jullie W Pan, Graeme F Mason, Dorothy Adams, Michael J Vaughn, Donald B Twieg, and Gerald M Pohost. Quantitative H-1 spectroscopic imaging of human brain at 4.1 T using image segmentation. Magnetic Resonance in Medicine, 36(1):21–29, 1996.

[308] H. P. Hetherington, G. F. Mason, J. W. Pan, S. L. Ponder, J. T. Vaughan, D. B. Twieg, and G. M. Pohost. Evaluation of cerebral gray and white matter metabolite diﬀerences by spectroscopic imaging at 4.1T. Magnetic Resonance in Medicine, 32(5):565–571, 1994.

[309] Norbert Schuﬀ, Frank Ezekiel, Anthony C Gamst, Diane L Amend, Andres A Capizzano, An- drew A Maudsley, and M W Weiner. Region and tissue diﬀerences of metabolites in normally aged brain using multislice 1H magnetic resonance spectroscopic imaging. Magnetic Resonance in Medicine, 45(5):899–907, 2001.

[310] Joel Lavoie, Jean-Francois Giguère, Gilles Pomier Layrargues, and Roger F Butterworth. Amino acid changes in autopsied brain tissue from cirrhotic patients with hepatic encephalopathy. Journal of neurochemistry, 49(3):692–697, 1987.

[311] TL Perry, S Berry, S Hansen, S Diamond, and C Mok. Regional Distiubution Of Amino Acids In Human Brain Obtained At Autopsy. Journal of neurochemistry, 18:513–519, 1971.

[312] Saadallah Ramadan, Alexander Lin, and Peter Stanwell. Glutamate and glutamine: A review of in vivo MRS in the human brain, 2013.

[313] Ileana Hancu. Optimized glutamate detection at 3T. Journal of Magnetic Resonance Imaging, 30(5):1155–62, nov 2009.

[314] Paul Gerald Mullins, Hongji Chen, Jing Xu, Arvind Caprihan, and Charles Gasparovic. Com- parative reliability of proton spectroscopy techniques designed to improve detection of J-coupled metabolites. Magnetic Resonance in Medicine, 60(4):964–969, 2008.

[315] Paul G. Mullins, David J. McGonigle, Ruth L. O’Gorman, Nicolaas A J Puts, Rishma Vidyasagar, C. John Evans, Richard A E Edden, Matthew J. Brookes, Adrian Garcia, Bradley R. Foerster, Myria Petrou, Darren Price, Bhavana S. Solanky, Inês R. Violante, Steve Williams, and Martin Wilson. Current practice in the use of MEGA-PRESS spectroscopy for the detection of GABA. NeuroImage, 86:43–52, 2014.

[316] Ruth L. O’Gorman, Lars Michels, Richard A. Edden, James B. Murdoch, and Ernst Martin. In vivo detection of GABA and glutamate with MEGA-PRESS: Reproducibility and gender eﬀects. Bibliography 177

Journal of Magnetic Resonance Imaging, 33(5):1262–1267, 2011.

[317] Richard J Maddock and Michael H Buonocore. Comment regarding increased striatal glutamate in schizophrenia. Neuropsychopharmacology : oﬃcial publication of the American College of Neu- ropsychopharmacology, 37(4):1067–8; author reply 1069, mar 2012.

[318] Helle S Waagepetersen, Ursula Sonnewald, and Arne Schousboe. Compartmentation of glutamine, glutamate, and GABA metabolism in neurons and astrocytes: functional implications. The Neu- roscientist: a review journal bringing neurobiology, neurology and psychiatry, 9(5):398–403, 2003.

[319] Douglas L. Rothman, Kevin L. Behar, Fahmeed Hyder, and Robert G. Shulman. In vivo NMR studies of the glutamate neurotransmitter ﬂux and neuroenergetics: implications for brain function. Annual review of physiology, 65(1):401–427, 2003.

[320] Nicola R Sibson, A Dhankhar, GF Mason, KL Behar, DL Rothman, and RG Shulman. In vivo 13C NMR measurements of cerebral glutamine synthesis as evidence for glutamate-glutamine cycling. Proceedings of the National Academy of Sciences of the United States of America, 94(6):2699–2704, 1997.

[321] Nicola R Sibson, Ajay Dhankhar, Graeme F Mason, Douglas L Rothman, Kevin L Behar, and Robert G Shulman. Stoichiometric coupling of brain glucose metabolism and glutamatergic neuronal activity. Proceedings of the National Academy of Sciences of the United States of America, 95(January):316–321, 1998.

[322] Hiroaki Tani, Chris G. Dulla, Zoya Farzampour, Amaro Taylor-Weiner, John R. Huguenard, and Richard J. Reimer. A local glutamate-glutamine cycle sustains synaptic excitatory transmitter release. Neuron, 81(4):888–900, 2014.

[323] Jun Shen. Glutamate. In Charlotte Stagg and Douglas L. Rothman, editors, Magnetic Resonance Spectroscopy: Tools for Neuroscience Research and Emerging Clinical Applications, pages 111–121. Academic Press, 2013.

[324] Jonathan G. Best, Charlotte J. Stagg, and Andrea Dennis. Other Signiﬁcant Metabolites: Myo- Inositol, GABA, Glutamine, and Lactate. In Charlotte Stagg and Douglas L. Rothman, editors, Magnetic Resonance Spectroscopy: Tools for Neuroscience Research and Emerging Clinical Appli- cations, pages 122–138. Academic Press, 2013.

[325] Ileana Hancu and John Port. The case of the missing glutamine. NMR in Biomedicine, 24(5):529– 535, 2011.

[326] Nivedita Agarwal and P. F. Renshaw. Proton MR spectroscopy - Detectable major neurotransmitters of the brain: Biology and possible clinical applications. American Journal of Neuroradiology, 33(4):595–602, 2012.

[327] Ricardo C G Landim, Richard A E Edden, Bernd Foerster, Li Min Li, Roberto J M Covolan, and Gabriela Castellano. Investigation of NAA and NAAG dynamics underlying visual stimulation using MEGA-PRESS in a functional MRS experiment. Magnetic Resonance Imaging, 34(3):239– 245, 2016.

[328] Anant B Patel, Robin a de Graaf, Graeme F Mason, Douglas L Rothman, Robert G Shulman, and Kevin L Behar. The contribution of GABA to glutamate/glutamine cycling and energy metabolism in the rat cortex in vivo. Proceedings of the National Academy of Sciences of the United States of America, 102(15):5588–5593, 2005.

[329] Nicolas Audet, Elodie Archer-Lahlou, Mélissa Richard-Lalonde, and Graciela Piñeyro-Filpo. Func- tional selectivity: theoretical consideration and future direction. In KA Neve, editor, Functional Selectivity of G Protein-Coupled Receptor Ligands: New Opportunities for Drug Discovery, pages 9–24. Humana Press, New York, 2009.

[330] Richard Mailman, Yan-Min Wang, Andrew Kant, and Justin Brown. Functional selectivity at dopamine deceptors. In Kim Neve, editor, Functional Selectivity of G Protein-Coupled Receptor Ligands: New Opportunities for Drug Discovery, pages 177–209. Humana Press, New York, 2009.

[331] Takashi Hamamura and Toshiki Harada. Unique pharmacological proﬁle of aripiprazole as the phasic component buster. Psychopharmacology, 191(3):741–743, 2007. Bibliography 178

[332] Guo Fen Ma, Noora Raivio, Josefa Sabria, and Jordi Ortiz. Agonist and Antagonist Eﬀects of Aripiprazole on D2-Like Receptors Controlling Rat Brain Dopamine Synthesis Depend on the Dopaminergic Tone. International Journal of Neuropsychopharmacology, 18(4):pyu046, 2015.

[333] Nina Segnitz, Thomas Ferbert, Andrea Schmitt, Peter Gass, Peter J. Gebicke-Haerter, and Mathias Zink. Eﬀects of chronic oral treatment with aripiprazole on the expression of NMDA receptor subunits and binding sites in rat brain. Psychopharmacology, 217(1):127–142, 2011.

[334] Nina Segnitz, Andrea Schmitt, Peter J. Gebicke-Härter, and Mathias Zink. Diﬀerential expression of glutamate transporter genes after chronic oral treatment with aripiprazole in rats. Neurochem- istry International, 55(7):619–628, 2009.

[335] Nina Peselmann, Andrea Schmitt, Peter J. Gebicke-Haerter, and Mathias Zink. Aripiprazole diﬀerentially regulates the expression of Gad67 and γ-aminobutyric acid transporters in rat brain. European Archives of Psychiatry and Clinical Neuroscience, 263(4):285–297, 2013.

[336] Monica E. Lemmon, Marco Grados, Tina Kline, Carol B. Thompson, Syed F. Ali, and Harvey S. Singer. Eﬃcacy of glutamate modulators in tic suppression: A double-blind, randomized control trial of D-serine and riluzole in Tourette Syndrome. Pediatric Neurology, 52(6):629–634, 2015.

[337] J Wayne Aldridge and Kent C Berridge. Coding of serial order by neostriatal neurons: a "natural action" approach to movement sequence. The Journal of neuroscience: the oﬃcial journal of the Society for Neuroscience, 18(7):2777–2787, 1998.

[338] Kent C Berridge, J Wayne Aldridge, Kimberly R Houchard, and Xiaoxi Zhuang. Sequential super- stereotypy of an instinctive ﬁxed action pattern in hyper-dopaminergic mutant mice: a model of obsessive compulsive disorder and Tourette’s. BMC biology, 3(1):4, 2005.

[339] Howard C Cromwell and Kent C Berridge. Implementation of action sequences by a neostriatal site: a lesion mapping study of grooming syntax. The Journal of neuroscience: the oﬃcial journal of the Society for Neuroscience, 16(10):3444–3458, 1996.

[340] David V. Smith, Amanda V. Utevsky, Amy R. Bland, Nathan Clement, John a. Clithero, Anne E W Harsch, R. McKell Carter, and Scott a. Huettel. Characterizing individual diﬀerences in functional connectivity using dual-regression and seed-based approaches. NeuroImage, 95:1–12, 2014.

[341] Daniel F English, Osvaldo Ibanez-Sandoval, Eran Stark, Fatuel Tecuapetla, György Buzsáki, Karl Deisseroth, James M Tepper, and Tibor Koos. GABAergic circuits mediate the reinforcement- related signals of striatal cholinergic interneurons. Nature Neuroscience, 15(1):123–130, 2011.

[342] Christoph Straub, Nicolas X. Tritsch, Nellwyn A. Hagan, Chenghua Gu, and Bernardo L. Sabatini. Multiphasic modulation of cholinergic interneurons by nigrostriatal aﬀerents. The Journal of neuroscience: the oﬃcial journal of the Society for Neuroscience, 34(25):8557–69, 2014.

[343] Mitsuko Watabe-Uchida, Lisa Zhu, Sachie K. Ogawa, Archana Vamanrao, and Naoshige Uchida. Whole-Brain Mapping of Direct Inputs to Midbrain Dopamine Neurons. Neuron, 74(5):858–873, 2012.

[344] Lei Phillip Wang, Fei Li, Dong Wang, Kun Xie, Deheng Wang, Xiaoming Shen, and Joe Z. Tsien. NMDA Receptors in Dopaminergic Neurons are Crucial for Habit Learning. Neuron, 72(6):1055– 1066, 2011.

[345] Brian Keating and Thomas Ernst. Real-time dynamic frequency and shim correction for single- voxel magnetic resonance spectroscopy. Magnetic Resonance in Medicine, 68(5):1339–1345, 2012.

[346] Jessica Schulz, Thomas Siegert, Enrico Reimer, Christian Labadie, Julian Maclaren, Michael Herbst, Maxim Zaitsev, and Robert Turner. An embedded optical tracking system for motion- corrected magnetic resonance imaging at 7T. Magnetic Resonance Materials in Physics, Biology and Medicine, 25(6):443–453, 2012.

[347] Ahmad Seif Kanaan, Alfred Anwander, Andreas Schäfer, Berkin Bilgic, Torsten Schlumm, Jamie Near, Kirsten Müller-Vahl, and Harald E. Möller. QSM meets MRS: The inﬂuence of subcortical iron on glutamatergic neurotransmission in a movement disorder population. In Proceedings of the 25th Annual Meeting of ISMRM, page 4649, 2017. Bibliography 179

[348] Gregory J Anderson and Christopher D Vulpe. The Cellular Physiology of Iron, pages 3–29. Humana Press, Totowa, NJ, 2010.

[349] John Beard. Iron deﬁciency alters brain development and functioning. The Journal of nutrition, 133(5 Suppl 1):1468S–72S, 2003.

[350] Andrew J Dwork, Gregory Lawler, Patricia A Zybert, Margaret Durkin, Mohammed Osman, Nicholas Willson, and Amiram I Barkai. An autoradiographic study of the uptake and distristri- bution of iron by the brain of the youg rat. Brain Research, 518:31–39, 1990.

[351] J M Hill. The distribution of iron in the brain. In Brain iron: Neurochemical and behavioural aspects, pages 1–24. Taylor and Francis London, 1988.

[352] Moussa BH Youdim, Dorit Ben-Shachar, and Shlomo Yehuda. Putative biological mechanisms of the eﬀect of iron deﬁciency on brain biochemistry and behavior. American Journal of Clinical Nutrition, 50:607–617, 1989.

[353] S Yehuda. Neurochemical basis of behavioral eﬀects of brain iron deﬁciency in anemia. Brain, behavior and iron in the infant diet, pages 63–81, 1990.

[354] D Li. Eﬀects of iron deﬁciency on iron distribution and gamma-aminobutyric acid (GABA) metabolism in young rat brain tissues. Hokkaido Igaku Zasshi. Hokkaido Journal of Medical Science, 73(3):215–225, 1998.

[355] Rama Devi Mittal, Amita Pandey, Balraj Mittal, and Kailash Nath Agarwal. Eﬀect of latent iron deﬁciency on GABA and glutamate neuroreceptors in rat brain. Indian Journal of Clinical Biochemistry, 17(2):1–6, 2002.

[356] Veena Taneja, Kamalapati Mishra, and Kailash N Agarwal. Eﬀect of early iron deﬁciency in rat on the gamma-aminobutyric acid shunt in brain. J Neurochem, 46(6):1670–1674, 1986.

[357] Joel G. Anderson, Paula T. Cooney, and Keith M. Erikson. Brain manganese accumulation is inversely related to γ-amino butyric acid uptake in Male and Female rats. Toxicological Sciences, 95(1):188–195, 2007.

[358] Keith M Erikson, Zakariya K Shihabi, Judy L Aschner, and Michael Aschner. Manganese ac- cumulates in iron-deﬁcient rat brain regions in a heterogeneous fashion and is associated with neurochemical alterations. Biological trace element research, 87(1-3):143–156, 2002.

[359] Kailash N. Agarwal. Iron and the brain: neurotransmitter receptors and magnetic resonance spectroscopy. British Journal of Nutrition, 85(Suppl 2):S147– S150, 2007.

[360] Raghavendra Rao, Ivan Tkac, Elise L Townsend, Rolf Gruetter, and Michael K Georgieff. Perinatal iron deficiency alters the neurochemical profile of the developing rat hippocampus. The Journal of nutrition, 133(10):3215–3221, 2003.

[361] Kay L Ward, Ivan Tkac, Yuezhou Jing, Barbara Felt, John Beard, James Connor, Timothy Schallert, Michael K Georgieﬀ, and Raghavendra Rao. Gestational and lactational iron deﬁciency alters the developing striatal metabolome and associated behaviors in young rats. The Journal of nutrition, 137(4):1043–1049, 2007.

[362] Mu-Hong Chen, Tung-Ping Su, Ying-Sheue Chen, Ju-Wei Hsu, Kai-Lin Huang, Wen-Han Chang, Tzeng-Ji Chen, and Ya-Mei Bai. Association between psychiatric disorders and iron deﬁciency anemia among children and adolescents: a nationwide population-based study. BMC psychiatry, 13(1):161, 2013.

[363] Samuele Cortese, Michel Lecendreux, Bernardo Dalla Bernardina, Marie Christine Mouren, An- drea Sbarbati, and Eric Konofal. Attention-deﬁcit/hyperactivity disorder, Tourette’s syndrome, and restless legs syndrome: the iron hypothesis. Medical hypotheses, 70(6):1128–32, jan 2008.

[364] Matan Avrahami, Ran Barzilay, Miki HarGil, Abraham Weizman, and Nathan Watemberg. Serum Ferritin Levels Are Lower in Children With Tic Disorders Compared with Children Without Tics: A Cross-Sectional Study. Journal of Child and Adolescent Psychopharmacology, XX(Xx):cap.2016.0069, 2016.

[365] Bradley S Peterson, John C Gore, Mark A Riddle, Donald J Cohen, and James F Leckman. Bibliography 180

Abnormal magnetic resonance imaging T2 relaxation time asymmetries in Tourette’s syndrome. Psychiatry research, 55(4):205–21, 1994.

[366] Wei Wang, Mary Ann Knovich, Lan G Coﬀman, Frank M Torti, and Suzy V Torti. Serum ferritin: Past, present and future. Biochimica et biophysica acta, 1800(8):760–9, 2010.

[367] Stefan Ropele and Christian Langkammer. Iron quantiﬁcation with susceptibility. NMR in Biomedicine, doi: 10.10, 2016.

[368] Andreas Deistung, V Endmayr, S Hametner, H Lassmann, JR Reichenbach, S Robinson, T Haider, H Traxler, E Haimburger, S Trattnig, and G Grabner. Toward Iron Distribution Mapping using Quantitative Susceptibility Mapping (QSM):A Comparison of Histological Iron Concentration Maps with Magnetic Susceptibility Maps. Proceedings of the International Society for Magnetic Resonance in Medicine, 24:#30, 2016.

[369] Christian Langkammer, Nikolaus Krebs, Walter Goessler, Eva Scheurer, Franz Ebner, Kathrin Yen, Franz Fazekas, and Stefan Ropele. Quantitative MR Imaging of Brain Iron: A Postmortem Validation Study 1. Radiology, 257(November):455–462, 2010.

[370] Ferdinand Schweser, Xiang Feng, Rosa Mach Batlle, Daniel Güllmar, Andreas Deistung, Michael G Dwyer, Robert Zivadinov, and Jürgen R Reichenbach. Improved deep gray matter segmentation using anatomical information from quantitative susceptibility maps. In Proc Intl Soc Mag Reson Med 22 (2014), volume 1878, page 1787, Milan, Italy, 2014.

[371] David Atkinson, Derek L Hill, Paul N Stoyle, Paul E Summers, and Stephen F Keevil. Automatic correction of motion artifacts in magnetic resonance images using an entropy focus criterion. IEEE transactions on medical imaging, 16(6):903–10, 1997.

[372] Benedict Mortamet, Matt A. Bernstein, Cliﬀord R. Jack, Jeﬀrey L. Gunter, Chadwick Ward, Paula J. Britson, Reto Meuli, Jean Philippe Thiran, and Gunnar Krueger. Automatic quality assessment in structural brain magnetic resonance imaging. Magnetic Resonance in Medicine, 62(2):365–372, 2009.

[373] Betsy Lozoff and Michael K Georgieff. Iron Deficiency and Brain Development. Seminars in Pediatric Neurology, 13(3):158–165, 2006.

[374] Bruce C Kennedy, Diana J Wallin, Phu V Tran, and Michael K Georgieﬀ. Long-Term Brain and Behavioral Consequences of Early-Life Iron Deﬁciency, pages 295–316. Springer International Publishing, Cham, 2016.

[375] Betsy Lozoff, John Beard, James Connor, Barbara Felt, Michael Georgieff, and Timoth Schallert. Long-lasting neural and behavioral effects of iron deficiency in infancy. Nutr Rev, 64(5 Pt 2):S34– 43; discussion S72–91, 2006.

[376] Barbara T. Felt, John L. Beard, Timothy Schallert, Jie Shao, J. Wayne Aldridge, James R. Connor, Michael K. Georgieff, and Betsy Lozoff. Persistent neurochemical and behavioral abnormalities in adulthood despite early iron supplementation for perinatal iron deficiency anemia in rats. Behavioural Brain Research, 171(2):261–270, 2006.

[377] SL Clardy, X Wang, W Zhao, W Liu, GA Chase, JL Beard, B True Felt, and JR Connor. Acute and chronic eﬀects of developmental iron deﬁciency on mRNA expression patterns in the brain. Journal of neural transmission. Supplementum, 71:173–96, 2006.

[378] Yi Guo, Linyan Su, Jie Zhang, Jing Lei, Xiong Deng, Hongbo Xu, Zuocheng Yang, Shoujin Kuang, Jinsong Tang, Ziqiang Luo, and Hao Deng. Analysis of the BTBD9 and HTR2C variants in Chinese Han patients with Tourette syndrome. Psychiatric Genetics, 22(6):300–303, 2012.

[379] Jean-Baptiste Rivière, Lan Xiong, Anastasia Levchenko, Judith St-Onge, Claudia Gaspar, Yves Dion, Paul Lespérance, Geneviève Tellier, François Richer, Sylvain Chouinard, Guy A. Rouleau, and Montreal Tourette Study Group. Association of intronic variants of the BTBD9 gene with Tourette syndrome. Archives of neurology, 66(10):1267–72, 2009.

[380] Bernd Goellner and Hermann Aberle. The synaptic cytoskeleton in development and disease. Developmental Neurobiology, 72(1):111–125, 2012. Bibliography 181

[381] Betsy Lozoff. Early Iron Deficiency Has Brain and Behavior Effects Consistent with Dopaminergic Dysfunction. J. Nutr, 141:740–746, 2011.

[382] Amanda M. Ill, Todd R. Mitchell, Elizabeth B. Neely, and James R. Connor. Metabolic analysis of mouse brains that have compromised iron storage. Metabolic Brain Disease, 21(2-3):77–87, 2006.

[383] Nikolaus Krebs, Christian Langkammer, Walter Goessler, Stefan Ropele, Franz Fazekas, Kathrin Yen, and Eva Scheurer. Assessment of trace elements in human brain using inductively coupled plasma mass spectrometry. Journal of Trace Elements in Medicine and Biology, 28(1):1–7, 2014.

[384] Sarah Gerasch, Ahmad Seif Kanaan, Ewgeni Jakubovski, and Kirsten R Müller-Vahl. Aripiprazole Improves Associated Comorbid Conditions in Addition to Tics in Adult Patients with Gilles de la Tourette Syndrome. Frontiers in Neuroscience, 10:416, 2016.

[385] Arthur K Shapiro, E Shapiro, and H Wayne. Treatment of Tourette’s Syndrome with Haloperidol, Review of 34 Cases. Archives of General Psychiatry, 28(1):92–97, jan 1973.

[386] Toya M. Bowles and Gary M. Levin. Aripiprazole: A new atypical antipsychotic drug. Annals of Pharmacotherapy, 37(5):687–694, 2003.

[387] Ahmad Ghanizadeh and Alireza Haghighi. Aripiprazole versus risperidone for treating children and adolescents with tic disorder: A randomized double blind clinical trial. Child Psychiatry and Human Development, 45(5):596–603, 2014.

[388] Ahmad Ghanizadeh. Systemic review of aripiprazole for the treatment of children and adolescents with tic disorders. Neurosciences, 17(3):200–204, 2012.

[389] Tanya K Murphy, Michael A Bengtson, Ohel Soto, Paula J Edge, Muhammad W Sajid, Nathan Shapira, and Mark Yang. Case series on the use of aripiprazole for Tourette syndrome. Interna- tional Journal of Neuropsychopharmacology, 8(3):489–490, sep 2005.

[390] Christine Winter, Andreas Heinz, Andreas Kupsch, and Andreas Ströhle. Aripiprazole in a Case Presenting With Tourette Syndrome and Obsessive-Compulsive Disorder. Journal of Clinical Psychopharmacology, 28(4), 2008.

[391] Gabriele Masi, Antonella Gagliano, Rosamaria Siracusano, Stefano Berloﬀa, Tiziana Calarese, Giovanna Ilardo, Chiara Pfanner, Angela Magazù, and Clemente Cedro. Aripiprazole in Children with Tourette’s Disorder and Co-morbid Attention-Deﬁcit/Hyperactivity Disorder: A 12-Week, Open-Label, Preliminary Study. Journal of Child and Adolescent Psychopharmacology, 22(2):120– 125, feb 2012.

[392] Jan Frölich, Martina Starck, Tobias Banaschewski, and Gerd Lehmkuhl. Aripiprazole - a medical treatment alternative for Tourette Syndrome in childhood and adolescence. Zeitschrift fur Kinder- und Jugendpsychiatrie und Psychotherapie, 38:291–298, 2010.

[393] Mariangela Gulisano, Paola V Calì, Andrea E Cavanna, Clare Eddy, Hugh Rickards, and Re- nata Rizzo. Cardiovascular safety of aripiprazole and pimozide in young patients with Tourette syndrome. Neurological Sciences, 32(6):1213–1217, 2011.

[394] Cathy Budman, Barbara J Coﬀey, Rachel Shechter, Matthew Schrock, Natalie Wieland, Arie Spirgel, and Elizabeth Simon. Aripiprazole in Children and Adolescents with Tourette Disorder with and without Explosive Outbursts. Journal of Child and Adolescent Psychopharmacology, 18(5):509–515, oct 2008.

[395] Lisa Davies, Jeremy S. Stern, Niruj Agrawal, and Mary M. Robertson. A case series of patients with Tourette’s Syndrome in the United Kingdom treated with aripiprazole. Human Psychophar- macology, 21(7):447–453, 2006.

[396] Gholson J Lyon, Stephanie Samar, Rahil Jummani, Scott Hirsch, Arie Spirgel, Rachel Goldman, and Barbara J Coﬀey. Aripiprazole in Children and Adolescents with Tourette’s Disorder: An Open-Label Safety and Tolerability Study. Journal of Child and Adolescent Psychopharmacology, 19(6):623–633, dec 2009.

[397] Hanik K Yoo, Yoo Sook Joung, Jeong-Seop Lee, Dong Ho Song, Young Sik Lee, Jae-Won Kim, Bung-Nyun Kim, and Soo Churl Cho. A multicenter, randomized, double-blind, placebo-controlled study of aripiprazole in children and adolescents with Tourette’s disorder. The Journal of clinical Bibliography 182

psychiatry, 74(8):772–780, 2013.

[398] Christos Ganos, Ursula Kahl, Odette Schunke, Simone Kühn, Patrick Haggard, Christian Gerloﬀ, Veit Roessner, Götz Thomalla, and Alexander Münchau. Are premonitory urges a prerequisite of tic inhibition in Gilles de la Tourette syndrome? Journal of Neurology, Neurosurgery & Psychiatry, 83(10):975–978, 2012.

[399] Kirsten R. Müller-Vahl, Laura Riemann, and Stefanie Bokemeyer. Tourette patients’ misbelief of a tic rebound is due to overall diﬃculties in reliable tic rating. Journal of Psychosomatic Research, 76(6):472–476, 2014.

[400] T. Beblo and S Lautenbacher. Neuropsychologie der Depression. Hogrefe, Goettingen, 2015.

[401] Elena Olariu, José-Ignacio Castro-Rodriguez, Pilar Álvarez, Carolina Garnier, Marta Reinoso, Luis Miguel Martín-López, Jordi Alonso, and Carlos G Forero. Validation of clinical symptom IRT scores for diagnosis and severity assessment of common mental disorders. Quality of Life Research, 24(4):979–992, 2015.

[402] Mehdi Sayyah, Mohammad Sayyah, Hatam Boostani, Seyyed Mohammad Ghaﬀari, and Abedin Hoseini. Eﬀects of aripiprazole augmentation in treatment-resistant obsessive-compulsive disorder (a double blind clinical trial). Depression and Anxiety, 29(10):850–854, oct 2012.

[403] Markus Dold, Martin Aigner, Rupert Lanzenberger, and Siegfried Kasper. Antipsychotic Augmen- tation of Serotonin Reuptake Inhibitors in Treatment-Resistant Obsessive-Compulsive Disorder: An Update Meta-Analysis of Double-Blind, Randomized, Placebo-Controlled Trials. International Journal of Neuropsychopharmacology, 18(9):pyv047–pyv047, may 2015.

[404] Saeed Shoja Shafti and Hamid Kaviani. Aripiprazole versus quetiapine in treatment-resistant obsessive-compulsive disorder: a double-blind clinical trial. Therapeutic advances in psychopharmacology, 5(1):32–7, 2015.

[405] Gabriele Masi, Chiara Pfanner, and Paola Brovedani. Antipsychotic augmentation of selective serotonin reuptake inhibitors in resistant tic-related obsessive-compulsive disorder in children and adolescents: Anaturalistic comparative study. Journal of Psychiatric Research, 47(8):1007–1012, 2013.

[406] Roberto Delle Chiaie, Pierluigi Scarciglia, Massimo Pasquini, Maria Caredda, and Massimo Biondi. Aripiprazole augmentation in patients with resistant obsessive compulsive disorder: a pilot study. Clinical practice and epidemiology in mental health: CP & EMH, 7:107–11, 2011.

[407] X. J. Wen, L. M. Wang, Z. L. Liu, A. Huang, Y. Y. Liu, and J. Y. Hu. Meta-analysis on the eﬃcacy and tolerability of the augmentation of antidepressants with atypical antipsychotics in patients with major depressive disorder. Brazilian Journal of Medical and Biological Research, 47(7):605–616, 2014.

[408] Martin A Katzman. Aripiprazole: A clinical review of its use for the treatment of anxiety disorders and anxiety as a comorbidity in mental illness. Journal of Aﬀective Disorders, 128:S11–S20, 2011.

[409] Chi-Un Pae, Alessandro Serretti, Ashwin a Patkar, and Praksh S Masand. Aripiprazole in the treatment of depressive and anxiety disorders: a review of current evidence. CNS drugs, 22(5):367– 388, 2008.

[410] Louise Carton, Olivier Cottencin, Maryse Lapeyre-Mestre, Pierre Geoﬀroy, Jonathan Favre, Nico- las Simon, Regis Bordet, and Benjamin Rolland. Oﬀ-Label Prescribing of Antipsychotics in Adults, Children and Elderly Individuals: A Systematic Review of Recent Prescription Trends. Current Pharmaceutical Design, 21(23):3280–3297, 2015.

[411] Ahmad Ghanizadeh. A Systematic Review of the Eﬃcacy and Safety of Desipramine for Treating ADHD. Current Drug Safety, 8(3):169–174, 2013.

[412] Kirsten R Müller-Vahl, Ines Dodel, Norbert Müller, Alexander Munchau, Jens Peter Reese, Monika Balzer-Geldsetzer, Richard Dodel, and Wolfgang H Oertel. Health-related quality of life in patients with Gilles de la Tourette’s syndrome. Movement disorders: oﬃcial journal of the Movement Disorder Society, 25(3):309–314, 2010.

[413] Isabelle Jalenques, Fabienne Galland, Laurent Malet, Dominique Morand, Guillaume Legrand, Appendix 183

Candy Auclair, Andreas Hartmann, Philippe Derost, and Franck Durif. Quality of life in adults with Gilles de la Tourette Syndrome. BMC Psychiatry, 12(1):109, 2012.

[414] Tanvi Sambrani, Ewgeni Jakubovski, and Kirsten R. M??ller-Vahl. New insights into clinical characteristics of Gilles de la Tourette syndrome: Findings in 1032 patients from a single German center. Frontiers in Neuroscience, 10(SEP), 2016.

[415] Natsumi Matsuda, Toshiaki Kono, Maiko Nonaka, Miyuki Fujio, and Yukiko Kano. Self-initiated coping with Tourette’s syndrome: Eﬀect of tic suppression on QOL. Brain and Development, 38(2):233–241, 2016.

[416] S Wilhelm, Peterson AL, J Piacentini, and et Al. Randomized trial of behavior therapy for adults with tourette syndrome. Archives of General Psychiatry, 69(8):795–803, aug 2012. Part VII

APPENDIX

184 Appendix A

Related publications

The work presented in this thesis describes a set of methodological and pathophysiological investigations that led to the publication of peer-reviewed journal articles, the presentation of abstracts at international conferences, and the preparation of a manuscripts that are under submission. References to the published and unpublished material outlined below is detailed in the subheadings of the respective Chapters.

Published peer-reviewed Journal Articles

1. Kanaan A.S.*, Gerasch S., Garcia-Garcia I., Lampe L., Pampel A, Anwander A, Near J., Möller H.E.**, Müller-Vahl K.R.** Pathological glutamatergic neurotransmission in Gilles de la Tourette Syndrome. Brain 2017: 140 218-234

2. Forde NJ*, Kanaan A.S.*, Widomska J., Padmanabhuni S.S.,Nespoli E., Alexan- der J, Arranz J.I.R, Fan S., Houssari R., Nawaz M., ZilhÃčo N, Pagliaroli L. Rizzo F., Aranyi T., Barta C., Boeckers T., Boomsma D.,Buisman W., Buitelaar J.K.,Cath D., Dietrich A., Driessen N., Drineas P., Dunlap M., Gerasch S., Glen- non J., Hengerer B., van den Heuvel O., Jespersgaard C., Möller H.E., Müller-Vahl K.R., Openneer T., Poelmans G., Pouwels P.J., Scharf J, Stefansson H., Tümer Z., Veltman D., van der Werf Y.D., Hoekstra P., Ludolph A. and Paschou P. TS- EUROTRAIN: A European-wide investigation and training network on the aetiology and pathophysiology of Gilles de la Tourette Syndrome. Frontiers in Neuroscience 2016; 10: 1-9.

3. Gerasch S., Kanaan A.S., Jakubovski E. and Müller-Vahl K.R. Aripiprazole improves associated comorbid Conditions in addition to Tics in adult Patients with Gilles de la Tourette Syndrome. Frontiers in Neuroscience 2016; 10: 416.

185 Appendix 186

Unpublished manuscripts

5. Kanaan A.S.*, Anwander A., Bilgic B., Torsten S., Jamie N., Müller-Vahl K.R.**, Möller H.E.** Subcortical iron reductions associated with glutamatergic neurotransmission in Gilles de la Tourette syndrome.

6. Metere R.*, Kanaan A.S.*, Bilgic B., Torsten S., Möller H.E. Eﬀects of the coil combination algorithm on quantitative susceptibility mapping

Conference Proceedings

7. Kanaan A.S., Anwander A., Schäfer A., Metere R., Schlumm T., Near J., Bilgic B., Müller-Vahl K.R.,and Möller H.E. QSM meets MRS: The inﬂuence of subcortical iron on glutamatergic neurotransmission in a movement disorder population International Society for Magnetic Resonance Imaging in Medicine (ISMRM), Ab- stract # 4649, Honolulu, Hawaii, USA, April 22 - April 27, 2017.

8. Kanaan A.S., Anwander A., Gerasch S., Garcia-Garcia I., Lampe L., Pampel A, Anwander A, Bilgic B,. Near J., Müller-Vahl K.R., Möller H.E. Pathological glutamatergic neurotransmission associated with reduced subcortical iron in Tourette syndrome. Max Planck Institute 2017 Scientiﬁc Advisory board., Leipzig , Germany, February 7-8, 2016

9. Metere R., Kanaan A.S., Schäfer A., Bilgic B., Schlumm T., and Möller H.E. Ef- fects of diﬀerent coil reconstruction algorithms on Quantitative Suscep- tibility Mapping results. 4th International Workshop on MRI Phase Contrast & Quantitative Susceptibility Mapping. Graz, Austria. September 26-28, 2016.

10. Kanaan A.S., Schäfer A, Bilgic, B. Müller-Vahl K.R., Möller H.E. Quantitative Susceptibility Mapping in Gilles de la Tourette Syndrome. European So- ciety for the study of Tourette Syndrome (ESSTS) 2016 Annual Meeting. Warsaw, Poland. June 8-11 2016.

11. Gerasch S,.Kanaan A.S.*, Jakuvoski E., Müller-Vahl K.R. Aripiprazole improves associated comorbid conditions in addition to tics in adult patients with Gilles de la Tourette Syndrome. European Society for the study of Tourette Syndrome (ESSTS) 2016 Annual Meeting. Warsaw, Poland. June 8-11 2016.

12. Kanaan A.S.*, Gerasch, Garcia-Garcia I., Lampe L., Pampel A, Anwander A, Near J, Möller HE*, Müller-Vahl K*. Pathological glutamatergic neurotransmission in Gilles de la Tourette Syndrome. 24th Annual Meeting of the Appendix 187

International Society for Magnetic Resonance in Medicine (ISMRM), Abstract # 2420, Singapore, Singapore, May 7 - 13, 2016

13. Kanaan A.S.*, Pamel A., Müller-Vahl K.R., Möller H.E. Test-Retest Quanti- tation of Absolute Metabolite Concentrations with Partial-Volume correction using diﬀerent segmentation methods. International Society for Magnetic Resonance Imaging in Medicine (ISMRM), Abstract # 1974, Toronto, Ontario, Canada, May 30 - June 5, 2015.

14. Kanaan A.S.*, Margulies D.S., Anwander A., Möller H.E., Müller-Vahl K.R. Retrospective control for motion-artefacts in functional neuroimaging datasets using Wavelet and ICA based methods , 1st World Congress on Tourette Syndrome & Tic Disorders, Abstract # 202, 1st Tourette World Congress. London, UK. June 24-26 2015.

15. Kanaan A.S.*, Forde N.J., et al. TS-EUROTRAIN: European wide investigation of the etiology and pathophysiology of Gilles de la Tourette syndrome and related disorders. 1st World Congress on Tourette Syndrome & Tic Disorders, Abstract # 111. London, UK. June 24-26 2015.

16. Kanaan A.S.*, Gerasch S., Pampel A., Lampe L., Schäfer A., Margulies D., Möller H.E., Müller-Vahl K.R. Elemental, neurochemical and network based analysis of the pathophysiological mechanisms of Gilles de la Tourette Syndrome. European Society for the study of Tourette Syndrome (ESSTS) 2014 Annual Meeting. Paris, France. April 25-26 2014, Appendix B

Contribution of authours

The work presented in this thesis is the result of a close collaboration between the author and his supervisors Kirsten R. Müller-Vahl and Harald E. Möller. We jointly designed the experiments, conducted the analyses, and wrote the primary manuscripts. The following list summarizes the contributions of all other authors in alphabetical order:

1. Anwander, Alfred: Max Planck Institute for Human Cognitive and Brain Sci- ences. Provided ample ideas and technical support in solving varied image processing and pipelining issues.

2. Bilgic, Berkin: A.A. Martinos Center for Biomedical Imaging and Department of Radiology, Harvard Medical School, Boston, MA, USA. Developed and provided access and support to the ESPIRiT-SVD coil combination algorithm.

3. Garcia Garcia, Isabel: Max Planck Institute for Human Cognitive and Brain Sciences. Provided assistance in performing statistical analyses, acquiring data and prooreading manuscripts.

4. Gerasch, Sarah: Hannover Medical School. Performed patient recruitment and acquisition of the clinical data.

5. Jakubovski, Ewgeni: Provided assistance in clinical result interpretation and the writing of the aripiprazole manuscript.

6. Lampe, Leonie: Hannover Medical School. Provided support in debrieﬁng patients, acquiring imaging data and blood samples.

7. Pascho, Peristera: Hannover Medical School. Coordinated the TS-EUROTRAIN Initial Training Network. Drafted the articles with the joint ﬁrst authors.

188 Appendix 189

8. Pampel, Andre: Max Planck Institute for Human Cognitive and Brain Sciences. Provided technical support in developing and testing imaging sequences.

9. Margulies, Daniel: Max Planck Institute for Human Cognitive and Brain Sci- ences. Provided advice and support in image processing and motion correction techniques.

10. Metere, Riccardo: Max Planck Institute for Human Cognitive and Brain Sci- ences. Provided technical support in image processing of QSM data and drafted the coil combination manuscript.

11. Near, Jamie: Douglas Mental Health University Institute and Department of Psychiatry, McGill University, Montreal, QC, Canada. Developed and provided technical support to the frequency and phase correction technique and the FID- appliance toolbox.

12. Schäfer, Andreas: Siemens Healthcare GmbH and Max Planck Institute for Hu- man Cognitive and Brain Sciences. Provided access and support to the QSM reconstruction pipeline.

13. Schlumm, Torsten: Max Planck Institute for Human Cognitive and Brain Sci- ences. Re-developed the GRE sequence to export RAW multichannel data. Appendix C

Research compliance certiﬁcates

• Ethics committee approval from Hannover Medical School

• Ethics committee approval from the Medical faculty of the University of Leipzig

190

Appendix D

MR Imaging sequence parameters

• AutoAlignHead scout sequence:

– AAHScout_32ch.

• T1-weighted MP2RAGE anatomical image sequence:

– mp2rage_p3_602B.

• FASTESTMAP 1H-MRS shimming sequence:

– fastestmap_577_st.

• Single-voxel 1H-MRS from the anterior mid-cingulate cortex:

– Water suppressed: svs_se_30_acc. – Water reference: svs_se_30_ref_head_acc.

• Single-voxel 1H-MRS from the bi-lateral thalamus:

– Water suppressed: svs_se_30_th. – Water reference: svs_se_30_ref_head_th.

• Single-voxel 1H-MRS from the left striatum:

– Water suppressed: svs_se_30_st. – Water reference: svs_se_30_ref_head_st.

• Resting state fMRI ( used for inspection of motion âĂŞ see Chapter 8:

– cmrr_mbep2d_resting.

• Gradient Echo sequence (for Quantiative Susceptibility Mapping):

– as_gre_TE17ms_nifti.

197 SIEMENS MAGNETOM Verio syngo MR B17

\\USER\ASK\TOURETTE\AA8P150321\AAHScout_32ch TA: 0:14 PAT: 3 Voxelgröße: 1.6×1.6×1.6 mm Rel. SNR: 1.00 SIEMENS: AALScout

Rohdaten Aus Eigenschaften Elliptischer Filter Aus Prio Rekonstr. Aus Vor der Messung Geometrie Nach der Messung Mehrschichtmodus Sequenziell Load to viewer Ein Serie Aufsteigend Inline movie Aus System Auto store images Ein Body Aus Load to stamp segments Ein HEP Ein Bilder in großes Bildsegment Ein HEA Ein laden ------Auto open inline display Aus Positionierungsmodus REF Start measurement without Aus Tischposition H further preparation Tischposition 0 mm Auf Start duch Benutzer Aus MSMA S - C - T warten Sagittal R >> L Start measurements single Coronar A >> P Routine Transversal F >> H 3D-Block-Gruppe 1 Unkombiniert speichern Aus 3D-Blöcke 1 Kanalkombination Adaptive Combine Autom. Spulenanwahl Default Distanzfaktor 20 % ------Position L0.0 P20.0 H0.0 Shim-Modus Tune-Up Orientierung Sagittal Mit Körperspule justieren Aus Phasenkod.-Richt. A >> P Freq. Justage bestät. Aus Rotation 0 Grad von Silikon ausgehen Aus AutoAlign Kopf ? Ref. Amplitude 1H 0.000 V Phasen-Oversampling 0 % Justagetoleranz Auto Schicht-Oversampling 0.0 % Justagevolumen Schichten im 3D-Block 128 Position Isozentrum FoV Auslese 260 mm Orientierung Transversal FoV Phase 100.0 % Rotation 0.00 Grad Schichtdicke 1.6 mm R >> L 350 mm TR 3.15 ms A >> P 263 mm TE 1.37 ms F >> H 350 mm Mittelungen 1 Verknüpfungen 1 Inline Filter Prescan Normalisierung Zeit bis k-Raummitte 6.2 s Spulenelemente HEA;HEP Sequenz Kontrast Einleitung Ein Flipwinkel 8 Grad Dimension 3D ------Asymmetrisches Echo Schwach Mittelungsmodus Kurzzeit Kontraste 1 Rekonstruktion Betrag Bandbreite 550 Hz/Px Messungen 1 ------HF-Puls-Typ Schnell Auflösung Gradientenmodus Normal Basis-Auflösung 160 Anregung Nichtsel. Phasen-Auflösung 100 % HF-Spoiler Ein Schicht-Auflösung 69 % Phasen Partial Fourier 6/8 Schicht Partial Fourier 6/8 ------PAT Modus GRAPPA Beschl. Faktor PE 3 Ref. Zeilen PE 24 Beschl. Faktor 3D 1 Matrix Spulen Modus Auto (Triple) Referenzmessungsmodus Integriert ------Image Filter Aus Verzeichn. Korr. Aus Ungefilterte Bilder Aus Prescan Normalisierung Ein Normalisierung Aus B1-Filter Aus

1/+ SIEMENS MAGNETOM Verio syngo MR B17

\\USER\ASK\TOURETTE\AA8P150321\mp2rage_p3_602B TA: 8:22 PAT: 3 Voxelgröße: 1.0×1.0×1.0 mm Rel. SNR: 1.00 USER: mp2rage_wip602B

Referenzmessungsmodus Integriert Eigenschaften ------Prio Rekonstr. Aus Image Filter Aus Vor der Messung Verzeichn. Korr. Aus Nach der Messung Ungefilterte Bilder Aus Load to viewer Ein Prescan Normalisierung Ein Inline movie Aus Normalisierung Aus Auto store images Ein B1-Filter Aus Load to stamp segments Aus Rohdaten Aus Bilder in großes Bildsegment Aus Elliptischer Filter Aus laden Geometrie Auto open inline display Aus Start measurement without Ein Mehrschichtmodus Einzelmess. Serie Verschachtelt further preparation ------Auf Start duch Benutzer Aus warten System Start measurements single Body Aus HEP Ein Routine HEA Ein 3D-Block-Gruppe 1 ------3D-Blöcke 1 Positionierungsmodus FIX Distanzfaktor 50 % Tischposition H Position L0.6 P4.8 F23.8 Tischposition 0 mm Orientierung S > C-1.1 > T-0.5 MSMA S - C - T Phasenkod.-Richt. A >> P Sagittal R >> L Rotation 0.00 Grad Coronar A >> P Phasen-Oversampling 0 % Transversal F >> H Schicht-Oversampling 0.0 % Unkombiniert speichern Aus Schichten im 3D-Block 176 Kanalkombination Adaptive Combine FoV Auslese 256 mm AutoAlign Kopf > Gehirn FoV Phase 93.8 % Autom. Spulenanwahl Default ------Schichtdicke 1.00 mm Shim-Modus Standard TR 5000 ms Mit Körperspule justieren Aus TE 2.98 ms Freq. Justage bestät. Aus Mittelungen 1 von Silikon ausgehen Aus Verknüpfungen 1 ? Ref. Amplitude 1H 0.000 V Filter Prescan Normalisierung Justagetoleranz Auto Spulenelemente HEA;HEP Justagevolumen Kontrast Position L0.6 P4.8 F23.8 Magn. Präparation Nichtsel. IR Orientierung S > C-1.1 > T-0.5 TI 1 700 ms Rotation 0.00 Grad TI 2 2500 ms F >> H 256 mm Flipwinkel 1 4 Grad A >> P 240 mm Flipwinkel 2 5 Grad R >> L 176 mm Fettunterdr. Keine Physio Wasserunterdr. Keine 1.Signal/Modus Kein 2nd Inversion-Contrast Ein ------Dark Blood Aus Mittelungsmodus Langzeit ------Rekonstruktion Betrag Atemkontrolle Aus Messungen 1 Inline Mehrere Serien Jede Messung Subtrahieren Aus Auflösung Std-Abw.-Sag Aus Basis-Auflösung 256 Std-Abw.-Cor Aus Phasen-Auflösung 100 % Std-Abw.-Tra Aus Schicht-Auflösung 100 % Std-Abw.-Zeit Aus Phasen Partial Fourier Aus MIP-Sag Aus Schicht Partial Fourier Aus MIP-Cor Aus Interpolation Aus MIP-Tra Aus ------MIP-Zeit Aus PAT Modus GRAPPA Originalbilder speichern Ein Beschl. Faktor PE 3 Ref. Zeilen PE 32 Sequenz Beschl. Faktor 3D 1 Einleitung Ein Matrix Spulen Modus Auto (Triple) Dimension 3D 2/+ SIEMENS MAGNETOM Verio syngo MR B17

Elliptische Abtastung Aus Asymmetrisches Echo Aus Kontraste 1 Bandbreite 240 Hz/Px Flusskomp. Nein Echoabstand 7.1 ms ------HF-Puls-Typ Schnell Gradientenmodus Normal Anregung Nichtsel. HF-Spoiler Ein ------FFT Scale Factor 100 % Line/Partition Swap Aus Homodyne Phase Filter Aus Flat Image Ein T1 Map Ein Division Image Ein ExtInvPulseOn Aus

3/+ SIEMENS MAGNETOM Verio syngo MR B17

\\USER\ASK\TOURETTE\AA8P150321\fastestmap_577_st TA: 0:12 VoI: 20 ×15 ×20 mm Rel. SNR: 1.00 USER: fastestmap_577

Position L18.9 A8.9 H5.3 Eigenschaften Orientierung Transversal Prio Rekonstr. Aus Rotation 89.15 Grad Vor der Messung A >> P 20 mm Nach der Messung R >> L 15 mm Load to viewer Ein F >> H 20 mm Inline movie Aus Auto store images Ein Physio Load to stamp segments Aus 1.Signal/Modus Kein Bilder in großes Bildsegment Aus Sequenz laden Phasenzyklierung Kein Auto open inline display Aus Bandbreite 100000 Hz Start measurement without Ein Messdauer 2 ms further preparation ------Auf Start duch Benutzer Aus Type of fit Full 6-bar warten VoI fit factor 150 % Start measurements single Refocus pulses Normal Routine Excitation pulse duration 5760 ms Position L18.9 A8.9 H5.3 Refocus pulse duration 5120 ms Orientierung Transversal Bar FoV 384 mm Rotation -0.85 Grad Bar thickness 5.0 mm VoI A >> P 20 mm Multi-echo acquisition Ein VoI R >> L 15 mm Number of echoes 12 VoI F >> H 20 mm Inversion pulse Aus TR 2000 ms TE 50.00 ms Mittelungen 1 Filter Keine Spulenelemente HEA;HEP Kontrast Tau 5.00 ms Excite flip angle 90 Grad Refocus flip angle 180 Grad Messungen 1 ------Auflösung Vektorgröße 256 ------Matrix Spulen Modus Auto (Triple) Geometrie System Body Aus HEP Ein HEA Ein ------Positionierungsmodus FIX Tischposition H Tischposition 0 mm MSMA S - C - T Sagittal R >> L Coronar A >> P Transversal F >> H Unkombiniert speichern Aus AutoAlign Kopf > Gehirn Autom. Spulenanwahl Default ------Shim-Modus Tune-Up Wasserunterdr. just. Aus Mit Körperspule justieren Aus Freq. Justage bestät. Aus von Silikon ausgehen Aus ? Ref. Amplitude 1H 0.000 V Justagetoleranz Auto Justagevolumen 17/+ SIEMENS MAGNETOM Verio syngo MR B17

\\USER\ASK\TOURETTE\AA8P150321\svs_se_35_acc TA: 4:12 VoI: 25 ×16 ×16 mm Rel. SNR: 1.00 SIEMENS: svs_se

von Silikon ausgehen Aus Eigenschaften ? Ref. Amplitude 1H 0.000 V Prio Rekonstr. Aus Justagetoleranz Auto Vor der Messung Justagevolumen Nach der Messung Position L0.2 A19.1 H32.7 Load to viewer Ein Orientierung T > C27.0 > S-0.5 Inline movie Aus Rotation 89.05 Grad Auto store images Ein A >> P 25 mm Load to stamp segments Aus R >> L 16 mm Bilder in großes Bildsegment Aus F >> H 16 mm laden Auto open inline display Aus Physio Start measurement without Ein 1.Signal/Modus Kein further preparation Sequenz Auf Start duch Benutzer Aus Präparationsscans 4 warten Delta Frequenz -2.5 ppm Start measurements single Phasenzyklierung Auto Routine Bandbreite 1200 Hz Position L0.2 A19.1 H32.7 Messdauer 853 ms Orientierung T > C27.0 > S-0.5 Entferne Oversampling Ein Rotation -0.95 Grad VoI A >> P 25 mm VoI R >> L 16 mm VoI F >> H 16 mm TR 3000 ms TE 30 ms Mittelungen 80 Filter Prescan Normalisierung Spulenelemente HEA;HEP Kontrast Flipwinkel 90 Grad Wasserunterdr. Wassersättig. H2O-unterdr. Bandbr. 50 Hz Spektrale Unterdr. Keine Messungen 1 ------Auflösung Prescan Normalisierung Ein Vektorgröße 1024 ------Matrix Spulen Modus CP ------Ungefilterte Bilder Aus Geometrie System Body Aus HEP Ein HEA Ein ------Positionierungsmodus FIX Tischposition H Tischposition 0 mm MSMA S - C - T Sagittal R >> L Coronar A >> P Transversal F >> H Unkombiniert speichern Aus AutoAlign Kopf > Gehirn Autom. Spulenanwahl Default ------Shim-Modus Standard Wasserunterdr. just. Ein Mit Körperspule justieren Aus Freq. Justage bestät. Aus 25/+ SIEMENS MAGNETOM Verio syngo MR B17

\\USER\ASK\TOURETTE\AA8P150321\svs_se_30_ref_head_ACC TA: 0:24 VoI: 25 ×16 ×16 mm Rel. SNR: 1.00 SIEMENS: svs_se

? Ref. Amplitude 1H 0.000 V Eigenschaften Justagetoleranz Auto Prio Rekonstr. Aus Justagevolumen Vor der Messung Position L0.2 A19.1 H32.7 Nach der Messung Orientierung T > C27.0 > S-0.5 Load to viewer Ein Rotation 89.05 Grad Inline movie Aus A >> P 25 mm Auto store images Ein R >> L 16 mm Load to stamp segments Aus F >> H 16 mm Bilder in großes Bildsegment Aus laden Physio Auto open inline display Aus 1.Signal/Modus Kein Start measurement without Ein Sequenz further preparation Präparationsscans 4 Auf Start duch Benutzer Aus Delta Frequenz -2.5 ppm warten Phasenzyklierung Auto Start measurements single Bandbreite 1200 Hz Routine Messdauer 1706 ms Position L0.2 A19.1 H32.7 Entferne Oversampling Ein Orientierung T > C27.0 > S-0.5 Rotation -0.95 Grad VoI A >> P 25 mm VoI R >> L 16 mm VoI F >> H 16 mm TR 3000 ms TE 30 ms Mittelungen 4 Filter Prescan Normalisierung Spulenelemente HEA;HEP Kontrast Flipwinkel 90 Grad Wasserunterdr. Keine Spektrale Unterdr. Keine Messungen 1 ------Auflösung Prescan Normalisierung Ein Vektorgröße 2048 ------Matrix Spulen Modus CP ------Ungefilterte Bilder Aus Geometrie System Body Aus HEP Ein HEA Ein ------Positionierungsmodus FIX Tischposition H Tischposition 0 mm MSMA S - C - T Sagittal R >> L Coronar A >> P Transversal F >> H Unkombiniert speichern Aus AutoAlign Kopf > Gehirn Autom. Spulenanwahl Default ------Shim-Modus Standard Wasserunterdr. just. Aus Mit Körperspule justieren Aus Freq. Justage bestät. Aus von Silikon ausgehen Aus 23/+ SIEMENS MAGNETOM Verio syngo MR B17

\\USER\ASK\TOURETTE\AA8P150321\svs_se_35_th TA: 4:12 VoI: 16 ×28 ×16 mm Rel. SNR: 1.00 SIEMENS: svs_se

von Silikon ausgehen Aus Eigenschaften ? Ref. Amplitude 1H 0.000 V Prio Rekonstr. Aus Justagetoleranz Auto Vor der Messung Justagevolumen Nach der Messung Position R0.3 P10.4 H7.4 Load to viewer Ein Orientierung Transversal Inline movie Aus Rotation -0.33 Grad Auto store images Ein R >> L 28 mm Load to stamp segments Aus A >> P 16 mm Bilder in großes Bildsegment Aus F >> H 16 mm laden Auto open inline display Aus Physio Start measurement without Ein 1.Signal/Modus Kein further preparation Sequenz Auf Start duch Benutzer Aus Präparationsscans 4 warten Delta Frequenz -2.5 ppm Start measurements single Phasenzyklierung Auto Routine Bandbreite 1200 Hz Position R0.3 P10.4 H7.4 Messdauer 853 ms Orientierung Transversal Entferne Oversampling Ein Rotation -0.33 Grad VoI A >> P 16 mm VoI R >> L 28 mm VoI F >> H 16 mm TR 3000 ms TE 30 ms Mittelungen 80 Filter Prescan Normalisierung Spulenelemente HEA;HEP Kontrast Flipwinkel 90 Grad Wasserunterdr. Wassersättig. H2O-unterdr. Bandbr. 50 Hz Spektrale Unterdr. Keine Messungen 1 ------Auflösung Prescan Normalisierung Ein Vektorgröße 1024 ------Matrix Spulen Modus CP ------Ungefilterte Bilder Aus Geometrie System Body Aus HEP Ein HEA Ein ------Positionierungsmodus FIX Tischposition H Tischposition 0 mm MSMA S - C - T Sagittal R >> L Coronar A >> P Transversal F >> H Unkombiniert speichern Aus AutoAlign Kopf > Gehirn Autom. Spulenanwahl Default ------Shim-Modus Standard Wasserunterdr. just. Ein Mit Körperspule justieren Aus Freq. Justage bestät. Aus 29/+ SIEMENS MAGNETOM Verio syngo MR B17

\\USER\ASK\TOURETTE\AA8P150321\svs_se_35_ref_head_th TA: 0:24 VoI: 16 ×28 ×16 mm Rel. SNR: 1.00 SIEMENS: svs_se

? Ref. Amplitude 1H 0.000 V Eigenschaften Justagetoleranz Auto Prio Rekonstr. Aus Justagevolumen Vor der Messung Position R0.3 P10.4 H7.4 Nach der Messung Orientierung Transversal Load to viewer Ein Rotation -0.33 Grad Inline movie Aus R >> L 28 mm Auto store images Ein A >> P 16 mm Load to stamp segments Aus F >> H 16 mm Bilder in großes Bildsegment Aus laden Physio Auto open inline display Aus 1.Signal/Modus Kein Start measurement without Ein Sequenz further preparation Präparationsscans 4 Auf Start duch Benutzer Aus Delta Frequenz -2.5 ppm warten Phasenzyklierung Auto Start measurements single Bandbreite 1200 Hz Routine Messdauer 1706 ms Position R0.3 P10.4 H7.4 Entferne Oversampling Ein Orientierung Transversal Rotation -0.33 Grad VoI A >> P 16 mm VoI R >> L 28 mm VoI F >> H 16 mm TR 3000 ms TE 30 ms Mittelungen 4 Filter Prescan Normalisierung Spulenelemente HEA;HEP Kontrast Flipwinkel 90 Grad Wasserunterdr. Keine Spektrale Unterdr. Keine Messungen 1 ------Auflösung Prescan Normalisierung Ein Vektorgröße 2048 ------Matrix Spulen Modus CP ------Ungefilterte Bilder Aus Geometrie System Body Aus HEP Ein HEA Ein ------Positionierungsmodus FIX Tischposition H Tischposition 0 mm MSMA S - C - T Sagittal R >> L Coronar A >> P Transversal F >> H Unkombiniert speichern Aus AutoAlign Kopf > Gehirn Autom. Spulenanwahl Default ------Shim-Modus Standard Wasserunterdr. just. Aus Mit Körperspule justieren Aus Freq. Justage bestät. Aus von Silikon ausgehen Aus 27/+ SIEMENS MAGNETOM Verio syngo MR B17

\\USER\ASK\TOURETTE\AA8P150321\svs_se_30_st TA: 6:36 VoI: 20 ×15 ×20 mm Rel. SNR: 1.00 SIEMENS: svs_se

Sättigungsregion 5 Eigenschaften Dicke 20 mm Prio Rekonstr. Aus Position L1.9 A35.1 F15.8 Vor der Messung Orientierung Coronar Nach der Messung Sätt. Delta Frequenz 0.00 ppm Load to viewer Ein Sättigungsregion 6 Inline movie Aus Dicke 20 mm Auto store images Ein Position R1.2 P14.4 H0.1 Load to stamp segments Aus Orientierung C > T-0.7 > S0.3 Bilder in großes Bildsegment Aus Sätt. Delta Frequenz 0.00 ppm laden Auto open inline display Aus System Start measurement without Ein Body Aus further preparation HEP Ein Auf Start duch Benutzer Aus HEA Ein warten ------Start measurements single Positionierungsmodus FIX Tischposition H Routine Tischposition 0 mm Position L18.9 A8.9 H5.3 MSMA S - C - T Orientierung Transversal Sagittal R >> L Rotation -0.85 Grad Coronar A >> P VoI A >> P 20 mm Transversal F >> H VoI R >> L 15 mm Unkombiniert speichern Aus VoI F >> H 20 mm AutoAlign Kopf > Gehirn TR 3000 ms Autom. Spulenanwahl Default TE 30 ms ------Mittelungen 128 Shim-Modus Standard Filter Prescan Normalisierung Wasserunterdr. just. Ein Spulenelemente HEA;HEP Mit Körperspule justieren Aus Freq. Justage bestät. Aus Kontrast von Silikon ausgehen Aus Flipwinkel 90 Grad ? Ref. Amplitude 1H 0.000 V Wasserunterdr. Wassersättig. Justagetoleranz Auto H2O-unterdr. Bandbr. 50 Hz Justagevolumen Spektrale Unterdr. Keine Position L18.9 A8.9 H5.3 Messungen 1 Orientierung Transversal ------Rotation 89.15 Grad Auflösung A >> P 20 mm Prescan Normalisierung Ein R >> L 15 mm Vektorgröße 1024 F >> H 20 mm ------Matrix Spulen Modus CP Physio ------1.Signal/Modus Kein Ungefilterte Bilder Aus Sequenz Geometrie Präparationsscans 4 Sättigungsregion 1 Delta Frequenz -2.5 ppm Dicke 20 mm Phasenzyklierung Auto Position L1.7 A5.2 H27.6 Bandbreite 1200 Hz Orientierung T > C0.9 > S0.2 Messdauer 853 ms Sätt. Delta Frequenz 0.00 ppm Entferne Oversampling Ein Sättigungsregion 2 Dicke 20 mm Position R4.2 A17.1 F0.3 Orientierung S > C-1.1 > T-0.5 Sätt. Delta Frequenz 0.00 ppm Sättigungsregion 3 Dicke 20 mm Position L42.1 A16.8 H0.0 Orientierung S > T-0.4 > C-0.2 Sätt. Delta Frequenz 0.00 ppm Sättigungsregion 4 Dicke 20 mm Position L0.8 A7.6 F19.6 Orientierung T > C0.7 Sätt. Delta Frequenz 0.00 ppm 20/+ SIEMENS MAGNETOM Verio syngo MR B17

\\USER\ASK\TOURETTE\AA8P150321\svs_se_30_ref_head_st TA: 0:36 VoI: 20 ×15 ×20 mm Rel. SNR: 1.00 SIEMENS: svs_se

Dicke 20 mm Eigenschaften Position L1.9 A35.1 F15.8 Prio Rekonstr. Aus Orientierung Coronar Vor der Messung Sätt. Delta Frequenz 0.00 ppm Nach der Messung Sättigungsregion 6 Load to viewer Ein Dicke 20 mm Inline movie Aus Position R0.5 P10.8 H1.1 Auto store images Ein Orientierung C > T-0.7 > S0.3 Load to stamp segments Aus Sätt. Delta Frequenz 0.00 ppm Bilder in großes Bildsegment Aus laden System Auto open inline display Aus Body Aus Start measurement without Ein HEP Ein further preparation HEA Ein Auf Start duch Benutzer Aus ------warten Positionierungsmodus FIX Start measurements single Tischposition H Tischposition 0 mm Routine MSMA S - C - T Position L18.9 A8.9 H5.3 Sagittal R >> L Orientierung Transversal Coronar A >> P Rotation -0.85 Grad Transversal F >> H VoI A >> P 20 mm Unkombiniert speichern Aus VoI R >> L 15 mm AutoAlign Kopf > Gehirn VoI F >> H 20 mm Autom. Spulenanwahl Default TR 3000 ms ------TE 30 ms Shim-Modus Standard Mittelungen 8 Wasserunterdr. just. Aus Filter Prescan Normalisierung Mit Körperspule justieren Aus Spulenelemente HEA;HEP Freq. Justage bestät. Aus von Silikon ausgehen Aus Kontrast ? Ref. Amplitude 1H 0.000 V Flipwinkel 90 Grad Justagetoleranz Auto Wasserunterdr. Keine Justagevolumen Spektrale Unterdr. Keine Position L18.9 A8.9 H5.3 Messungen 1 Orientierung Transversal ------Rotation 89.15 Grad Auflösung A >> P 20 mm Prescan Normalisierung Ein R >> L 15 mm Vektorgröße 2048 F >> H 20 mm ------Matrix Spulen Modus CP Physio ------1.Signal/Modus Kein Ungefilterte Bilder Aus Sequenz Geometrie Präparationsscans 4 Sättigungsregion 1 Delta Frequenz -2.5 ppm Dicke 20 mm Phasenzyklierung Auto Position L1.7 A5.2 H27.6 Bandbreite 1200 Hz Orientierung T > C0.9 > S0.2 Messdauer 1706 ms Sätt. Delta Frequenz 0.00 ppm Entferne Oversampling Ein Sättigungsregion 2 Dicke 20 mm Position R4.7 A15.6 H0.7 Orientierung S > C-1.1 > T-0.5 Sätt. Delta Frequenz 0.00 ppm Sättigungsregion 3 Dicke 20 mm Position L37.5 A15.4 H1.0 Orientierung S > T-0.4 > C-0.2 Sätt. Delta Frequenz 0.00 ppm Sättigungsregion 4 Dicke 20 mm Position L0.8 A7.6 F19.6 Orientierung T > C0.7 Sätt. Delta Frequenz 0.00 ppm Sättigungsregion 5 18/+ SIEMENS MAGNETOM Verio syngo MR B17

\\USER\ASK\TOURETTE\AA8P150321\cmrr_mbep2d_resting TA: 10:01 PAT: Aus Voxelgröße: 2.3×2.3×2.3 mm Rel. SNR: 1.00 USER: cmrr_mbep2d_bold

Eigenschaften System Prio Rekonstr. Aus Body Aus Vor der Messung HEP Ein Nach der Messung HEA Ein Load to viewer Ein ------Inline movie Aus Positionierungsmodus FIX Auto store images Ein Tischposition H Load to stamp segments Aus Tischposition 0 mm Bilder in großes Bildsegment Aus MSMA S - C - T laden Sagittal R >> L Auto open inline display Aus Coronar A >> P Start measurement without Ein Transversal F >> H further preparation Kanalkombination Quadratsumme Auf Start duch Benutzer Ein AutoAlign Kopf > Gehirn Autom. Spulenanwahl Default warten ------Start measurements single Shim-Modus Standard Routine Mit Körperspule justieren Aus Schichtgruppe 1 Freq. Justage bestät. Aus Schichten 64 von Silikon ausgehen Aus Distanzfaktor 0 % ? Ref. Amplitude 1H 0.000 V Position L0.6 P10.9 H19.3 Justagetoleranz Auto Orientierung T > C-11.9 Justagevolumen Phasenkod.-Richt. A >> P Position L0.6 P10.9 H19.3 Rotation 0.00 Grad Orientierung T > C-11.9 Phasen-Oversampling 0 % Rotation 0.00 Grad FoV Auslese 202 mm R >> L 202 mm FoV Phase 100.0 % A >> P 202 mm Schichtdicke 2.30 mm F >> H 148 mm TR 1400 ms Physio TE 30.0 ms 1.Signal/Modus Kein Multi-band accel. factor 4 Filter Keine BOLD Spulenelemente HEA;HEP GLM Statistiken Aus Dynamische t-Karten Aus Kontrast Anfangsmess. ignorieren 0 MTC Aus Ignoriere nach Übergang 0 Magn. Präparation Kein Modelliere Übergänge Aus Flipwinkel 69 Grad Temp. Hochpass Filter Aus Fettunterdr. Fettsättig. Schwellwert 4.00 ------Mittelungsmodus Langzeit Paradigmengröße 3 Rekonstruktion Betrag Mess.[1] Baseline Messungen 422 Mess.[2] Baseline Verzögerung in TR 0 ms Mess.[3] Aktiv Mehrere Serien Aus Bewegungskorrektur Aus Räumlicher Filter Aus Auflösung Sequenz Basis-Auflösung 88 Phasen-Auflösung 100 % Einleitung Aus Phasen Partial Fourier 7/8 Kontraste 1 Interpolation Aus Bandbreite 1776 Hz/Px ------Flusskomp. Nein PAT Modus Keiner Freier Echoabstand Aus Matrix Spulen Modus Auto (CP) Echoabstand 0.67 ms ------Verzeichn. Korr. Aus EPI Faktor 88 Prescan Normalisierung Aus Gradientenmodus Schnell Rohdaten Ein HF-Spoiler Aus Elliptischer Filter Aus ------Hamming Aus Excite pulse duration 5120 us Single-band images Aus Geometrie MB LeakBlock kernel Aus Mehrschichtmodus Verschachtelt MB dual kernel Aus Serie Verschachtelt MB RF phase scramble Aus ------SENSE1 coil combine Aus Spez. Sättiger Keine Invert RO/PE polarity Aus 7/+ SIEMENS MAGNETOM Verio syngo MR B17

PF omits higher k-space Aus Online multi-band recon. Online FFT scale factor 1.20 Physio recording Off Triggering scheme Standard

8/+ SIEMENS MAGNETOM Verio syngo MR B17

\\USER\ASK\TOURETTE\AA8P150321\as_gre_TE17ms_nifti TA: 7:51 PAT: 2 Voxelgröße: 0.8×0.8×0.8 mm Rel. SNR: 1.00 UNBEKANNT:

Image Filter Aus Eigenschaften Verzeichn. Korr. Aus Prio Rekonstr. Aus Prescan Normalisierung Aus Vor der Messung Normalisierung Aus Nach der Messung B1-Filter Aus Load to viewer Ein Rohdaten Aus Inline movie Aus Elliptischer Filter Aus Auto store images Ein Load to stamp segments Aus Geometrie Bilder in großes Bildsegment Aus Mehrschichtmodus Verschachtelt laden Serie Verschachtelt Auto open inline display Aus ------Start measurement without Ein Sättigungsmodus Standard Spez. Sättiger Keine further preparation ------Auf Start duch Benutzer Aus ------warten Tim CT Modus Aus Start measurements single System Routine Body Aus 3D-Block-Gruppe 1 HEP Ein 3D-Blöcke 1 HEA Ein ------Distanzfaktor 20 % Positionierungsmodus REF Position R0.5 P19.7 H36.7 Tischposition H Orientierung T > C-25.0 Tischposition 0 mm Phasenkod.-Richt. R >> L MSMA S - C - T Rotation 90.00 Grad Sagittal R >> L Phasen-Oversampling 0 % Coronar A >> P Schicht-Oversampling 10.0 % Transversal F >> H Schichten im 3D-Block 160 Unkombiniert speichern Aus FoV Auslese 205 mm Kanalkombination Adaptive Combine FoV Phase 81.3 % AutoAlign Kopf > Basis Schichtdicke 0.80 mm Autom. Spulenanwahl Aus TR 30 ms ------TE 17.3 ms Shim-Modus Standard Mittelungen 1 Mit Körperspule justieren Aus Verknüpfungen 1 Freq. Justage bestät. Aus Filter Keine von Silikon ausgehen Aus Spulenelemente HEA;HEP ? Ref. Amplitude 1H 0.000 V Justagetoleranz Auto Kontrast Justagevolumen MTC Aus Position R0.5 P19.7 H36.7 Magn. Präparation Kein Orientierung T > C-25.0 Flipwinkel 13 Grad Rotation 90.00 Grad Fettunterdr. Keine A >> P 205 mm Wasserunterdr. Keine R >> L 167 mm SWI Aus F >> H 128 mm ------Mittelungsmodus Kurzzeit Physio Rekonstruktion Betrag/Phase 1.Signal/Modus Kein Messungen 1 Segmente 1 Mehrere Serien Jede Messung ------Dark Blood Aus Auflösung ------Basis-Auflösung 256 Atemkontrolle Aus Phasen-Auflösung 100 % Inline Schicht-Auflösung 100 % Subtrahieren Aus Phasen Partial Fourier 6/8 Leber Registrierung Aus Schicht Partial Fourier Aus Std-Abw.-Sag Aus Interpolation Aus ------Std-Abw.-Cor Aus PAT Modus GRAPPA Std-Abw.-Tra Aus Beschl. Faktor PE 2 Std-Abw.-Zeit Aus Ref. Zeilen PE 24 MIP-Sag Aus Beschl. Faktor 3D 1 MIP-Cor Aus Matrix Spulen Modus Auto (Triple) MIP-Tra Aus Referenzmessungsmodus Integriert MIP-Zeit Aus ------Originalbilder speichern Ein 30/+ SIEMENS MAGNETOM Verio syngo MR B17

------Wash - In Aus Wash - Out Aus TTP Aus PEI Aus MIP-Zeit Aus Sequenz Einleitung Ein Dimension 3D Elliptische Abtastung Aus Phasenstabilisierung Ein Asymmetrisches Echo Aus Kontraste 1 Bandbreite 150 Hz/Px Flusskomp. Ja Erlaubte Verzögerung 0 s ------HF-Puls-Typ Normal Gradientenmodus Flüster Anregung 3D-Block sel. HF-Spoiler Ein ------length exc pulse 4000 us Ernst Angle? Ein T1 1200 ms FFT scale factor 1.0

31/+

Appendix E

Author portfolio

• Curriculum vitae

213 Publications Peer-reviewed Journal Articles

Kanaan A.S.*, Gerasch S., Garcia-Garcia I., Lampe L., Pampel A, Anwander A, Near J., Möller H.E.**, Müller-Vahl K.R.** Pathological glutamatergic neurotransmission in Gilles de la Tourette Syndrome. Brain 2017; 140: 218–234

Kanaan A.S.*, Jakubovski E.*, Müller-Vahl K.R. Signiﬁcant Tic Reduction in An Otherwise Treatment-Resistant Patient with Gilles de la Tourette Syn- drome Following Treatment with Nabiximols Brain Sciences 2017; 7(5): 47

Forde NJ*, Kanaan A.S.*, Widomska J., Padmanabhuni S.S.,Nespoli E., Alexander J, Arranz J.I.R, Fan S., Houssari R., Nawaz M., Zilhão N, Pagliaroli L. Rizzo F., Aranyi T., Barta C., Boeckers T., Boomsma D.,Buisman W., Buitelaar J.K.,Cath D., Dietrich A., Driessen N., Drineas P., Dunlap M., Gerasch S., Glennon J., Hengerer B., van den Heuvel O., Jespersgaard C., Möller H.E., Müller-Vahl K.R., Openneer T., Poelmans G., Pouwels P.J., Scharf J, Stefansson H., Tümer Z., Veltman D., van der Werf Y.D., Hoekstra P., Ludolph A. and Paschou P. TS-EUROTRAIN: A European-wide investigation and training network on the aetiology and pathophysiology of Gilles de la Tourette Syndrome.Frontiers in Neuroscience 2016; 10: 1–9. (Shared ﬁrst authorship).

Gerasch S., Kanaan A.S., Jakubovski E. and Müller-Vahl K.R. Aripiprazole improves associated comorbid Conditions in addition to Tics in adult Patients with Gilles de la Tourette Syndrome.Frontiers in Neuroscience 2016; 10: 416.

Kanaan, A. S.*, Frank, F., Maedler-Kron, C., Verma, K., Sonenberg, N., Nagar, B. Crystallization and preliminary X-ray diﬀraction analysis of the mid- dle domain of Paip1. Acta Crystallographica Section F: Structural Biology and Crystallization Communications 65 (10), S. 1060-1064, 2009.

Manuscripts under review

Kanaan A.S.*, Anwander A., Metere, R, Torsten S., Jamie N., Bilgic B., Müller-Vahl K.R.**, Möller H.E.** Disturbed iron homeostasis in Gilles de la Tourette syndrome. Under review.

Kanaan A.S.*, Metere R.*, Bilgic B., Torsten S., Möller H.E. Eﬀects of the coil combination algorithm on quantitative susceptibility mapping. Under review.

Book Chapters

Kanaan A.S.* and Müller-Vahl K.R. Cannabinoid-Based Medicines for the Treatment of Gilles de la Tourette Syndrome. In Press, Handbook of Cannabis and Related Pathologies, Elsevier, 2017.

de la Iglesia-Vaya, M., Molina-Mateo, J., Escarti-Fabra, J., Kanaan, A. S., Martí- Bonmatí, L.: Brain connections – Resting state fMRI functional connectivity. In: Novel Frontiers of Advanced Neuroimaging, S. 51-66, 2013 Conference Presentations

Kanaan A.S., Anwander A., Schäfer A., Bilgic B., Müller-Vahl K.R.,and Möller H.E. Putative neurochemical abnormalities inﬂuenced by subcortical iron de- ﬁciency in Tourette syndrome. The Organization for Human Brain Mapping. Vancouver, BC, Canada, June 25-29, 2017

Kanaan A.S., Müller-Vahl K.R.,and Möller H.E. Towards robust, large-scale and collaborative investigation of the macroscale cerebral architecture of Gilles de la Tourette syndrome pathophysiology. European Society for the study of Tourette Syndrome (ESSTS) 2017 Annual Meeting. Sevilla, Spain. June 14-16 2016.

Kanaan A.S., Anwander A., Schäfer A., Metere R., Schlumm T., Near J., Bilgic B., Müller-Vahl K.R.,and Möller H.E. QSM meets MRS: The inﬂuence of subcortical iron on glutamatergic neurotransmission in a movement disorder population International Society for Magnetic Resonance Imaging in Medicine (ISMRM), Abstract # 4649, Honolulu, Hawaii, USA, April 22 - April 27, 2017.

Metere R., Kanaan A.S., Schäfer A., Bilgic B., Schlumm T., and Möller H.E. Eﬀects of diﬀerent coil reconstruction algorithms on Quantitative Susceptibility Mapping results. 4th International Workshop on MRI Phase Contrast & Quanti- tative Susceptibility Mapping. Graz, Austria. September 26-28, 2016.

Kanaan A.S., Schäfer A, Bilgic, B. Müller-Vahl K.R., Möller H.E. Quantitative Susceptibility Mapping in Gilles de la Tourette Syndrome. European Society for the study of Tourette Syndrome (ESSTS) 2016 Annual Meeting. Warsaw, Poland. June 8-11 2016.

Kanaan A.S.*, Gerasch, Garcia-Garcia I., Lampe L., Pampel A, Anwander A, Near J, Möller HE*, Müller-Vahl K*. Pathological glutamatergic neurotransmission in Gilles de la Tourette Syndrome. 24th Annual Meeting of the International Society for Magnetic Resonance in Medicine (ISMRM), Abstract # 2420, Singapore, Singapore, May 7 - 13, 2016

Kanaan A.S.*, Pampel A., Müller-Vahl K.R., Möller H.E. Test-Retest Quanti- tation of Absolute Metabolite Concentrations with Partial-Volume correction using diﬀerent segmentation methods. International Society for Mag- netic Resonance Imaging in Medicine (ISMRM), Abstract # 1974, Toronto, Ontario, Canada, May 30 - June 5, 2015.

Kanaan A.S.*, Margulies D.S., Anwander A., Möller H.E., Müller-Vahl K.R. Retro- spective control for motion-artefacts in functional neuroimaging datasets using Wavelet and ICA based methods , 1st World Congress on Tourette Syn- drome & Tic Disorders, Abstract # 202, 1st Tourette World Congress. London, UK. June 24-26 2015.

Kanaan A.S.* and Müller-Vahl K, Signiﬁcant tic reduction in a treatment resistant Gilles de la Tourette Syndrome patient following treatment with nabiximols (Sativex R ). 1st World Congress on Tourette Syndrome & Tic Dis-

orders, Abstract # 156. 1st Tourette World Congress. London, UK. June 24-26 2015. Kanaan A.S.*, Forde N.J., et al. TS-EUROTRAIN: European wide investigation of the etiology and pathophysiology of Gilles de la Tourette syndrome and related disorders. 1st World Congress on Tourette Syndrome & Tic Disorders, Abstract # 111. London, UK. June 24-26 2015.

Kanaan A.S.*, Gerasch S., Pampel A., Lampe L., Schäfer A., Margulies D., Möller H.E., Müller-Vahl K.R. Elemental, neurochemical and network based analysis of the pathophysiological mechanisms of Gilles de la Tourette Syndrome. European Society for the study of Tourette Syndrome (ESSTS) 2014 Annual Meeting. Paris, France. April 25-26 2014.

Moreno-Dominguez, D., Watanabe, A., Gorgolewski, K. J., Schäfer, A., Goulas, A., Kipping, J., Kanaan, A. S., Anwander, A., Toro, R., Margulies, D. S. Multi-modal parcellation of the frontal lobe. Proceedings of the 20th Annual Meeting of the Organization for Human Brain Mapping. Hamburg, Germany, June 8 - 12, 2014.

Oral European Society for the Study of Tourette Syndrome (ESSTS) Conference, 2016 Presentation Pathological glutamatergic neurotransmission in Gilles de la Tourette Syn- drome. Warsaw, Poland. June 8-11 2016.

Kongress der Deutschen Gesellschaft für Neurologie (DGN) Magnetic resonance spectroscopy in Gilles de la Tourette syndrome. Düsseldorf, Germany, September 24, 2015

Awards Professor Mary Robertson Award for distinguished research contribution European Society for the Study of Tourette Syndrome (ESSTS) Conference Warsaw, Poland, June 10, 2016.

Best Poster Presentation Award International Max Planck Research School (NEUROCOM) Summer School Leipzig, Germany, July 6, 2016.

Marie Curie Doctoral Fellowship Award, 2013 European Commission - 7th Framework Programme.

American Crystollographic Association Travel Award Groupe De Recherche Axé Sur La Structure Des Protéines, July, 2009, Montreal, Canada.

Memberships Int. Society for Magnetic Resonance Imaging in Medicine (ISMRM): 2014-present European Society for the Study of Tourette Syndrome (ESSTS): 2014 - present Organization for Human Brain Mapping (OHBM): 2014-present American Crystallographic Association (ACA): 2009-2010

Academic Ad-hoc reviewer Service Neuroimage, Frontiers in Neuroscience, Frontiers in Psychiatry