Final Thesis File

THE UNIVERSITY OF MELBOURNE

TREMOR IN MULTIPLE SCLEROSIS

Neuroimaging perspective

by Frederique Maria Christina Boonstra

ORCID: 0000-0002-5861-4174

October 2019

A thesis presented to the University of Melbourne in fulfilment of the degree of Doctor in Philosophy in Medicine, Health and Dentistry Science

Abstract Multiple sclerosis (MS) is a common autoimmune disorder of the central nervous system (CNS), characterised by inflammation, demyelination and neurodegeneration. The clinical presentation and disease course of MS is heterogeneous, which reflects the multifocal nature of damage within the CNS. Almost half of the MS population experiences a tremor in the later stages of the disease. Tremor significantly increases disease severity and worsens patients’ quality of life. Current understanding of tremor pathophysiology in MS is incomplete and mostly based on treatment studies, clinical observation studies, or neuroimaging studies of parkinsonian tremor and essential tremor. Focused imaging assessments of defined neural pathways associated with tremor can help improve our understanding of complex pathophysiology of MS tremor. This could benefit the current lack of effective, noninvasive and long-term treatment options for tremor.

This thesis provides a comprehensive examination of the pathophysiology of tremor in MS. The first experimental chapter confirms the hypothesis that the cerebello-thalamo-cortical tract is involved in tremor pathophysiology. Specifically, the pilots study finds a positive correlation between thalamic and superior cerebellar peduncle atrophy and tremor severity. In the second experimental chapter, we aimed to develop a functional imaging task that will allow in vivo imaging of tremor pathology. We introduced a novel joystick task, which showed to elicit the MS-related tremor while playing the game. Furthermore, we showed good reproducibility, which is great for the longitudinal part of this thesis. In the third experimental chapter, we applied the joystick imaging task in a large sample of tremor and non-tremor MS patients. We supported that pathology along the cerebello-thalamo-cortical tract is instrumental in tremor pathology. Interestingly, we also found increased functional activation within sensorimotor integration and motor planning areas in MS tremor, which negatively correlated to tremor severity. In the final experimental chapter, we examined the central effects of onabotulinumtoxinA (BoNT-A) for the treatment of tremor in a randomized controlled trial. We found that patients that received BoNT-A had improved tremor and reduced activation within the sensorimotor integration regions. The change in tremor severity correlated with the change in activation, indicated that BoNT-A has a central effect as well as a local effect.

Collectively, these findings suggest that the clinical presentation of tremor in MS is influenced by a tremor network consisting of both structural and functional aspects. Specifically, atrophy and inflammation along the cerebello-thalamo-cortical tract is thought to be more causal to tremor. Contrarily, functional activation is thought to be compensatory to alleviate tremor severity. Intramuscular injection of BoNT-A reduced tremor severity and the activation within the sensorimotor integration area. The central effect of BoNT-A is thought to be due to the lower need for the compensatory functional activation. Together, both structural and functional aspects of the tremor network in MS need to be consider when trying to monitor tremor over time and to find effective treatments options.

Declaration

I, Frederique Maria Christina Boonstra, certify that:

i. This thesis comprises only my original work towards the PhD. ii. Due acknowledgements have been made in the text to all other material used. iii. This thesis is less than 100,000 words in length, exclusive of tables, figures, and bibliographies.

Frederique Maria Christina Boonstra

iii

Preface The author of this thesis has performed the extensive review of literature and research involved in all of the experimental chapters. Work was supervised by Dr Scott Kolbe and Dr Anneke van der Walt.

The first experimental chapter is part of a randomized, double-blind clinical trial with a crossover design (clinical trial identifier: NCT01018485). Funding for this project was provided by the Box Hill MS Research Unit. All patient recruitment was complete before enrolment in the PhD, but all analyses, interpretation and writing has been done afterwards. Grace Florescu contributed to data analyses, interpretation and writing the manuscript. Dr Scott Kolbe and Dr Anneke van der Walt both contributed to interpretation and writing of the manuscript. Other co-authors provided general comments and/or helped with the project at large.

The second, third and fourth experimental chapter are part of a large randomized controlled clinical trial studying the efficacy of onabotulinumtoxinA for MS-related tremor (ACTRN12617000379314). This project was funded by a National Health and Medical Research Council (NHMRC) Project Grant (1085461 CIA Van der Walt) and Fellowship (1135683 Vogel). The Bionics Institute receives Operational Infrastructure Support from the Victorian Government. This trial started in early 2015 before enrolment in the PhD program; however, all participants included in the presented work were recruited after enrolment. All other aspects of work towards this thesis have been completed after enrolment. Dr Anneke van der Walt is the principal investigator of the study and was of great importance in the participant recruitment, data collection, data interpretation and writing on the manuscripts. Dr Scott Kolbe has significantly contributed to data analyses, interpretation and editing of the manuscripts. Dr Andrew Evans has contributed to the study design, data interpretation and writing of the manuscripts. All other co-authors of respective manuscripts provided general comments and/or helped with the project at large.

None of this work has been submitted for other qualifications.

None of this work was carried out prior to this PhD.

Publications, Presentations and Scholarships Published journal articles Motor speech signature of behavioral variant frontotemporal dementia: refining the phenotype. Vogel, A. P., Poole, M. L., Pemberton, H., Caverlé, M. W. J., Boonstra, F. M. C., Low, E., Darby, D., Brodtmann, A. Neurology (2017)

Tremor in multiple sclerosis is associated with cerebello-thalamic pathology. Boonstra, F. M. C., Florescu, G., Evans, E. H., Steward, C., Mitchell, P., Desmond, P., Moffat, B. A., Butzkueven, H., Kolbe, S. C., Van der Walt, A. Journal of Neural Transmission (2017)

Technologies for Advanced Gait and Balance Assessments in People with Multiple Sclerosis. Boonstra, F. M. C., Shanahan, C. J., Cofré Lizama, L. E., Strik, M., Moffat, B. A., Khan, F., Kilpatrick, T. J., Van der Walt, A., Galea, M. P., Kolbe, S. C. Frontiers in Neurology (2018)

Novel Functional MRI Task for Studying the Neural Correlates of Upper Limb Tremor. Boonstra, F. M. C., Perera, T., Noffs, G., Marotta, C., Vogel, A. P., Evans, A. H., Butzkueven, H., Moffat, B. A., Van der Walt, A., Kolbe, S. C. Frontiers in Neurology (2018)

Validation of a precision tremor measurement system for multiple sclerosis. Perera, T., Lee, W., Yohanandan, S. A. C., Nguyen, A., Cruse, B., Boonstra, F. M. C., Noffs, G., Vogel, A. P., Kolbe, S. C., Butzkueven, H., Evans, A. H., Van der Walt, A. Journal of Neuroscience Methods (2018)

What speech can tell us: A systematic review of dysarthria characteristics in Multiple Sclerosis. Noffs, G., Perera, T., Kolbe, S. C., Shanahan, C. J., Boonstra, F. M. C., Evans, A. H., Butzkueven, H., Van der Walt, A., Vogel, A. P. Autoimmunity Reviews (2018)

Functional neuroplasticity in response to cerebello-thalamic injury underpins the clinical presentation of tremor in multiple sclerosis. Boonstra, F. M. C., Noffs, G., Perera, T., Jokubaitis, V. G., Vogel, A. P., Moffat, B. A., Butzkueven, H., Evans, A. H., Van der Walt, A., Kolbe, S. C. Multiple Sclerosis Journal (in press)

Journal articles under review Reduced fMRI response in central sensory-motor integration areas following intramuscular onabotulinumtoxinA for tremor in multiple sclerosis. Boonstra, F. M. C., Evans, A. H., Noffs, G., Perera, T., Jokubaitis, V. G., Stankovic, J., Vogel, A. P., Moffat, B. A., Butzkueven, H., Kolbe, S. C., Van der Walt, A. Manuscript submitted for publication.

Conference abstracts The thalamus as a neural correlate for tremor in MS. Florescu G., Boonstra F. M. C., Mitchell, P., Steward C., Kolbe, S. C., Van der Walt A. Australian and New-Zealand association of neurologists (2016)

The cerebello-thalamic tract as a neural correlate for tremor in MS. Boonstra F. M. C., Florescu G., Kolbe S. C., Steward C., Evans A. H., Butzkueven H., Mitchell P., Van der Walt A. European conference of treatment and intervention in multiple sclerosis (ECTRIMS) (2016)

Pathophysiology of MS tremor: an fMRI study. Boonstra, F. M. C., Noffs, G., Perera, T., Shanahan, C. J., Vogel, A. P., Evans, A. H., Butzkueven, H., Van der Walt, A., Kolbe, S. C. International Society for Magnetic Resonance in Medicine (2017), ECTRIMS (2017) & Multiple sclerosis research Australia (2017)

Scholarships Melbourne International Research Scholarship Melbourne International Fee Remission Scholarship

vii

Acknowledgements I would like to thank my family and friends, who have shown nothing but support over the last three to four years. To my partner in life, Martin, who has supported me throughout this PhD and without whom I could never have completed this or any other project. To my family, Renger, Heleen, Matthijs and Coen, thank you for support of me through my whole life, even if it brought me far away from my family home. To my friends, who have helped me make a life in Melbourne, and gave me the much-needed distraction with tennis, games and exploring Melbourne.

I would like to greatly thank my supervisors, A/Prof Anneke van der Walt, Dr. Scott Kolbe and A/Prof Brad Moffat. Anneke and Scott, you have shown tremendous believe in me and given me the opportunity to keep working with both of you. Furthermore, you both have provided great guidance, and taught me a great deal about the MS field and all the aspects involved in project management and research. To Brad, you have been instrumental for me in starting my PhD and have been a supports person throughout my PhD.

I would like to thank all colleagues from the different groups I have been a part of. To the TremTox team, for their knowledge and insight into this field of research. To all the imaging radiographers at the Murdoch Child Research Institute, for being incredibly helpful and organized, while at the same time keeping me entertained through hours and hours of scanning. To my colleagues in MBCIU, for creating an amazing environment to work in. Finally, a special thanks to Sanuji and Myrte who are not only my colleagues but have become great friends.

Finally, I would like to thank the participants who have been instrumental in completing these studies and from who taught me a great deal about generosity and reliability.

viii

Table of Content

Abstract ...... i

Declaration ...... iii

Acknowledgements ...... viii

Table of Content ...... ix

List of Figures and Tables ...... xiv

List of Abbreviations ...... xvi 1 Background and Motivation ...... 1

1.1. Multiple sclerosis ...... 2 1.1.1. Epidemiology and phenotypes ...... 2 1.1.2. Clinical course of multiple sclerosis ...... 2 1.1.3. Symptomatology of multiple sclerosis ...... 3 1.1.4. Prognosis of multiple sclerosis ...... 4 1.1.5. Diagnosis of multiple sclerosis ...... 5 1.1.6. Pathology of multiple sclerosis ...... 5 Inflammation ...... 6 Demyelination ...... 7 Neurodegeneration ...... 9 Lesions ...... 10 Clinical stages pathology ...... 12 1.1.7. Conclusion ...... 13

1.2. Tremor in multiple sclerosis ...... 13 1.2.1. Tremor presentation in multiple sclerosis ...... 13 1.2.2. Tremor Assessment in multiple sclerosis ...... 14 Subjective rating scales ...... 15 Objective performance tests ...... 16 1.2.3. Treatment of tremor ...... 18 Non-pharmaceutical treatments ...... 18 Pharmaceutical treatments ...... 19 Surgical treatments ...... 22 1.2.4. Conclusion ...... 22

1.3. Pathology of tremor in multiple sclerosis ...... 23 1.3.1. Pathophysiological insights non-imaging studies ...... 23

1.3.2. Structural imaging ...... 25 The basics of structural imaging ...... 25 Detection of lesion in multiple sclerosis ...... 26 Detection of volumetric changes in multiple sclerosis ...... 27 Structural imaging in multiple sclerosis tremor ...... 28 1.3.3. Functional imaging ...... 28 The basics of functional imaging ...... 29 Functional imaging of multiple sclerosis tremor ...... 30 1.3.4. Conclusion ...... 30

1.4. Summary/motivation and thesis overview ...... 30

1.5. References ...... 32 2 Project Design ...... 48

2.1. Chapter 3 ...... 48

2.2. Chapters 4-6 ...... 48 2.0.1. Investigators ...... 48 2.0.2. Overall study aims ...... 49 2.0.3. Participants recruitment ...... 49 2.0.4. Ethics ...... 50 2.0.5. Inclusion criteria ...... 50 2.0.6. Exclusion criteria ...... 51 2.0.7. Contribution ...... 51

2.3. References ...... 52 3 Pilot Study ...... 53

Tremor in multiple sclerosis is associated with cerebello-thalamic pathology...... 53

3.1. Abstract ...... 54

3.2. Introduction ...... 54

3.3. Methods ...... 56 Participants ...... 56 Clinical assessments ...... 56 Neuroimaging ...... 56 Statistical Analyses ...... 58

3.4. Results ...... 58

3.5. Discussion ...... 61 x

3.6. Author’s roles ...... 62

3.7. Financial Disclosures of all authors ...... 63

3.8. References ...... 63 4 Validation Study ...... 67

Novel functional MRI task for studying the neural correlates of upper limb tremor ...... 67

4.1. Abstract ...... 68

4.2. Introduction ...... 68

4.3. Methods ...... 70 Participants ...... 70 MRI acquisition ...... 70 Joystick task ...... 71 Joystick task analysis ...... 72 Imaging analysis ...... 73 Reproducibility ...... 73

4.4. Results ...... 74 Head movement ...... 74 Validation of tremulous motion during task performance in tremor patients ...... 74 Task analyses ...... 75 Reproducibility ...... 78

4.5. Discussion ...... 78

4.6. Acknowledgement & funding ...... 80

4.7. References ...... 81

4.8. Supplementary table ...... 85 5 Pathophysiology Study ...... 87

Functional neuroplasticity in response to cerebello-thalamic injury underpins the clinical presentation of tremor in multiple sclerosis ...... 87

5.1. Abstract ...... 88

5.2. Introduction ...... 88

5.3. Methods ...... 89 Study design ...... 89 xi

Participants ...... 90 Standard protocol approvals, registrations, and patient consents ...... 90 Imaging protocols ...... 90 Joystick task ...... 91 Joystick task analysis ...... 91 Functional imaging analyses ...... 91 Structural imaging analyses ...... 92 Statistical analyses ...... 93

5.4. Results ...... 93 Demographic and clinical data ...... 93 Structural group differences ...... 94 Tremulous movement during task ...... 96 Functional group differences ...... 96 Functional task activations and tremor severity ...... 98 MRI and general disability in patients with tremor ...... 99

5.5. Discussion ...... 99

5.6. Author contributions ...... 102

5.7. Acknowledgement ...... 103

5.8. Disclosures ...... 103

5.9. Study funding ...... 104

5.10. References ...... 104

5.11. Supplementary figures ...... 108 6 Treatment Effect Study ...... 111

OnabotulinumtoxinA treatment for MS-tremor modifies fMRI tremor response in central sensory-motor integration areas ...... 111

6.1. Abstract ...... 112

6.2. Introduction ...... 112

6.3. Methods ...... 113 Study design ...... 113 Participants ...... 114 Randomization and BoNT-A injections ...... 114 Clinical assessments ...... 114

xii

Strength measurements ...... 114 Tremor measurement ...... 114 Imaging protocols ...... 115 Joystick fMRI task ...... 115 Imaging analyses ...... 115 Statistical analyses ...... 116

6.4. Results ...... 117 Participants ...... 117 Effects of BoNT-A treatment on MS-related tremor ...... 117 MRI structural and functional measures and tremor improvement after BoNT-A ...... 119 Functional neuroplasticity in response to BoNT-A ...... 119

6.5. Discussion ...... 120

6.6. Authors contributions ...... 123

6.7. Acknowledgement ...... 123

6.8. Disclosures ...... 123

6.9. Study funding ...... 124

6.10. References ...... 124

6.11. Supplementary tables ...... 127 7 General Discussion ...... 131

7.1. Summary of findings ...... 131

7.2. Integration into the literature ...... 133

7.3. Contributions to research and clinical treatment and management ...... 134

7.4. Limitations and Future directions ...... 135

7.5. References ...... 137 8 Conclusion ...... 142 9 Appendices ...... 143

2.1. Appendix 1. Bain Tremor Rating Score: ...... 143

2.2. Appendix 2. Bain sheet: ...... 145

xiii

List of Figures and Tables

Chapter 1 Figure 1. The clinical stages of MS. Figure 2. Inflammatory processes in MS. Figure 3. Schematic view of a neuron. Figure 4. Mechanisms of neurodegeneration in MS. Figure 5. Schematic view of temporal evolution of MS lesions. Figure 6. Mechanisms of neuromuscular blockade by botulinum toxin. Figure 7. Anatomical overview of the cerebello-thalamo-cortical tract. Figure 8. Different MRI sequences to visualise MS lesions.

Chapter 2 Figure 1. Participant groups overview. Table 1. Visit protocol

Chapter 3 Figure 1. Alignment and manual labelling of the SCP. Table 1. Patient demographics. Table 2. Correlation between tremor and dystonia severity scores and MRI measures.

Chapter 4 Figure 1. The joystick used outside the scanner (A) and inside the scanner (B) Figure 2. The tremor-like movement during MOVE (top) and PLAY (bottom) for both healthy controls and tremor patients. Figure 3. Statistical parametric maps for the PLAY > WATHC (top), PLAY > MOVE (middle) and PLAY > WATCH + MOVE (bottom) contrasts, z-stat threshold of >2.3.

Chapter 5 Table 1. Demographic and clinical tremor data for multiple sclerosis participants with tremor and without tremor Table 2. Volumetric and lesion load data for multiple sclerosis participants with tremor and without tremor

xiv

Figure 1. Tremor during fMRI task. Figure 2. Activation during fMRI task.

Chapter 6 Table 1. Participants’ demographics and disease characteristics at baseline. Table 2. Baseline strength, tremor severity scores and fMRI activation with posttreatment change at 6 and 12 weeks. Figure 1. The relation between the change in tremor and the change in functional activation within the ipsilateral inferior parietal lobule (IPL).

List of Abbreviations

ANOVA analysis of variance ANTs advanced normalisation tool BBB blood-brain barrier BOLD blood-oxygen-level dependent BoNT-A botulinum toxin type a BW bandwidth CIS clinically isolated syndrome CNS central nervous system CoV coefficient of variation CSF cerebrospinal fluid CTC cerebello-thalamo-cortical tract DBS deep brain stimulation DIR double inversion recovery DMT disease modifying therapy EDSS expanded disability status scale EEG electroencephalogram EMG electromyography ET essential tremor ETL echo train length FEAT fMRI expert analysis tool FLAIR fluid-attenuated inversion recovery fMRI functional magnetic resonance imaging FMRIB functional magnetic resonance imaging of the brain FOV field of view FSL FMRIB software library GDS global dystonia scale GLM general linear model GM grey matter 9-HPT nine-hole-peg-test ICC intercorrelation coefficient ICV intracranial volume IPL inferior parietal lobule

xvi

IQR interquartile range LST lesion segmentation tool MPRAGE magnetization-prepared rapid acquisition with gradient echo MR magnetic resonance MRC medical research council MRI magnetic resonance imaging MS multiple sclerosis PD Parkinson’s disease PPMS primary progressive multiple sclerosis RF radiofrequency RRMS relapsing-remitting multiple sclerosis RNS reactive nitrogen species ROS reactive oxygen species SARA scale for the assessment and rating of ataxia SCP superior cerebellar peduncle SMA supplementary motor area SNARE soluble N-ethylmaleimide–sensitive factor attachment protein receptor SNAP-25 synaptosomal nerve-associated protein 25 SPMS secondary progressive multiple sclerosis SPSS statistical package for the social sciences T1 a rate constant relation to the longitudinal recovery of magnetisation T2 a rate constant relation to the transverse recovery of magnetisation TE echo time TI inversion time TR repetition time WHO world health organisation WM white matter

xvii 1

Background and Motivation

Multiple sclerosis (MS) is the leading cause of neurological disability in young adults, and comes with a life-long disability and significant cost to families and the community (Browne et al., 2014). This chronic disease is characterized by a widespread distribution of lesions, resulting in a highly variable symptomatology and ergo the term ‘a disease of a thousand images’ (Heesen et al., 2008). One of the most common symptoms is tremor, which is a very disabling and embarrassing symptom, for which little treatment is available. Significant progress is needed to help understand tremor and improve patients’ quality of life.

This thesis is part of a double-blind randomised controlled clinical trial, TREMTOX, whose primary aim is to provide class I evidence for the treatment of tremor in MS using botulinum toxin type A (BoNT-A). Additional aims of this project are to improve assessment of tremor in the clinic using electrophysiological measurements, to characterise the network associated with tremor in MS and finally to assess how BoNT-A affects the tremor network and identify baseline parameters that predict response to BoNT-A. This thesis will focus on the two latter aims. The first part of this thesis (Chapter 1: sections 1- 5), provides the background and motivation of the study including an overview of what is currently known about MS and its pathology in general, followed by a more detailed treatment of tremor specifically including its clinical presentation in MS, the means by which it is assessed, current treatments available, and a detailed overview of what is currently known about the pathophysiology of MS tremor.

1.1. Multiple sclerosis

1.1.1. Epidemiology and phenotypes Multiple sclerosis is the most common demyelinating disease, affecting 2.3 million people worldwide according to the world health organisation (WHO) (WHO, 2013). The prevalence is region and gender dependant. North America and Europa have the highest prevalence and MS is twice more common in women compared to men. The reason for the difference in regions and gender remains poorly understood and genetic and environmental factors are thought to play a role (Handel, Handunnetthi, Giovannoni, Ebers, & Ramagopalan, 2010). The average age of onset of MS is early (30 years) (Confavreux & Vukusic, 2008), however, the life expectancy of MS patients is only slightly lower than the healthy population (Kingwell et al., 2012; Lunde, Assmus, Myhr, Bo, & Grytten, 2017). Thus many patients remain affected by the disease for the majority of their adult life which results in an enormous social and economic impact for patients, their families and carers (WHO, 2013).

1.1.2. Clinical course of multiple sclerosis The course is divided into four clinical stages of MS (figure 1). There are two stages prior to clinically definite MS. There is radiological evidence to support a preclinical stage of MS, named radiologically isolated syndrome (Lebrun, Kantarci, Siva, Pelletier, & Okuda, 2018). The first clinical presentation of MS is known as clinically isolated syndrome (CIS). CIS patients present with their first episode of focal neurological symptom(s), lasting at least 24 hours, which represents an area of inflammation or demyelination in the optic nerve, brain (Marcus & Waubant, 2013) or spinal cord (Arrambide et al., 2018). The first relapse following CIS denoting the transition from CIS to relapsing-remitting MS (RRMS). Diagnosis of RRMS can also be made substituting dissemination in time and space using imaging features of lesion distribution and the development of new lesions (Polman et al., 2011). RRMS is the most common type of MS with 85% of patients diagnosed with this phenotype. Clinical relapses are followed by periods of remission (periods of partial or complete recovery) and no progression between relapses. Progression of disability still occurs by incomplete recovery after relapse, and as mentioned above this is a sign of poorer prognosis and earlier development of secondary progressive MS (SPMS) (Novotna et al., 2015). In historical epidemiology studies, 80% of all RRMS patients eventually lose

the stable periods between relapses and show irreversible progression, which marks the transition to SPMS. One third of patients with SPMS experience relapses and the disease course can plateau for period of stability and then further progression.

In addition, about ten to fifteen percent of patients present with primary-progressive MS (PPMS). PPMS is characterized by a steady progression of disability from the onset of symptoms without prior periods of relapses or remission (Lublin et al., 2014; Lunde et al., 2017; Thompson et al., 2018). Fluctuations of progression rate, periods of stable disease and relapse might still occur.

1.1.3. Symptomatology of multiple sclerosis The clinical course of MS is heterogenous and difficult to predict on the individual level. The variability in symptomatology is already presents at onset. The initial symptom or episode that brings people to the neurologist is most often a motor or sensory problem. Specifically, typical presentation of patients are unilateral optic neuritis, focal brainstem or cerebellar syndrome or partial myelopathy (D. H. Miller et al., 2008; Thompson et al., 2018). While the motor and sensory symptoms are more likely to prompt a consultation with a neurologist, others such as fatigue are more difficult to measure and therefore may go unrecorded. Similarly, symptoms related to urinary and sexual issues may be less likely to be discussed by people with MS in some cultures.

Over time, the symptomatology of MS includes the following domains in order of most common: sensory, motor, visual, fatigue, balance, sexual, urinary, pain and cognition (WHO, 2013). Even though not all symptoms present in the early stages of MS, all domains of disability worsen with disease progression (Kister et al., 2013). Interestingly, the early symptomological phenotype of the patient is thought to have prognostic value. Several studies found that patients with episodes that impact bowel/bladder and pyramidal (voluntary control of body and face muscles) functions have a poorer long-term disability outcome (Amato, Ponziani, Bartolozzi, & Siracusa, 1999; Kantarci et al., 1998; Leone et al., 2008; Riise et al., 1992; Simone et al., 2002). Contrarily, episodes resulting in optic neuritis were associated with relative favourable long-term disability outcome (Baghizadeh, Sahraian, & Beladimoghadam, 2013; Leone et al., 2008; Riise et al., 1992). Finally, episodes affecting the cerebellar (symptoms include ataxia, loss of coordination or tremor) and

brainstem (symptoms include problems with speech, swallowing and nystagmus) functions have mixed disability outcome (Langer-Gould et al., 2006).

Figure 1. The clinical stages of MS. CIS patients present with one isolated symptom that can coincide with radiological activity. Patients with RRMS presents with multiple relapses with stable periods between relapses. After RRMS, 80% progress to SPMS, which main difference is the continues progression between relapses. PPMS is similar to SPMS however this starts without a period of RRMS prior. Taken from: https://my-ms.org/ms_types.htm

1.1.4. Prognosis of multiple sclerosis The reason the phenotype is important in prognosis is due to difference in severity and recovery after an acute relapse. Relapses involving the pyramidal and bowel/bladder function result in a bigger change in disability and are therefore regarded as more severe (Stewart et al., 2017). Furthermore, complete recovery of function is more likely after sensory, visual and brainstem episode, compared to pyramidal, bowel/bladder, cognitive 4

and cerebellar episodes (Kalincik et al., 2014). Repeated poor recovery from episodes leads to accumulation of disability. This has been thought to explain the differences prognostic outcome for males and females. Kalincik et al. (2014) showed that females experience more repeated relapses, however, they often involve the sensory and visual functions. Contrarily, men are thought to have fewer relapses, but they more often affect pyramidal, brainstem or cerebellar functions. This indicates that women, who are more likely to experience more frequent but less severe relapses with better recovery, will also show better long-term disease outcomes than men.

1.1.5. Diagnosis of multiple sclerosis The heterogeneity in presentation can complicate the diagnosis of MS in the clinic. Currently, the clinical diagnosis is based on the McDonald criteria (Polman et al., 2011) that assesses the presence of inflammatory-demyelinating injury in the central nervous system (CNS) with dissemination in space and time. Inflammatory-demyelinating injury is measured using a combination of the patients clinical history, a neurologic examination, CNS imaging and exclusion of diseases with similar symptoms (Polman et al., 2011). Dissemination in space and time are terms that indicate that chronicity and widespread damage to the CNS needs to have occurred i.e. inflammatory activity in more than one CNS location and at different times. Previously, the dissemination in time required either two separate relapses and/or two brain images at different times. However, in 2010 the revised McDonald criteria highlighted the importance of imaging in the diagnoses and allowed dissemination in time to be assessed using a brain image at one time point (for how this is done please read “lesions” on page 10). Finally, the 2017 revised McDonald criteria further improve the dissemination in time by allowing specific measures in the spinal fluid (Thompson et al., 2018).

1.1.6. Pathology of multiple sclerosis The pathology in MS is complex and consists of inflammation, demyelination and neurodegeneration. We will discuss all three pathological processes. Followed by an overview of how each pathological process plays a role in the different clinical stages of MS.

Inflammation MS is often regarded as an autoimmune disorder, which means that the body’s immune system recognizes its own antigens as foreign and attacks its own cells. This results in inflammation within the CNS which includes the brain, optic nerves and spinal cord (Abou-Raya & Abou-Raya, 2006). The initial trigger causing inflammation in the CNS remains unknown, but the dominating theory is the CNS extrinsic or molecular mimicry theory (Malpass, 2012), which proposes that malfunctional peripheral immune cells invade the CNS. This theory is opposed to the CNS intrinsic theory (Louveau et al., 2015; Ransohoff & Engelhardt, 2012) that proposes the neuronal damage is triggered by events inside of the CNS (Heneka, Kummer, & Latz, 2014; Hussain et al., 2014).

The peripheral immune cells, particularly the T-cells, are autoreactive and migrate through the blood-brain-barrier (BBB) to enter the CNS. In the CNS, they are involved in the production of proinflammatory cytokines (figure 2) (Baecher-Allan et al., 2018; Malpass, 2012; Steinman, 2008). Proinflammatory cytokines are signalling molecules and initiate a cascade of pathological processes inside the CNS. Furthermore, they activate the CNS innate immune cells (microglia and astrocytes). The microglia and astrocytes contribute to the pathological processes (Amedei, Prisco, & D’Elios, 2012). In addition to T-cells, B-cells play a role by presenting antigens to activate T-cells and CNS innate immune cells, and the production of autoantigens (which may cause neuronal damage (Fraussen, Claes, de Bock, & Somers, 2014)) and proinflammatory cytokines (Wekerle, 2017).

Interestingly, all currently available disease modifying therapies (DMT) affect the peripheral immune response and are effective in reducing both the CNS inflammation and clinical relapses (Grand’Maison et al., 2018). Unfortunately, none are successful in stopping to stop progression. This is suggested to be due to two other processes involved in MS pathology: demyelination and neurodegeneration. Furthermore, even though inflammation is involved in the different clinical stages of MS, it is not always the primary driver of pathology. The role of inflammation in the different stages of MS will be discussed in the at the end of this section.

Figure 2. Inflammatory processes in MS. The T-cells migrate through the BBB and once inside the CNS they interact with the myelin antigen and recognizes it as foreign. This causes activation of the autoreactive T-cell which produce cytokines and activate the CNS innate immune cells (macrophages and microglia). Macrophages and microglia further contribute to cytokine production and directly cause demyelination. B-cells are also involved in the pathology of MS by presenting antigens (using the complement system) and producing autoantigens and proinflammatory cytokines. In red are depicted on which place the different disease modulating therapy acts. Together, the inflammatory processes driven by peripheral and CNS innate immune cells lead to demyelination. Taken from: Steinman (2008).

Demyelination Demyelination is the loss of myelin, which is the result of an immune cell mediated attack against myelin antigens (Kane & Oware, 2012). Myelin is an insulating layer around nerves, also called axons, consisting of glial cells, the oligodendrocyte, in the CNS (Kandel, Schwartz, & Jessel, 2000; Natrajan, 2015) (see figure 3). Propagation of action potentials along the axons depends on the combined action of on sodium channels, potassium channels and the sodium/calcium exchanger. To further increase the signal propagation of action potentials, myelin is wrapped around the axon with short uncovered areas called the ‘Nodes of Ranvier’. Due to the low capacitance of myelin, the action potentials “hop” from

node to node, creating a saltatory conduction. Myelination around the axon combined with high expression of channels within the Nodes of Ranvier significantly increases the rate of conduction speed of action potentials. Therefore, demyelination of the axons impairs or blocks the conduction of action potentials. Subsequently, slower signal propagation causes deficiency in sensation, movement, cognition, and/or other functions depending on the location of the involved axons (Charil et al., 2003).

Figure 3. Schematic view of a neuron. The oligodendrocytes extend their membranes to create multiple myelin sheets around the axons. The areas between myelin sheets are called Node of Ranvier. The action potential along the axon jumps from node of Ranvier to node of Ranvier, creating the saltatory impulse (red dashed line). When the action potential reaches the terminals and the end of the axon, neurotransmitters will be released. Taken from: Natrajan et al. 2015

Restoration of the conduction speed can occur in two ways and clinically represents the recovery after relapses. First, when the immune cell mediate attack subsides, remyelination can occur by generation and differentiation of the precursor cells of oligodendrocytes (Patrikios et al., 2006). The oligodendrocyte precursor cells migrate to sites of active demyelination where it will start the formation of new myelin (Biname, Sakry, Dimou, Jolivel, & Trotter, 2013). Although remyelination can allow the axon to regain function, new myelin is often thinner and shorter than the original resulting in a slower conduction (Ludwin & Maitland, 1984). As a second mechanism to restore axon function is a change in channel density on the demyelinated axons. In response to chronic demyelination, sodium and calcium channels can be distributed throughout the axon instead of localized at the

nodes. However, this will only partially recover axonal function and comes with an increase in energy demand on mitochondria, the powerhouse of the cell (Craner et al., 2004; Zambonin et al., 2011). When the energy demand is in excess of the supply, it will lead to hypoxia, which over time will lead to irreversible degeneration of axons (Trapp & Stys, 2009).

Neurodegeneration Neurodegeneration is an umbrella term for the progressive loss of structure and function of neurons, and it is the strongest correlate of irreversible neurologic disability (Friese, Schattling, & Fugger, 2014). Neurodegeneration is already present in the early stages of MS (Barkhof, Calabresi, Miller, & Reingold, 2009; Bjartmar, Kidd, Mork, Rudick, & Trapp, 2000; Chard et al., 2002; Ferguson, Matyszak, Esiri, & Perry, 1997; D. H. Miller, Barkhof, Frank, Parker, & Thompson, 2002; Novakova et al., 2018) and the degree of neurodegeneration early in the disease may be an indicator of disease progression and outcome (Brex et al., 2002). There are immune dependant and immune independent mechanisms driving the neurodegeneration (figure 4).

The immune dependent component consists of microglia, astrocytes and B cells. The B cells are most important as they are linked to meningeal inflammation, cortical demyelination and an increased rate of progression (Choi et al., 2012; Magliozzi et al., 2010). However, the neurodegeneration is predominantly driven by immune independent mechanisms (Waxman, 2000). This also partly explains why most DMT treatment targeting inflammation fail to halt neurodegeneration (Correale, Gaitan, Ysrraelit, & Fiol, 2017). The immune-independent component consists of different pathways that ultimately leading to atrophy and degeneration. Briefly, we will discuss how mitochondrial dysfunction, oxidative stress, ion channel dysfunction lead to neurodegeneration.

First, mitochondrial dysfunction is induced by oxidative injured as a result of the increasing levels of reactive oxygen and nitrogen species (ROS and RNS) (Mahad et al., 2009). As the powerhouse of the cell, mitochondrial dysfunction leads to energy deficiency. Initially, loss of energy can lead to alteration in the function of the cell without resulting in structural damage; however, when the energy deficiency surpasses a certain threshold it will contribute to axonal degeneration (Correale et al., 2017). Furthermore, mitochondrial dysfunction contributes to the production of ROS and RNS creating a vicious cycle of

neurodegeneration. Second, demyelinating of the axon in active MS lesions produces iron that is by itself non-toxic. However, iron can react with ROS and create highly damaging ROS which further increases the oxidative stress in MS (Kell, 2009). Finally, an excessive production and reduced clearance of the excitatory neurotransmitters, glutamate, is found in MS lesions. Glutamate excitotoxicity causes a massive influx of calcium through ion channels which leads to an ionic imbalance. Together, all three immune independent mechanisms contribute to degeneration of neurons.

Figure 4. Mechanisms of neurodegeneration in MS. The immune dependent mechanisms include the microglia, astrocytes and the B-cells. The immune independent mechanisms include mitochondrial injury due to ROS and RSN production, oxidative stress due demyelination related to extracellular iron accumulation, and ionic imbalance due to glutamate excitotoxicity. Taken from: Baecher-Allan, KasKow & Weiner (2018).

Lesions Lesions (also called plaques) are the regions of injury in the brain, which are caused by a combination of inflammation, demyelination and neurodegeneration (Popescu & Lucchinetti, 2012). Lesions occurs in both white matter (WM; includes the axons and myelin) and grey matter (GM; includes cell bodies of neurons, axon terminals, little myelin and dendrites (receiving part of cells)) of the CNS. The temporal evolution of the lesions

has been classified in different ways (figure 5). The most recent classification divides MS lesions as active, mixed active/inactive, or inactive lesions based on the presence and distribution of immune cells (Bö et al., 1994; Jonkman et al., 2015; Kuhlmann et al., 2017; Peterson, Bo, Mork, Chang, & Trapp, 2001; Trapp et al., 1998). There are thought to be pre-active lesions, characterized by activated microglia, minor loss of myelin and the absence of other immune cells.

Figure 5. Schematic view of the temporal evolution of MS lesions. Active and demyelinating lesions are characterised by a high level of inflammation with different immune cells present and demyelination. Mixed active/inactive lesions are characterised by lower amount of inflammation, and the immune cells collects around the border of the lesions. Inactive lesions are characterised by sharp demarcated, hypocellular and almost complete demyelination. Several radiological reports describe the re- enhancing of established MS lesions; therefore, inactive lesions or mixed active/inactive lesions might be infiltrated by a second wave of macrophages/microglia resulting again in an active lesion. Taken from: Kuhlmann et al. 2017.

Active and demyelinating lesions are the initial phenotype of MS lesions and are common in patients with short disease durations. They are characterized by hypercellularity including T cells, microglia and macrophages, and the loss of myelin. There is a relative preservation of axonal function and the immune cells are kept within the boundaries of the glial limitans. They can develop into active and post-demyelinating lesions or mixed/inactive lesions. Mixed active/inactive lesions can be demyelinating and post-demyelinating. They are characterised by hypocellularity and the immune cells present collect around the border of the lesions. These lesions are common in patients with a long (10+ years) disease duration (Frischer et al., 2015). Finally, inactive lesions are characterised by sharp demarcated and hypocellular. There is substantial loss of axons and myelin and indicates neurodegeneration. These lesions are predominant in patients with a very long (15+ years) disease duration and/or SPMS (Frischer et al., 2015). The presence of remyelination may be present in all types of lesions; however, the quality of repair is thought to be much lower in the progressive stages. Finally, studies have shown the re-activation of inactive MS lesions by infiltration of immune cells (Campbell et al., 2012).

Clinical stages pathology Inflammation, demyelination and neurodegeneration are involved in all different clinical stages of MS; however, whether they are the important driver or play a smaller role in the background varies. In the relapse-remitting stage of the disease, the pathology is driven by inflammation and demyelination, with slow build of neurodegeneration. At the same time remyelination and redistribution of ion channel occurs which results in the remission. In RRMS, the remission can recover function back to normal. When the amount of damage to the brain exceeds the brain’s capacity to remyelinate and compensate for the damage, the result is irreversible progressive neurological disability (Nave & Trapp, 2008; Trapp et al., 1998). This can present as partial recovery of function, seen in RRMS. However, this can also result in no recovery, indicating the stage when MS patients convert from RRMS to SPMS. Furthermore, in a small number of patients, PPMS starts without relapses yet this is thought to be driven by the same pathological processes as SPMS (Baecher-Allan et al., 2018). One of the key characteristics in the progressive stages is neurodegeneration. Furthermore, contrarily to the strongly immune system driven RRMS, the progressive MS is thought to have immune dependant and immune independent mechanisms driving the disease progression.

1.1.7. Conclusion MS is a multifocal demyelinating disease that affects many young people. The theorized mechanisms of injury in the CNS highlight the complexity of the disease. And depending on the location of the pathology in the brain, MS patients present with varying symptoms. In the next chapter we will focus on a specific symptom: tremor.

1.2. Tremor in multiple sclerosis One of the most common symptoms in MS is tremor and will be our focus for the next paragraphs.

1.2.1. Tremor presentation in multiple sclerosis The clinical definition of tremor is “a rhythmic involuntary movement that involves oscillations of one or more parts of the body, resulting from repeated contraction and relaxation of opposing muscle groups” (Deuschl, Wenzelburger, & Raethjen, 2000). In MS, tremor is by far the most common movement disorder with a recent study in North America finding a prevalence ranging from 45% to 46.8% (Rinker et al., 2015). MS-related tremor often emerges in the relapse-remitting or secondary progressive stage of the disease, on average 11 years after MS onset (Alusi, Worthington, Glickman, & Bain, 2001). Patients with a mild tremor often have an earlier onset of MS and tremor, while a more severe tremor was found in patients with later onset of MS and tremor (Rinker et al., 2015). Regardless of the severity, the presence of tremor itself significantly adds to disability (Pittock, McClelland, Mayr, Rodriguez, & Matsumoto, 2004). In 3%-15% of MS patients, the tremor causes severe disability; furthermore, even mild MS tremor is associated with a worsening of the disability (measured using the expanded disability status scale [EDSS] (Kurtzke, 1983)) and increased likelihood for patients to be wheelchair dependent (Alusi, Worthington, et al., 2001). Pittock et al. (2004) reported an increase in EDSS by a median of 2 points in MS patients with tremor compared to those without.

To complicate matters, tremor can present differently in different people and presentations complicated further by accompanying symptoms. Tremor is classified according to the consensus statement of the Movement Disorder Society (Deuschl et al., 2000) into different types. There are two main types of tremor: rest tremor and action tremor. The rest tremor presents itself in a body part that is not voluntarily activated and is completely

supported against gravity, while the action tremor is produced by voluntary muscle contraction. The action tremor contains further sub-classifications; common types are postural tremor, intention tremor, kinematic tremor, task-specific tremor and isometric tremor. In MS, the action tremor is most common and specifically, the intention and postural tremors, while a rest tremor is rarely seen (Alusi, Worthington, et al., 2001). Furthermore, tremor usually affects the upper limbs unilaterally or bilaterally; however, the tremor can also affect the lower limbs, head, truck and vocal cords (Alusi, Worthington, et al., 2001; Rinker et al., 2015).

1.2.2. Tremor Assessment in multiple sclerosis MS tremor often presents as a complex movement disorder influenced by ataxia, dysarthria (difficulty articulating), dysmetria (overshoot or undershoot of intended movement), and dysdiadochokinesia (impaired ability to complete a sequence rapid of movements). The assessment of tremor severity needs to take these and other influences (e.g. weakness and disease related disability as measured by the EDSS) into account. The EDSS (Kurtzke, 1983) and is the best known and most widely used scale to monitor disease progression in MS and as outcome measures in clinical trials (Balcer, 2001). The EDSS is an ordinal rating scale ranging from 0 (normal neurological status) to 10 (death due to MS) with steps of 0.5. The lower scale values of the EDSS measure impairments based on the neurological examination, while the upper range of the scale (> EDSS 6) measures handicaps of patients with MS. The determination of EDSS 4 – 6 is heavily dependent on aspects of walking ability. Within the EDSS you have eight functional system scores. Tremor is rated as part of the cerebellar functional system score; however, this score is largely driven by gait ataxia.

Overall, there are a variety of subjective and objective techniques that are currently used in the assessment of the different subtypes of tremor, tremor severity and its impact on the lives of patients. These techniques are used in clinical and research settings to monitor the progression of tremor and to examine the efficacy of treatments towards tremor. In the clinical setting, the techniques that can be used are constrained by cost, time and space required for the assessment technique. Therefore, clinical assessment usually includes subjective rating scales for a semi-quantitative assessment of frequency and magnitude of tremor. These clinical constraints are not so important in the research setting; therefore, more extensive assessment techniques, collectively called objective performance tests, are

used. The subjective rating scales and the objective performance test will be discussed subsequently.

Subjective rating scales Tremor scales have historically been developed to rate hyperkinetic movement disorders such as Parkinson’s disease (PD) and essential tremor (ET). Fahn and colleagues were the first to devise a comprehensive tremor scale for non-Parkinsonian tremor in 1984 (Fahn, Tolosa, & Marin, 1988). This rating scale assesses tremor in nine body parts on a 0-4 scale with increasing score represents increasing severity. The scale consists of three different settings: at rest, fixed posture and action and intention. It also includes an assessment of arm tremor while writing and pouring water as well as a patient-reported functional disability resulting from the tremor (speaking, eating, drinking, social activities, etc.). Evaluation of the intra-rater reliability showed good to excellent results; however, the interrater reliability is often poor to moderate (Hooper, Taylor, Pentland, & Whittle, 1998; Stacy, Elble, Ondo, Wu, & Hulihan, 2007). This is related to the fact that this scale requires standardized training of the raters. Furthermore, even though sensitivity to change has been found in ET treatment studies, there is a ceiling effect for upper extremity tremor (R. Elble et al., 2013; R. J. Elble, Lyons, & Pahwa, 2007). In the literature, this scale is rarely used in its original form and only some studies use a modified and heavily shortened version of this scale. Most studies assess tremor by clinical examination (e.g. finger-to-nose testing, drinking from a cup, nine-hole-peg-test (9-HPT) or writing and drawing tasks) or use a simple ordinal severity scale, often classifying tremor as absent, mild, moderate or severe.

In 1993, Bain et al. (Bain et al., 1993) introduced a battery of tremor rating scales for assessing rest, postural and kinetic/intention tremor severity in the head, voice and all four limbs. Originally, it was devised as a simple 0-10 tremor. However, as most people struggled to differentiate beyond 7 levels (G. A. Miller, 1956), they redefined the scale to 0=no tremor, 1-3=mild tremor, 4-6= moderate tremor, 7-9=severe tremor, and 10=extremely severe tremor. This results in a 5-level scale with intermediate gradations (see appendices 1 and 2). The Bain tremor severity scale has been validated in MS tremor and it is the only reliable tremor severity scale available for MS tremor. The Bain scale rates tremor from 0 to 10 with increasing scores representing increasing severity. Two studies validation that most patients with MS are shown to have a moderate tremor as represented

by a median Bain composite tremor score between 4-5 (Van der Walt et al., 2015; Van Der Walt et al., 2012). The inter-rater reliability is shown to be moderate to good. Furthermore, Van der Walt et al. (Van Der Walt et al., 2012) has shown high to near-perfect intra-rater reliability in tremor assessment in MS patients. Finally, the Bain spiral score showed to be sensitive to change in MS (Alusi, Aziz, et al., 2001).

Objective performance tests The limitations of subjective rating necessitate more objective and quantitative measurements. These tests need to be capable to detect subtle disease-related changes throughout the entire disease course. Two tools that have been applied in movement disorder research for many years are accelerometry and electromyography (EMG) (Marshall & Walsh, 1956). The accelerometer calculates the acceleration of the body by measuring the force and mass, by use of Newton’s Second Law (F [force] = a [acceleration] x m [mass]). The EMG is a measure of muscle activity by recording the electric potential of muscles. Muscle contraction is the result of an electric activation of that muscle and as tremor is the result of repeated contraction and relaxation of opposing muscle groups, EMG is the perfect way get a direct measure of tremor. By extracting the frequency of movement or frequency of the electric charges using the accelerometer and EMG respectively, you get a measure of tremor frequency.

Alusi et al. (Alusi, Worthington, et al., 2001) used the accelerometer to calculate the frequency and amplitude of the tremor in MS and showed that measurements of frequency together with severity and associated dysmetria enable differentiation between the different subtypes of tremor. A fine distal postural tremor had a frequency range of 2.6-6Hz, a coarse distal postural tremor had a frequency range of 3-7Hz, a proximal postural tremor had a frequency range of 3.6-4.8Hz, a distal and proximal postural tremor had a frequency range of 3.3-5Hz and an intention tremor had a frequency range of 2.6-3.7Hz. Recently, Ayache et al. (S. S. Ayache et al., 2015) assessed the accelerometer-EMG coupling and showed that in 18 MS patients with a clinically defined tremor only one has a neurophysiological tremor (defined by a frequency peak with coherence between accelerometer and EMG). They concluded that a clinical examination might be insufficient to differentiate tremor from a pseudo-rhythmic activity. Hence, neurophysiological investigation of tremor in MS including both accelerometry and EMG measurements is particularly relevant to confirm the presence and severity of tremor.

Newer objective tremor measurement tools are available and many of which utilize the inertial sensor (Rovini, Maremmani, & Cavallo, 2017). The inertial sensor can consist of a combination of sensors, but the most promising in 3D gait analysis is the combined three- axis accelerometer, three-axis gyroscope, and three-axis magnetometer. These accelerometers are different from the previously described conventional accelerometer in that they have three-axis orientation. Gyroscopes are used to measure rotations around an axis and changes in the orientation. The magnetometer measures changes in the orientation of the body in relation to the magnetic north to provide a global reference measure for body orientation that can be compared across different people. As all three types of sensors are three-axis this enables the capture of 3D information on movement. Inertial sensors are widely available, portable, and inexpensive; however, they are not perfect. One of the problems is the difficulty in analysing the data and making the output comprehensible for clinicians without losing the quality and quantity of the data.

Only one study has assessed the use of a inertial sensor in upper limb function in MS. Carpinella et al. (Carpinella, Cattaneo, & Ferrarin, 2014) did a cluster analysis to create three MS subgroups (mild, moderate and severe impairment) based on the Action Research Arm Test, the 9-HPT and the Fahn Tremor Rating Scale. First, they validated the inertial sensor to abovementioned clinical tests. Second, they showed that patients who scored maximally on the Action Research Arm Test and are categorized into the mild impairment group still show significant decreased performance (task duration and task jerk index) when incorporating the inertial sensor in the assessment of the Action Research Arm Test during the grasp, grip pinch and gross tasks. In other words, the inertial sensor is able to differentiate mild upper limb dysfunction between healthy controls and MS, while a standardized ordinal scale was insufficient to do so. However, as a limitation they mentioned the complexity of the data analyses and the lack of sensors on more than one location of the arm. Regardless, these findings highlight the advantageous of inertial sensor with good accuracy and capability of quantifying tremor.

The third aim of the TREMTOX study was the optimize tremor assessment using inertial sensors and help characterisation of tremor in MS. Furthermore, we aimed to validate the use of a benchtop solution for the assessment of tremor. TREMBAL is an instrument that quantifies tremor using electromagnetic motion tracking sensors and has an accuracy of

0.45mm (Perera et al., 2016; Perera, Yohanandan, McKay, & McDermott, 2015). In our current study, the TREMBAL sensors were positioned at standardized locations on the 2nd phalanx of the index finger, wrist dorsum (3 cm distal to the wrist joint), as well as forearm (10 cm proximal to wrist joint) and upper-arm (10 cm proximal to the lateral epicondyle). It was found that measurements at a single location was sufficient to provide adequate data to quantify tremor. In addition, translational velocity of the index finger showed the strongest correlation (r = 0.81) with clinical tremor rating. Finally, the sensitivity to change was substantially better compared to the clinical observations. To conclude, these results are promising as they could be used in the clinical setting and with automated analyses could provide easy to interpret, sensitive data that is essential in monitoring tremor and assessing treatment efficacy.

1.2.3. Treatment of tremor Currently, pharmaceutical treatment options for tremor in MS can be divided into “insufficient evidence…”, “probably does not reduce tremor…” and “it should not be considered…”, while surgical treatment is very invasive and should be used a last resort option (Labiano-Fontcuberta & Benito-León, 2012). There are very limited numbers of clinical studies examining potential drugs for treating MS tremor. Therapies should be used to reduce functional disability or embarrassment and to improve health-related quality of life. Furthermore, it is important that a treatment option is able to provide benefits that out way any side effects. Therefore, patients generally begin with non-pharmaceutical treatment option to help with mild tremor. Subsequently, pharmaceutical treatment options, and as a last resort surgical treatment options, need to be considered for each patient. Unfortunately, treating MS-related tremor is challenging and unsuccessful in most cases.

Non-pharmaceutical treatments Non-pharmaceutical treatment options are often impractical and provide short-lived relief. Recent work discusses the option of limb cooling as a treatment for tremor. Cooling is done locally on the muscles involved in the tremor. Feys et al. (Feys, Helsen, et al., 2005) showed that cooling to both 25 and 18 degrees Celsius reduces the tremor amplitude and frequency. The effect of treatment gradually disappears within 30min, and the effect is inversely proportional to the temperature of cooling (Feys, Helsen, et al., 2005). However, due to the short effect duration, this treatment is a task directed solution. Mechanical loading has shown beneficial effects in three MS patients (Dahlin-Webb, 1986); however,

effect on tremor is minimal and unlikely will reflect in a functional improvement. Other non-pharmaceutical treatment options include physiotherapy ("MS, Parkinson's disease and physiotherapy," 2002), orthotics (Aisen, Arnold, Baiges, Maxwell, & Rosen, 1993) and lifestyle changes; however, they have limited scientific evidence supporting reduction in tremor severity (Koch, Mostert, Heersema, & De Keyser, 2007). Finally, a case study examined the effect of transcranial magnetic stimulation of the thalamus in a patient with SPMS with severe upper limb tremor (Samar S. Ayache, Ahdab, Neves, Nguyen, & Lefaucheur, 2009). They found a reduction in the amount of tremor, but due to the nature of the study were unable to ascertain the relation between the treatment and effect.

Pharmaceutical treatments Effective pharmaceutical treatment options are currently non-existent, and of those tested, the mechanisms of action are often unknown. Drugs tested were often previously trailed in ET with sufficient evidence, including propranolol (Koller, 1984), primidone (Naderi, Javadi, Motamedi, & Sahraian, 2012), topiramate (Schroeder, Linker, Lukas, Kraus, & Gold, 2010), isoniazid (Bozek, Kastrukoff, Wright, Perry, & Larsen, 1987; Hallett, Lindsey, Adelstein, & Riley, 1985; Koller, 1984), and levetiracetam (Feys, D'Hooghe M, Nagels, & Helsen, 2009). However, the small trials competed using these drugs did not support their use for treatment of MS-related tremor. Furthermore, two drugs commonly used to treat nausea and vomiting caused by cancer treatment (ondansetron and dolasetron) have been examined as they are suggested to also reduce symptoms of cerebellar ataxia (Trouillas, Brudon, & Adeleine, 1988). Respectively, these drugs have either inconsistent results (Gbadamosi, Buhmann, Moench, & Heesen, 2001; Rice, Lesaux, Vandervoort, Macewan, & Ebers, 1997) or negative results (Monaca-Charley et al., 2003). The anticonvulsant drug, carbamazepine (Sechi et al., 1989), has been tested as it has shown reduces tremor assessed with a clinical rating scale and accelerometer. However, they failed to assess whether this also improves function. Similar shortcomings are encountered when examining, gluthetimide (Aisen, Holzer, Rosen, Dietz, & McDowell, 1991) (a hypnotic sedative) and intrathecal baclofen (Sadiq & Poopatana, 2007) (a beta-blocker). Monoclonal antibody rituximab has been tested in one case study and they found a significant reduction in head tremor (Chansakul, Moguel-Cobos, & Bomprezzi, 2011). However, further research is needed as this depletes the B-cells from the immune system and should only be used as a last resort. Together, there is a clear lack of evidence that any of the above-mentioned drugs alleviate tremor in MS patients. Interestingly, recent interest has been drawn to the

use of botulinum for the treatment of tremor in MS which will be discussed in the next paragraph.

Botulinum toxin (BoNT) is an enzyme produced by a spore-forming bacillus, clostridium botulinum (figure 6). It is the most acutely lethal toxin known, with a median lethal dose of 1.3-2.1 ng/kg intramuscularly. This means that with only one gram of pure BoNT you can kill millions of people. After intramuscular injection, the receptor-binding domain (heavy chain) of BoNT binds to the presynaptic receptors of the neuromuscular junction. The toxin then enters the neuronal cell via receptor-mediated endocytosis. The light chain (toxic moiety) enters the neuroplasm, where it cleaves one or more of the proteins that form soluble N-ethylmaleimide–sensitive factor attachment protein receptor (SNARE) protein complex (complex formed by synaptosomal nerve-associated protein 25 (SNAP-25), Syntaxin and the vesicle associated membrane protein). There are 7 different types (A to G) that work slightly different as they cleave to different proteins of the SNARE complex. SNARE protein complex normally allow neurotransmitter, acetylcholine to leave the cell and transmits a nerve impulse to a muscle, signalling the muscle to contract. As BoNT prevents the formation of the SNARE protein complex by cleaving the proteins that form SNARE protein complex, the acetylcholine release is blocked. As the result, signal transmission between the nerve and muscle is stopped causing botulism (paralysis). The most common type used for treatment is BoNT-A and this affects SNAP-25. Transmission of acetylcholine is restored over 2 to 4 months in the neuromuscular junction; however, clinical effects may persist for 6 months (Birklein, Eisenbarth, Erbguth, & Winterholler, 2003).

The use of BoNT-A has been previously used successfully in patients with ET. Jankovic et al. (Jankovic, Schwartz, Clemence, Aswad, & Mordaunt, 1996) and Brin et al. (Brin et al., 2001) showed positive effect of BoNT-A in ET measured by clinical rating scales; however, no significant improvements on functional scales. Furthermore, common side effects of BoNT-A including local muscle weakness has been reported by both studies. Recently, Van der Walt et al. (Van Der Walt et al., 2012) performed a double-blind, randomized, controlled crossover study of targeted BoNT-A injections in 33 tremor-affected limbs of MS patients (Van Der Walt et al., 2012). They reported significant improvement (p<0.001) after BoNT-A compared with placebo treatment in the Bain score for tremor severity as well as writing and drawing of an Archimedes Spiral at 6 weeks and 12 weeks. With respect

to the side effects, more patients developed weakness after BoNT-A treatment (42.2%) than after placebo injection (6.1%; (p<0.0005). However, weakness was generally mild (just detectable) to moderate (still able to use limb) and on average, resolved within 2 weeks. This study provided class III evidence of the efficacy of BoNT-A for the treatment of a disabling MS symptom but translation into clinical practice requires class I evidence and further study. Small numbers made any analyses of tremor-type subgroups and the correlation between baseline clinical factors and treatment response impossible. Furthermore, the study did not include any para-clinical measures of MS tremor severity and response to BoNT-A.

Figure 6. Mechanism of neuromuscular blockade by botulinum toxin. In normal muscle tissue (left), there are normal SNARE proteins that aid to the release of acetylcholine into the synaptic cleft at the neuromuscular junction. At right, the SNARE proteins are cleaved by botulinum toxin, resulting in inhibition of synaptic vesicle fusion and absence of acetylcholine release. Depending on the type of botulinum toxin (A-G), different SNARE proteins are targeted. Taken from Dickerson & Janda (2006).

Surgical treatments Tremor reduction can be obtained with stereotactic thalamotomy, deep brain stimulation (DBS) and gamma knife thalamotomy (Alusi, Worthington, et al., 2001; Mathieu, Kondziolka, Niranjan, Flickinger, & Lunsford, 2007; Raju, Niranjan, Monaco, Flickinger, & Lunsford, 2018). Most of the studies into the effects of surgical treatment in MS tremor are small observational retrospective studies and the majority of studies are remarkably imprecise in providing basic information on the length of follow-up, on adverse effects and most importantly on the effect on functional outcome and tremor associated disability. Because of many shortcomings of the published studies, caution is warranted when interpreting these results. Seventeen studies have examined the effect of thalamic DBS and sixteen showed a reduction of tremor in 60-100% of MS patients. Similar results are found for stereotactic thalamotomy with twenty-six studies of which twenty-five report a reduction of tremor in 50-100% of MS patients (Koch et al., 2007). However, surgical treatment is not easily accessible and both come with serious adverse effects (Bittar et al., 2005; Schuurman et al., 2000). In addition, the duration of the response is unknown given the lack of long-term post-operative outcome studies (Hassan, Ahlskog, Rodriguez, & Matsumoto, 2012). A relative newer option with fewer studies is gamma knife thalamotomy (Mathieu et al., 2007; Raju et al., 2018). The two studies available have shown 86-100% of the patient experiencing improvement in the tremor. Furthermore, this option is labelled as non-invasive; however, between 13-17% experiences temporarily adverse effects and 1 patient experienced a large thalamic cyst that required intervention to be managed.

1.2.4. Conclusion Tremor is a common and disabling symptom of MS. In the outpatient clinic, the assessment of tremor in MS is limited by observation and subjective rating scales. However, new technological advances such as our TREMBAL device offer accurate, sensitive and quantitative tools. Future clinical applicable tools will be beneficial to monitoring the development of tremor during disease progression, thereby improving tremor characterization in MS. Furthermore, improving the assessment of tremor will aid the treatment development by having a more sensitive measure of change. This is essential as there is an urgent need for more extensive and longitudinal trials into the effects of abovementioned drug and surgical treatment. Specifically, BoNT-A treatment is a promising drug that requires further study to translate use of BoNT-A for the treatment of tremor in MS into clinical practice.

Another aspect of MS tremor that is vital for the development of new treatments, is the understanding of what causes tremor in MS. The next sections discuss what is currently known about the pathophysiology of MS tremor

1.3. Pathology of tremor in multiple sclerosis 1.3.1. Pathophysiological insights non-imaging studies The pathophysiology of MS tremor remains largely unknown. Traditional views of MS tremor emphasize a central and important role for the cerebellum. This is related to the correlation between tremor severity and dysmetria, dysarthria and dysdiadochokinesia (Diener & Dichgans, 1992). These three conditions are characterized by lack of coordination and fine motor movement, characteristics often linked to cerebellar function. Alusi et al. (Alusi, Worthington, et al., 2001) found no correlation between tremor severity and cognition, Barthel disability index, walking speed and grip strength. This supports cerebellar involvement and, furthermore, indicates it is localised rather than diffuse damage to the cerebellum.

The effect of most non-pharmaceuticals treatment options is through modulation of peripheral afferents in the presence of increased long latency stretch reflexes (Deuschl, Raethjen, Lindemann, & Krack, 2001; Sohn, Niu, & Sanger, 2015). The long latency stretch reflex starts in the muscle spindles, which provide peripheral input that relayed in the thalamus prior to integration by the primary motor cortex (Klein et al., 2012; Middleton & Strick, 2000). Subsequently, the motor cortex provides modulation to the motor function through cortico-cerebellar feed-forward loops (Middleton & Strick, 2000). Decreasing the sensitivity and excitability of the muscle spindles, by means of mechanical loads (Richard et al., 1997), peripheral cooling (Feys, Helsen, et al., 2005) and BoNT-A treatment (Jankovic et al., 1996; Van Der Walt et al., 2012), enables modulation of the cortico-cerebellar feed- forward loops and subsequently reduce tremor (Fischer & Schafer, 1999; Giladi, 1997). In addition to non-pharmaceutical treatment studies, surgical treatment studies have implicated the thalamus for the production of tremor as this is their main target (Alusi, Worthington, et al., 2001; Yap, Kouyialis, & Varma, 2007). Finally, animal studies have further highlighted the cerebellum (Conrad & Brooks, 1975; Gemba, Sasaki, Yoneda, Hashimoto, & Mizuno, 1980; Gilman, Carr, & Hollenberg, 1976) and the superior cerebellar peduncle (SCP) (Burrows, 1937) as damaging either structure resulted in tremor.

Taken together, MS tremor is thought to be associated with cerebellar dysfunction but also and maybe more importantly its afferent tracts.

This hypothesis is congruent with current findings about tremor pathophysiology in ET and PD (Muthuraman et al., 2018; Sharifi, Nederveen, Booij, & van Rootselaar, 2014). In both ET and PD there is only partial understanding of what causes tremor, but there is a more substantial amount of research to base a hypothesis on. The combination of structural and functional in vivo imaging has allowed them to also point towards the cerebello-thalamo-cortical as the original of the tremor network (Fang et al., 2013; Lin et al., 2013; Muthuraman et al., 2018; Novellino et al., 2013), a tract which primary function is thought to be motor adaptation (Aumann, 2002; Groiss & Ugawa, 2013). This tract originates in the ipsilateral cerebellum (compared to the tremulous hand) and exits through the ipsilateral SCP to the contralateral thalamus and ends up in the contralateral motor cortex (figure 7). A recent study by Muthuraman et al. (Muthuraman et al., 2018) showed that the cerebellar nuclei in which the tract originate slightly differs for ET, PD and healthy controls; however, the trajectories are otherwise identical. Furthermore, the direction of flow in ET and PD is from the cerebellum to the motor cortex, opposite to that in healthy controls. This could explain the involuntary nature due to the lack of the cortico-cerebellar feed-forward loop. Taken together, research done in ET and PD highlights the cerebello- thalamo-cortical tract as a driver of tremor.

Figure 7. Anatomical overview of the cerebello- thalamo-cortical tract. The tract originates in the cerebellum and flows via the superior cerebellar peduncle to the thalamus. From the thalamus it flows to the cortex. Taken from: http://humanphysiology.aca demy

1.3.2. Structural imaging 1. Magnetic Resonance Imaging (MRI) is the most common tool that enables in-vivo visualization of pathophysiological processes operative in MS patients, and for this reason has gained widespread popularity. Structural MRI is a great way to understand the infrastructure of the brain. Next, we will briefly introduce the mechanisms of structural MRI and how we derive the parameters that are useful for MS research: lesions and atrophy. Finally, we will discuss what the findings are regarding the pathophysiology of MS tremor. 2. The basics of structural imaging The MRI is a large magnet that consists of different electromagnetic components: main magnet, gradient coils, radio frequency (RF) transmitter and a receiver coil. The primary origin of the magnetic resonance (MR) signal is hydrogen nuclei, because water molecules are present in all parts of the body. Hydrogen nuclei have magnetic moments (spins) that behave like compass needles and without magnetization these compass needles are randomly orientated in to body. The main magnet aligns the nuclei parallel to the magnetic

field (B0). Three gradient coils (x, y, z) transform the static magnetic field into a gradient

magnetic field in direction of B0. The RF transmitter sends out RF pulses which create a B1

field perpendicular to the B0 field. Due to the gradient coils, the strength B0 field varies depending on the location and the resonance frequency of each nuclei varies as well. The nuclei with a resonance frequency similar to the RF pulse, change their orientation to align

with the B1 field. When the RF pulse is turned off, those nuclei return to B0 field orientation (relaxation of the nuclei) and release energy that they previously gained by the RF pulse. Subsequently, the released energy is recorded by the receiver coil. Changing the transmitted RF pulse, changes the location of the nuclei that will be affected. Thereby, with a range of RF pulses one gets a spatially localised image.

Spatial localization generates a 3D image but doesn’t explain the visualisation of different tissues on a MR image. This is achieved by differences in relaxation times. There are two

distinct relaxation processes: 1) regrowth of the B0 field and is known as the longitudinal

T1 relaxation and 2) decay of the B1 field and is known as the transverse T2 relaxation. Both T1 and T2 relaxation is dependent on tissue type and, more specifically, size of the molecules and mobility within the tissue structure. This important for the image visualization as it allows us to identify different structures in the body and, furthermore,

allows to extract lesions and atrophy measures in the brain that are two very important measurements in MS.

Detection of lesion in multiple sclerosis The MRI is a useful tool for the diagnosis of MS, monitoring disease progression and as an outcome measure in therapeutic trials. One of the reasons is its ability to identify lesions in MS patients (see figure 8A). Depending on the stage of the lesion, the lesion area will contain more and more liquid (called cerebrospinal fluid in the brain [CSF]) and, subsequently will appear more and more as a hypo-intensity on a T1-weighted image and hyper-intensity on a T2-weighted image. To optimize lesion detection, fluid-attenuated inversion recovery (FLAIR) image was introduced. This image provides a stronger contrast between the lesions and WM, by supressing pure CSF (figure 8A. left T2, middle FLAIR and right T1). Furthermore, to determine dissemination in space a T1-weighted image is used in combination with gadolinium. Gadolinium is injected intravenously and under normal circumstances does not pass the BBB; however, during active inflammation the BBB is damaged, and the gadolinium collects in the active lesion. Gadolinium shortens the T1 relaxation time; therefore, active lesion can be separated from older lesions that have repaired any BBB damage (figure 5 top right MRI image). Furthermore, GM lesions are very hard to detect compared to WM lesions. Specifically, 5% or less of cortical lesions are seen on T2-weighted and FLAIR scan, and for deep GM lesions, about 15% and 40% are seen for T2-weighted scan and FLAIR respectively (Geurts et al., 2005). A scan specifically equipped to detect GM lesions is a double inversion recovery (DIR), and it detects about nearly double the amount. The DIR suppresses both CSF and WM and enhances inflammatory lesions.

A B

Figure 8. Different MRI sequences to visualise MS lesions. A. Left: T2 (left), middle: FLAIR and right: T1. Taken from: https://www.reviewofoptometry.com/. B. DIR image of multiple sclerosis lesions.

Progression of MS has often been measured using increase in total lesion load. This holds true for the progression of CIS to RRMS, as this is shown to be predicted by the number of lesions on a T2 in the brain or spinal cord (Filippi, Rocca, & Comi, 2003). However, in the later stages of MS the number and volume of WM lesions have been found to correlate poorly with relapses and progression of disability (Brex et al., 2002; Filippi et al., 1995; Kappos et al., 1999). This is called the clinico-radiological paradox, and potential reason behind this are: the EDSS is often used to measure disease progression but it is nonlinear and heavily weighted toward ambulatory deficits; reorganization in the brain could contribute to functional recovery irrespective of lesions (Reddy et al., 2000); most studies are unable to accurately detect spinal cord and GM lesions; most studies fail to take lesion location into account; and finally dysfunction of normal appearing WM, or axonal damage, may contribute to disability (Filippi et al., 1995; Kappos et al., 1999). Many studies have shown that GM lesions are better predictors of disease progression (Calabrese et al., 2007; Calabrese et al., 2010; Klaver, De Vries, Schenk, & Geurts, 2013). Furthermore, Charil et al. (Charil et al., 2003) showed that lesion location is an important factor in determining the nature of the deficits in MS. Together it seems like lesion load is not a great predictor of disease progression, but might be better for specific symptoms with in MS.

Detection of volumetric changes in multiple sclerosis In addition to lesions, cerebral atrophy in both WM and GM is also a measurement derived from T1-weighted MRI images. For the visualisation of atrophy, automated programs separate the different structures based on intensity on the MR image and label them

according to a probabilistic map based on manually labelled brains (Fischl et al., 2002). The interpretation of atrophy in MS is tricky. Early processes in MS such as inflammation and oedema can cause inflation/swelling of the tissue which can mask atrophy (D. H. Miller et al., 2002). When the inflammation and oedema subside (either by itself or by use of DMT) the swelling reduces, and atrophy is more prominent. This is particularly relevant in clinical trials of DMT, that observe ‘pseudoatrophy’ in the first year after treatment that is actually the results of the resolution of oedema. Nevertheless, both GM and WM atrophy correlates with disability throughout the disease course (De Stefano, Battaglini, & Smith, 2007; Fisher et al., 2000); however, GM atrophy is thought to be a more sensitive and specific marker for neurodegeneration as it is a closer indicator of neuronal loss rather than axon and myelin loss combined (Jasperse et al., 2007). Finally, similarly to lesion load, location of atrophy is important for specific progression in different symptoms (Schoonheim et al., 2015).

Structural imaging in multiple sclerosis tremor A total of three structural MRI studies have been perform in MS tremor. The first case report reported that the MS patient with tremor had a lesion within the SCP, and the other tremor patients without MS had a lesion between the thalamus and red nucleus in the other case (Nakamura et al., 1993). A following case study of an MS patient with tremor reported a lesion within the middle superior peduncle (Fahn, 1986). Finally, a more recent study by Feys et al. (Feys, Maes, et al., 2005) demonstrated a link between tremor severity in MS patients and T2 lesion load in the contralateral pons. Furthermore, they found no relationship between the tremor severity and lesion load in the cerebellum. Taken together, these findings are minimal but still support the involvement of the cerebello-thalamo- cortical tract in MS tremor pathophysiology.

1.3.3. Functional imaging The brain is a complex machine and understanding the clinical presentation of a disease requires more than visualization of the structural infrastructure in the brain. Therefore, in concert with visualizing structure, one should assess the how the brain functions. Next, we will discuss the basics and interpretation of functional imaging and findings of fMRI in MS tremor.

The basics of functional imaging The technique to assess the function of the brain is functional MRI (fMRI). The fMRI uses a ‘Blood Oxygenation Level Dependent’ (BOLD) contrast that enables visualisation of ‘active’ areas in the brain. The principle behind the BOLD response is the haemodynamic response to neural activity. Neuronal activity results in a short decrease in oxygenated blood and increase in deoxygenated blood as surrounding cells will use up the available oxygen. This is followed by a delayed increase in local blood flow that supplies oxygenated blood. Oxygenated and deoxygenated haemoglobin in the blood have a different magnetic susceptibility. Specifically, oxygenated blood is diamagnetic and is repelled by the applied

B0 magnetic field. This creates a magnetic field in the opposite direction of the B0 field.

Deoxygenated blood is paramagnetic and will be attracted by the B0 field. This creates in magnetic field in the same directions as the B0 field. As mentioned in above, this results in different relaxation times for oxygenated and deoxygenated blood (Huettel, Song, & McCarthy, 2004). Together, increased neural activity will result change in ratio of oxygenated to deoxygenated blood, which subsequently changes the relaxation time and the MR signal. fMRI has been used in two different settings: 1) during a specific task (task-based), or 2) during rest (resting-state). For the scope of this thesis, we focus on task-based fMRI. Task- based fMRI aims to identify the regions of the brain involved while performing a particular task. Changes in functional activation can be difficult to interpret and can be directly or indirectly related to changes in the structural of the brain (Tomassini et al., 2012). Damage to the structure of the brain often coincides with loss of activation in the same area of the brain due to loss of infrastructure. However, as mentioned previously, structural MRI changes are not always reflected in the clinical presentation. To still perform the same task, the brain attempts to reorganise, which is reflected in abnormal activation. Functional reorganisation, also called functional neuroplasticity, can be maladaptive or beneficial and the direction is often difficult to ascertain cross-sectionally (Schoonheim, 2017). Evidence for functional reorganisation in MS comes from studies in perception, motor function and cognition (Tomassini et al., 2012). Understanding and identifying the presence of functional reorganisation helps to better understand the clinical presentation of symptoms in MS.

Functional imaging of multiple sclerosis tremor Unfortunately, to our knowledge, no task fMRI studies have been performed in tremor in MS. General upper limb function in MS has been assessed by one group. in this study, Reddy et al. (2000) found that MS patients with no clinical motor symptoms still had an altered activity pattern in the brain. Specifically, they had a fivefold increase in activation of ipsilateral sensorimotor and premotor cortex during a simultaneous four-finger flexion extension task. In a follow-up study, they showed that the highest activation within the ipsilateral hemisphere coincides with MS patients with the worst hand function (Reddy et al., 2002). Furthermore, they showed that more diffuse brain injury also coincides with stronger ipsilateral activation. This suggests that in MS patients, more structural damage causes the brain to functionally reorganise, but this ultimately results in worsening of the hand function.

1.3.4. Conclusion To conclude, there are a variety of imaging techniques that enable visualization the pathophysiology of MS tremor. These findings could help understanding networks involved in tremor and aid development of targeted treatments. Although in-vivo human imaging studies into MS tremor are scarce, the findings that have been reported are congruent with those observed in clinical, animal and interventional treatment studies. The cerebellum and its afferents and efferents play an important part in the pathophysiology of MS tremor. Combined with studies in PD and ET, the cerebello-thalamo-cortical tract seems to be the driver of MS tremor. However, more in vivo imaging studies, specifically functional fMRI, are needed to better characterise the tremor network in MS.

1.4. Summary/motivation and thesis overview Tremor is a common and disabling feature of MS. Almost half of patients with MS experience tremor at some point during the disease course. This significantly worsens the clinical severity of MS but is also extremely embarrassing for the patient during day the day tasks. Currently, there is no effective treatment to help alleviate or ameliorate tremor in MS. To help MS patients with tremor it is important to first understand its presentation in the clinic and the brain, and then ‘cure’ it. With regarding to the clinical presentation, the assessment currently used is subjective and insensitive. Our study introduces a new technique that provides an objective, quantitative and sensitive measure of tremor. This will

provide a very detailed characterization of tremor, that would be of great use in evaluating the effect of new treatments. Regarding the presentation of MS tremor in the brain, minimal is known. Current hypothesis involves the cerebello-thalamo-cortical pathway; however, very little headway is being made which hampers progress.

This thesis aimed to provide a detailed characterisation of the pathophysiology of tremor in MS. The importance of this study is two-fold. Initially, these results are important for the large clinical trial as it will help understand how the tremor network is influenced after treatment with BoNT-A and potentially select patients based on their brain image that are most likely to respond to treatment. More broadly, the results from this thesis will help understand what causes tremor, enable better monitoring of tremor, potentially develop new treatment that effectively target tremor. To achieve our aim, we created four different experimental chapters that will be discussed next. Due to the fact that all papers are in article format, repetition of information within the introduction within each chapter is likely.

In chapter 3, we aimed to verify the hypothesis of involvement of the cerebello-thalamo- cortical pathway in MS tremor in a pilot study. To achieve this, we examined a small sample of MS tremor patients using MRI with the focus on three structures along the cerebello- thalamo-cortical tract: the cerebellum, the SCP and the thalamus. Damage to the structures was measured by atrophy and/or lesion count. We found that damage to both the thalamus and SCP is associated with worsening of the tremor. Interestingly, we did not find an association between cerebellar damage and tremor severity. These results confirm involvement of the cerebello-thalamo-cortical tract rather than specific cerebellar damage.

In chapter 4, we introduce a novel joystick fMRI task specifically designed to assess the brain in vivo during tremulous movement. The aim of this paper is to test the feasibility of this task to assess tremor pathophysiology and treatment effects in future studies. This study consists of a cross-sectional and a longitudinal part. In the cross-sectional part, we assessed the validation of the tremor-evoking potential in four participants with MS tremor, and assessed the ability to control for task-related activation in 15 healthy controls. In the longitudinal part, we determined the reproducibility of the fMRI task in the same 15 healthy controls. We found an increase in tremulous movement while performing the task in MS tremor participants compared to healthy controls. Furthermore, both the ability to

control for task-related activation and the reproducibility show good results. These results indicate our task has the feasibility to be used a pathophysiology study of MS tremor.

In chapter 5, we aimed to provide a detailed characterisation of the tremor network in MS by combining structural and functional MRI. Therefore, we applied structural MRI and the previously introduced joystick fMRI task to 69 MS participants of which 42 with unilateral upper-limb tremor and 27 without tremor. We found significant decreased volume of the and increase lesion load with both the thalamus and cerebellum in the MS tremor group compared to the MS control group. MS tremor group showed an increased activation in two separate clusters within the brain. Furthermore, an increase in activation within both clusters correlated with less severe tremor. These results suggest that both damage to cerebello-thalamo-cortical and functional neuroplasticity play a role in the pathophysiology of MS tremor.

In experimental chapter 6, we aimed to apply our understanding of the tremor network to assess the effect of the promising drug BoNT-A. At baseline, we injected 21 MS tremor participants with BoNT-A and 22 MS tremor participants with placebo. All participants were scanned at baseline and at 6 weeks after injection to determine any changes in functional activation. Furthermore, we determined tremor severity at 6 weeks and 12 weeks after BoNT-A injection. Patients that received BoNT-A treatment showed improved tremor severity and decreased functional activation. This indicates a local effect of BoNT- A on the tremor that subsequently reduces the need for compensatory activation within the tremor network.

In the final chapter of this thesis (chapter 7) we will discuss key results of our experimental chapters and we will make concluding remarks.

1.5. References Abou-Raya, A., & Abou-Raya, S. (2006). Inflammation: a pivotal link between autoimmune diseases and atherosclerosis. Autoimmun Rev, 5(5), 331-337. doi:10.1016/j.autrev.2005.12.006 Aisen, M. L., Arnold, A., Baiges, I., Maxwell, S., & Rosen, M. (1993). The effect of mechanical damping loads on disabling action tremor. Neurology, 43(7), 1346-1350.

Aisen, M. L., Holzer, M., Rosen, M., Dietz, M., & McDowell, F. (1991). Glutethimide treatment of disabling action tremor in patients with multiple sclerosis and traumatic brain injury. Arch Neurol, 48(5), 513-515. Alusi, S. H., Aziz, T. Z., Glickman, S., Jahanshahi, M., Stein, J. F., & Bain, P. G. (2001). Stereotactic lesional surgery for the treatment of tremor in multiple sclerosis: a prospective case-controlled study. Brain, 124(Pt 8), 1576-1589. Alusi, S. H., Worthington, J., Glickman, S., & Bain, P. G. (2001). A study of tremor in multiple sclerosis. Brain, 124(Pt 4), 720-730. Amato, M. P., Ponziani, G., Bartolozzi, M. L., & Siracusa, G. (1999). A prospective study on the natural history of multiple sclerosis: clues to the conduct and interpretation of clinical trials. J Neurol Sci, 168(2), 96-106. Amedei, A., Prisco, D., & D’Elios, M. M. (2012). Multiple Sclerosis: The Role of Cytokines in Pathogenesis and in Therapies. International Journal of Molecular Sciences, 13(10), 13438-13460. doi:10.3390/ijms131013438 Arrambide, G., Rovira, A., Sastre-Garriga, J., Tur, C., Castillo, J., Rio, J., . . . Tintore, M. (2018). Spinal cord lesions: A modest contributor to diagnosis in clinically isolated syndromes but a relevant prognostic factor. Mult Scler, 24(3), 301-312. doi:10.1177/1352458517697830 Aumann, T. D. (2002). Cerebello-thalamic synapses and motor adaptation. Cerebellum, 1(1), 69-77. doi:10.1080/147342202753203104 Ayache, S. S., Ahdab, R., Neves, D. O., Nguyen, J.-P., & Lefaucheur, J.-P. (2009). Thalamic stimulation restores defective cerebellocortical inhibition in multiple sclerosis tremor. Movement Disorders, 24(3), 467-469. doi:doi:10.1002/mds.22416 Ayache, S. S., Chalah, M. A., Al-Ani, T., Farhat, W. H., Zouari, H. G., Creange, A., & Lefaucheur, J. P. (2015). Tremor in multiple sclerosis: The intriguing role of the cerebellum. J Neurol Sci, 358(1-2), 351-356. doi:S0022-510X(15)02448-X [pii]

10.1016/j.jns.2015.09.360 [doi] Baecher-Allan, C., Kaskow, B. J., & Weiner, H. L. (2018). Multiple Sclerosis: Mechanisms and Immunotherapy. Neuron, 97(4), 742-768. doi:https://doi.org/10.1016/j.neuron.2018.01.021 Baghizadeh, S., Sahraian, M. A., & Beladimoghadam, N. (2013). Clinical and demographic factors affecting disease severity in patients with multiple sclerosis. Iran J Neurol, 12(1), 1-8.

Bain, P. G., Findley, L. J., Atchison, P., Behari, M., Vidailhet, M., Gresty, M., . . . Marsden, C. D. (1993). Assessing tremor severity. J Neurol Neurosurg Psychiatry, 56(8), 868-873. Balcer, L. J. (2001). Clinical outcome measures for research in multiple sclerosis. J Neuroophthalmol, 21(4), 296-301. Barkhof, F., Calabresi, P. A., Miller, D. H., & Reingold, S. C. (2009). Imaging outcomes for neuroprotection and repair in multiple sclerosis trials. Nat Rev Neurol, 5(5), 256- 266. doi:10.1038/nrneurol.2009.41 Biname, F., Sakry, D., Dimou, L., Jolivel, V., & Trotter, J. (2013). NG2 regulates directional migration of oligodendrocyte precursor cells via Rho GTPases and polarity complex proteins. J Neurosci, 33(26), 10858-10874. doi:10.1523/jneurosci.5010-12.2013 Birklein, F., Eisenbarth, G., Erbguth, F., & Winterholler, M. (2003). Botulinum toxin type B blocks sudomotor function effectively: a 6 month follow up. J Invest Dermatol, 121(6), 1312-1316. doi:10.1046/j.1523-1747.2003.12620.x Bittar, R. G., Hyam, J., Nandi, D., Wang, S., Liu, X., Joint, C., . . . Aziz, T. Z. (2005). Thalamotomy versus thalamic stimulation for multiple sclerosis tremor. J Clin Neurosci, 12(6), 638-642. doi:10.1016/j.jocn.2004.09.008 Bjartmar, C., Kidd, G., Mork, S., Rudick, R., & Trapp, B. D. (2000). Neurological disability correlates with spinal cord axonal loss and reduced N-acetyl aspartate in chronic multiple sclerosis patients. Ann Neurol, 48(6), 893-901. Bö, L., Mörk, S., Kong, P. A., Nyland, H., Pardo, C. A., & Trapp, B. D. (1994). Detection of MHC class II-antigens on macrophages and microglia, but not on astrocytes and endothelia in active multiple sclerosis lesions. Journal of Neuroimmunology, 51(2), 135- 146. doi:http://dx.doi.org/10.1016/0165-5728(94)90075-2 Bozek, C. B., Kastrukoff, L. F., Wright, J. M., Perry, T. L., & Larsen, T. A. (1987). A controlled trial of isoniazid therapy for action tremor in multiple sclerosis. J Neurol, 234(1), 36-39. Brex, P. A., Ciccarelli, O., O'Riordan, J. I., Sailer, M., Thompson, A. J., & Miller, D. H. (2002). A Longitudinal Study of Abnormalities on MRI and Disability from Multiple Sclerosis. New England Journal of Medicine, 346(3), 158-164. doi:doi:10.1056/NEJMoa011341

Brin, M. F., Lyons, K. E., Doucette, J., Adler, C. H., Caviness, J. N., Comella, C. L., . . . Koller, W. C. (2001). A randomized, double masked, controlled trial of botulinum toxin type A in essential hand tremor. Neurology, 56(11), 1523-1528. Browne, P., Chandraratna, D., Angood, C., Tremlett, H., Baker, C., Taylor, B. V., & Thompson, A. J. (2014). Atlas of Multiple Sclerosis 2013: A growing global problem with widespread inequity. Neurology, 83(11), 1022-1024. doi:10.1212/WNL.0000000000000768 Burrows, T. E. (1937). The Effects of Lesions of the Superior Cerebellar Peduncle in the Macacus rhesus Monkey. (Bull. Neur. Inst. New York, vol. v, p. 165, Aug., 1936.) Ferraro, A., and Barrera, S. E. Journal of Mental Science, 83(342), 115-115. doi:10.1192/bjp.83.342.115 Calabrese, M., De Stefano, N., Atzori, M., Bernardi, V., Mattisi, I., Barachino, L., . . . Gallo, P. (2007). Detection of cortical inflammatory lesions by double inversion recovery magnetic resonance imaging in patients with multiple sclerosis. Arch Neurol, 64(10), 1416-1422. doi:10.1001/archneur.64.10.1416 Calabrese, M., Rocca, M. A., Atzori, M., Mattisi, I., Favaretto, A., Perini, P., . . . Filippi, M. (2010). A 3-year magnetic resonance imaging study of cortical lesions in relapse- onset multiple sclerosis. Ann Neurol, 67(3), 376-383. doi:10.1002/ana.21906 Campbell, Z., Sahm, D., Donohue, K., Jamison, J., Davis, M., Pellicano, C., . . . Bagnato, F. (2012). Characterizing contrast-enhancing and re-enhancing lesions in multiple sclerosis. Neurology, 78(19), 1493-1499. doi:10.1212/WNL.0b013e3182553bd2 Carpinella, I., Cattaneo, D., & Ferrarin, M. (2014). Quantitative assessment of upper limb motor function in Multiple Sclerosis using an instrumented Action Research Arm Test. J Neuroeng Rehabil, 11, 67. doi:10.1186/1743-0003-11-67 Chansakul, C., Moguel-Cobos, G., & Bomprezzi, R. (2011). Head tremor secondary to MS resolved with rituximab. Neurol Sci, 32(6), 1157-1160. doi:10.1007/s10072-011- 0600-x Chard, D. T., Griffin, C. M., Parker, G. J., Kapoor, R., Thompson, A. J., & Miller, D. H. (2002). Brain atrophy in clinically early relapsing-remitting multiple sclerosis. Brain, 125(Pt 2), 327-337. Charil, A., Zijdenbos, A. P., Taylor, J., Boelman, C., Worsley, K. J., Evans, A. C., & Dagher, A. (2003). Statistical mapping analysis of lesion location and neurological disability in multiple sclerosis: application to 452 patient data sets. NeuroImage, 19(3), 532-544. doi:http://dx.doi.org/10.1016/S1053-8119(03)00117-4

Choi, S. R., Howell, O. W., Carassiti, D., Magliozzi, R., Gveric, D., Muraro, P. A., . . . Reynolds, R. (2012). Meningeal inflammation plays a role in the pathology of primary progressive multiple sclerosis. Brain, 135(10), 2925-2937. doi:10.1093/brain/aws189 Confavreux, C., & Vukusic, S. (2008). The clinical epidemiology of multiple sclerosis. Neuroimaging Clin N Am, 18(4), 589-622. doi:10.1016/j.nic.2008.09.002 Conrad, B., & Brooks, V. B. (1975). [Cerebellar movement disorders in monkeys. Comparison of rapidly alternating and slower target movements during cooling of the dentate nucleus (author's transl)]. J Neurol, 209(3), 165-179. Correale, J., Gaitan, M. I., Ysrraelit, M. C., & Fiol, M. P. (2017). Progressive multiple sclerosis: from pathogenic mechanisms to treatment. Brain, 140(3), 527-546. doi:10.1093/brain/aww258 Craner, M. J., Newcombe, J., Black, J. A., Hartle, C., Cuzner, M. L., & Waxman, S. G. (2004). Molecular changes in neurons in multiple sclerosis: altered axonal expression of Nav1.2 and Nav1.6 sodium channels and Na+/Ca2+ exchanger. Proc Natl Acad Sci U S A, 101(21), 8168-8173. doi:10.1073/pnas.0402765101 Dahlin-Webb, S. R. (1986). A weighted wrist cuff. Am J Occup Ther, 40(5), 363-364. De Stefano, N., Battaglini, M., & Smith, S. M. (2007). Measuring brain atrophy in multiple sclerosis. J Neuroimaging, 17 Suppl 1, 10s-15s. doi:10.1111/j.1552-6569.2007.00130.x Deuschl, G., Raethjen, J., Lindemann, M., & Krack, P. (2001). The pathophysiology of tremor. Muscle Nerve, 24(6), 716-735. Deuschl, G., Wenzelburger, R., & Raethjen, J. (2000). Tremor. Curr Opin Neurol, 13(4), 437-443. Dickerson, T. J., & Janda, K. D. (2006). The use of small molecules to investigate molecular mechanisms and therapeutic targets for treatment of botulinum neurotoxin A intoxication. ACS Chem Biol, 1(6), 359-369. doi:10.1021/cb600179d Diener, H. C., & Dichgans, J. (1992). Pathophysiology of cerebellar ataxia. Mov Disord, 7(2), 95-109. doi:10.1002/mds.870070202 Elble, R., Bain, P., João Forjaz, M., Haubenberger, D., Testa, C., Goetz, C. G., . . . Schrag, A. (2013). Task force report: Scales for screening and evaluating tremor: Critique and recommendations. Movement Disorders, 28(13), 1793-1800. doi:10.1002/mds.25648

Elble, R. J., Lyons, K. E., & Pahwa, R. (2007). Levetiracetam is not effective for essential tremor. Clin Neuropharmacol, 30(6), 350-356. doi:10.1097/WNF.0b013E31807A32C6 Fahn, S. (1986). What is it? Case 1 1986. Presentation of a case. Movement Disorders, 1(4), 275-280. doi:10.1002/mds.870010408 Fahn, S., Tolosa, E., & Marin, C. (1988). Clinical rating scale for tremor. Baltimore: Urban & Schwarzenberg. Fang, W., Lv, F., Luo, T., Cheng, O., Liao, W., Sheng, K., . . . Zhang, H. (2013). Abnormal regional homogeneity in patients with essential tremor revealed by resting-state functional MRI. PLoS One, 8(7), e69199. doi:10.1371/journal.pone.0069199 Ferguson, B., Matyszak, M. K., Esiri, M. M., & Perry, V. H. (1997). Axonal damage in acute multiple sclerosis lesions. Brain, 120 ( Pt 3), 393-399. Feys, P., D'Hooghe M, B., Nagels, G., & Helsen, W. F. (2009). The effect of levetiracetam on tremor severity and functionality in patients with multiple sclerosis. Mult Scler, 15(3), 371-378. doi:10.1177/1352458508099142 Feys, P., Helsen, W., Liu, X., Mooren, D., Albrecht, H., Nuttin, B., & Ketelaer, P. (2005). Effects of peripheral cooling on intention tremor in multiple sclerosis. J Neurol Neurosurg Psychiatry, 76(3), 373-379. doi:10.1136/jnnp.2004.044305 Feys, P., Maes, F., Nuttin, B., Helsen, W., Malfait, V., Nagels, G., . . . Liu, X. (2005). Relationship between multiple sclerosis intention tremor severity and lesion load in the brainstem. Neuroreport, 16(12), 1379-1382. Filippi, M., Paty, D. W., Kappos, L., Barkhof, F., Compston, D. A. S., Thompson, A. J., . . . Miller, D. H. (1995). Correlations between changes in disability and T2-weighted brain MRI activity in multiple sclerosis: A follow-up study. Neurology, 45(2), 255- 260. Filippi, M., Rocca, M. A., & Comi, G. (2003). The use of quantitative magnetic- resonance-based techniques to monitor the evolution of multiple sclerosis. Lancet Neurol, 2(6), 337-346. Fischer, M., & Schafer, S. S. (1999). Temperature effects on the discharge frequency of primary and secondary endings of isolated cat muscle spindles recorded under a ramp-and-hold stretch. Brain Res, 840(1-2), 1-15.

Fischl, B., Salat, D. H., Busa, E., Albert, M., Dieterich, M., Haselgrove, C., . . . Dale, A. M. (2002). Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron, 33(3), 341-355. Fisher, E., Rudick, R. A., Cutter, G., Baier, M., Miller, D., Weinstock-Guttman, B., . . . Simonian, N. A. (2000). Relationship between brain atrophy and disability: an 8- year follow-up study of multiple sclerosis patients. Mult Scler, 6(6), 373-377. Fraussen, J., Claes, N., de Bock, L., & Somers, V. (2014). Targets of the humoral autoimmune response in multiple sclerosis. Autoimmun Rev, 13(11), 1126-1137. doi:https://doi.org/10.1016/j.autrev.2014.07.002 Friese, M. A., Schattling, B., & Fugger, L. (2014). Mechanisms of neurodegeneration and axonal dysfunction in multiple sclerosis. Nat Rev Neurol, 10(4), 225-238. doi:10.1038/nrneurol.2014.37 Frischer, J. M., Weigand, S. D., Guo, Y., Kale, N., Parisi, J. E., Pirko, I., . . . Lucchinetti, C. F. (2015). Clinical and pathological insights into the dynamic nature of the white matter multiple sclerosis plaque. Ann Neurol, 78(5), 710-721. doi:10.1002/ana.24497 Gbadamosi, J., Buhmann, C., Moench, A., & Heesen, C. (2001). Failure of ondansetron in treating cerebellar tremor in MS patients--an open-label pilot study. Acta Neurol Scand, 104(5), 308-311. doi:075 [pii] Gemba, H., Sasaki, K., Yoneda, Y., Hashimoto, S., & Mizuno, N. (1980). Tremor in the monkey with a cerebellar lesion. Exp Neurol, 69(1), 173-182. Geurts, J. J., Bo, L., Pouwels, P. J., Castelijns, J. A., Polman, C. H., & Barkhof, F. (2005). Cortical lesions in multiple sclerosis: combined postmortem MR imaging and histopathology. AJNR Am J Neuroradiol, 26(3), 572-577. Giladi, N. (1997). The mechanism of action of botulinum toxin type A in focal dystonia is most probably through its dual effect on efferent (motor) and afferent pathways at the injected site. J Neurol Sci, 152(2), 132-135. Gilman, S., Carr, D., & Hollenberg, J. (1976). Kinematic effects of deafferentation and cerebellar ablation. Brain, 99(2), 311-330. Grand’Maison, F., Yeung, M., Morrow, S. A., Lee, L., Emond, F., Ward, B. J., . . . Schecter, R. (2018). Sequencing of disease-modifying therapies for relapsing– remitting multiple sclerosis: a theoretical approach to optimizing treatment. Current Medical Research and Opinion, 1-12. doi:10.1080/03007995.2018.1458023 Groiss, S. J., & Ugawa, Y. (2013). Chapter 51 - Cerebellum. In A. M. Lozano & M. Hallett (Eds.), Handb Clin Neurol (Vol. 116, pp. 643-653): Elsevier.

Hallett, M., Lindsey, J. W., Adelstein, B. D., & Riley, P. O. (1985). Controlled trial of isoniazid therapy for severe postural cerebellar tremor in multiple sclerosis. Neurology, 35(9), 1374-1377. Handel, A. E., Handunnetthi, L., Giovannoni, G., Ebers, G. C., & Ramagopalan, S. V. (2010). Genetic and environmental factors and the distribution of multiple sclerosis in Europe. Eur J Neurol, 17(9), 1210-1214. doi:10.1111/j.1468-1331.2010.03003.x Hassan, A., Ahlskog, J. E., Rodriguez, M., & Matsumoto, J. Y. (2012). Surgical therapy for multiple sclerosis tremor: a 12-year follow-up study. Eur J Neurol, 19(5), 764- 768. doi:10.1111/j.1468-1331.2011.03626.x [doi] Heesen, C., Bohm, J., Reich, C., Kasper, J., Goebel, M., & Gold, S. M. (2008). Patient perception of bodily functions in multiple sclerosis: gait and visual function are the most valuable. Mult Scler, 14(7), 988-991. doi:10.1177/1352458508088916 Heneka, M. T., Kummer, M. P., & Latz, E. (2014). Innate immune activation in neurodegenerative disease. Nat Rev Immunol, 14(7), 463-477. doi:10.1038/nri3705 Hooper, J., Taylor, R., Pentland, B., & Whittle, I. R. (1998). Rater reliability of Fahn's tremor rating scale in patients with multiple sclerosis. Arch Phys Med Rehabil, 79(9), 1076-1079. Huettel, S. A., Song, A. W., & McCarthy, G. (2004). Functional Magnetic Resonance Imaging (Vol. 1): Sinauer Associates. Hussain, R. Z., Hayardeny, L., Cravens, P. C., Yarovinsky, F., Eagar, T. N., Arellano, B., . . . Stüve, O. (2014). Immune surveillance of the central nervous system in multiple sclerosis– Relevance for therapy and experimental models. Journal of Neuroimmunology, 276(0), 9-17. doi:10.1016/j.jneuroim.2014.08.622 Jankovic, J., Schwartz, K., Clemence, W., Aswad, A., & Mordaunt, J. (1996). A randomized, double-blind, placebo-controlled study to evaluate botulinum toxin type A in essential hand tremor. Mov Disord, 11(3), 250-256. doi:10.1002/mds.870110306 Jasperse, B., Vrenken, H., Sanz-Arigita, E., de Groot, V., Smith, S. M., Polman, C. H., & Barkhof, F. (2007). Regional brain atrophy development is related to specific aspects of clinical dysfunction in multiple sclerosis. NeuroImage, 38(3), 529-537. doi:10.1016/j.neuroimage.2007.07.056 Jonkman, L. E., Soriano, A. L., Amor, S., Barkhof, F., van der Valk, P., Vrenken, H., & Geurts, J. J. G. (2015). Can MS lesion stages be distinguished with MRI? A

postmortem MRI and histopathology study. Journal of Neurology, 262(4), 1074-1080. doi:10.1007/s00415-015-7689-4 Kalincik, T., Buzzard, K., Jokubaitis, V., Trojano, M., Duquette, P., Izquierdo, G., . . . Butzkueven, H. (2014). Risk of relapse phenotype recurrence in multiple sclerosis. Mult Scler, 20(11), 1511-1522. doi:10.1177/1352458514528762 Kandel, E. R., Schwartz, J. H., & Jessel, T. M. (2000). Principles of Neural Science (Fourth ed.). New York: McGraw-Hill. Kane, N. M., & Oware, A. (2012). Nerve conduction and electromyography studies. J Neurol, 259(7), 1502-1508. doi:10.1007/s00415-012-6497-3 Kantarci, O., Siva, A., Eraksoy, M., Karabudak, R., Sutlas, N., Agaoglu, J., . . . Demirkiran, M. (1998). Survival and predictors of disability in Turkish MS patients. Turkish Multiple Sclerosis Study Group (TUMSSG). Neurology, 51(3), 765-772. Kappos, L., Moeri, D., Radue, E. W., Schoetzau, A., Schweikert, K., Barkhof, F., . . . Filippi, M. (1999). Predictive value of gadolinium-enhanced magnetic resonance imaging for relapse rate and changes in disability or impairment in multiple sclerosis: a meta-analysis. The Lancet, 353(9157), 964-969. doi:http://dx.doi.org/10.1016/S0140-6736(98)03053-0 Kell, D. B. (2009). Iron behaving badly: inappropriate iron chelation as a major contributor to the aetiology of vascular and other progressive inflammatory and degenerative diseases. BMC Medical Genomics, 2, 2-2. doi:10.1186/1755-8794-2-2 Kingwell, E., van der Kop, M., Zhao, Y., Shirani, A., Zhu, F., Oger, J., & Tremlett, H. (2012). Relative mortality and survival in multiple sclerosis: findings from British Columbia, Canada. Journal of Neurology, Neurosurgery & Psychiatry, 83(1), 61-66. doi:10.1136/jnnp-2011-300616 Kister, I., Bacon, T. E., Chamot, E., Salter, A. R., Cutter, G. R., Kalina, J. T., & Herbert, J. (2013). Natural History of Multiple Sclerosis Symptoms. International Journal of MS Care, 15(3), 146-156. doi:10.7224/1537-2073.2012-053 Klaver, R., De Vries, H. E., Schenk, G. J., & Geurts, J. J. G. (2013). Grey matter damage in multiple sclerosis: A pathology perspective. Prion, 7(1), 66-75. doi:10.4161/pri.23499 Klein, J. C., Barbe, M. T., Seifried, C., Baudrexel, S., Runge, M., Maarouf, M., . . . Hilker, R. (2012). The tremor network targeted by successful VIM deep brain stimulation in humans. Neurology, 78(11), 787-795. doi:10.1212/WNL.0b013e318249f702

Koch, M., Mostert, J., Heersema, D., & De Keyser, J. (2007). Tremor in multiple sclerosis. J Neurol, 254(2), 133-145. doi:10.1007/s00415-006-0296-7 [doi] Koller, W. C. (1984). Pharmacologic trials in the treatment of cerebellar tremor. Arch Neurol, 41(3), 280-281. Kuhlmann, T., Ludwin, S., Prat, A., Antel, J., Bruck, W., & Lassmann, H. (2017). An updated histological classification system for multiple sclerosis lesions. Acta Neuropathol, 133(1), 13-24. doi:10.1007/s00401-016-1653-y Kurtzke, J. F. (1983). Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology, 33(11), 1444-1452. Labiano-Fontcuberta, A., & Benito-León, J. (2012). Understanding Tremor in Multiple Sclerosis: Prevalence, Pathological Anatomy, and Pharmacological and Surgical Approaches to Treatment. Tremor and Other Hyperkinetic Movements, 2, tre-02-109- 765-102. Langer-Gould, A., Popat, R. A., Huang, S. M., Cobb, K., Fontoura, P., Gould, M. K., & Nelson, L. M. (2006). Clinical and demographic predictors of long-term disability in patients with relapsing-remitting multiple sclerosis: a systematic review. Arch Neurol, 63(12), 1686-1691. doi:10.1001/archneur.63.12.1686 Lebrun, C., Kantarci, O. H., Siva, A., Pelletier, D., & Okuda, D. T. (2018). Anomalies Characteristic of Central Nervous System Demyelination: Radiologically Isolated Syndrome. Neurol Clin, 36(1), 59-68. doi:10.1016/j.ncl.2017.08.004 Leone, M. A., Bonissoni, S., Collimedaglia, L., Tesser, F., Calzoni, S., Stecco, A., . . . Monaco, F. (2008). Factors predicting incomplete recovery from relapses in multiple sclerosis: a prospective study. Mult Scler, 14(4), 485-493. doi:10.1177/1352458507084650 Lin, C.-H., Chen, C.-M., Lu, M.-K., Tsai, C.-H., Chiou, J.-C., Liao, J.-R., & Duann, J.-R. (2013). VBM Reveals Brain Volume Differences between Parkinson’s Disease and Essential Tremor Patients. Frontiers in Human Neuroscience, 7(247). doi:10.3389/fnhum.2013.00247 Louveau, A., Smirnov, I., Keyes, T. J., Eccles, J. D., Rouhani, S. J., Peske, J. D., . . . Kipnis, J. (2015). Structural and functional features of central nervous system lymphatic vessels. Nature, 523(7560), 337-341. doi:10.1038/nature14432 http://www.nature.com/nature/journal/v523/n7560/abs/nature14432.html#suppleme ntary-information

Lublin, F. D., Reingold, S. C., Cohen, J. A., Cutter, G. R., Sørensen, P. S., Thompson, A. J., . . . Polman, C. H. (2014). Defining the clinical course of multiple sclerosis: the 2013 revisions. Neurology, 83(3), 278-286. doi:10.1212/WNL.0000000000000560 Ludwin, S. K., & Maitland, M. (1984). Long-term remyelination fails to reconstitute normal thickness of central myelin sheaths. J Neurol Sci, 64(2), 193-198. Lunde, H. M. B., Assmus, J., Myhr, K. M., Bo, L., & Grytten, N. (2017). Survival and cause of death in multiple sclerosis: a 60-year longitudinal population study. J Neurol Neurosurg Psychiatry, 88(8), 621-625. doi:10.1136/jnnp-2016-315238 Magliozzi, R., Howell, O. W., Reeves, C., Roncaroli, F., Nicholas, R., Serafini, B., . . . Reynolds, R. (2010). A Gradient of neuronal loss and meningeal inflammation in multiple sclerosis. Ann Neurol, 68(4), 477-493. doi:10.1002/ana.22230 Mahad, D. J., Ziabreva, I., Campbell, G., Lax, N., White, K., Hanson, P. S., . . . Turnbull, D. M. (2009). Mitochondrial changes within axons in multiple sclerosis. Brain, 132(Pt 5), 1161-1174. doi:10.1093/brain/awp046 Malpass, K. (2012). 'Outside-in' demyelination in MS. Nature Reviews Neurology, 8, 61. doi:10.1038/nrneurol.2011.217 Marcus, J. F., & Waubant, E. L. (2013). Updates on Clinically Isolated Syndrome and Diagnostic Criteria for Multiple Sclerosis. The Neurohospitalist, 3(2), 65-80. doi:10.1177/1941874412457183 Marshall, J., & Walsh, E. G. (1956). Physiological tremor. J Neurol Neurosurg Psychiatry, 19(4), 260-267. Mathieu, D., Kondziolka, D., Niranjan, A., Flickinger, J., & Lunsford, L. D. (2007). Gamma knife thalamotomy for multiple sclerosis tremor. Surg Neurol, 68(4), 394- 399. doi:10.1016/j.surneu.2006.11.049 Middleton, F. A., & Strick, P. L. (2000). Basal ganglia and cerebellar loops: motor and cognitive circuits. Brain Res Brain Res Rev, 31(2-3), 236-250. Miller, D. H., Barkhof, F., Frank, J. A., Parker, G. J., & Thompson, A. J. (2002). Measurement of atrophy in multiple sclerosis: pathological basis, methodological aspects and clinical relevance. Brain, 125(Pt 8), 1676-1695. Miller, D. H., Weinshenker, B. G., Filippi, M., Banwell, B. L., Cohen, J. A., Freedman, M. S., . . . Polman, C. H. (2008). Differential diagnosis of suspected multiple sclerosis: a consensus approach. Mult Scler, 14(9), 1157-1174. doi:10.1177/1352458508096878

Miller, G. A. (1956). The magical number seven, plus or minus two: some limits on our capacity for processing information. Psychological Review, 63(2), 81-97. doi:10.1037/h0043158 Monaca-Charley, C., Stojkovic, T., Duhamel, A., De Seze, J., Ferriby, D., & Vermersch, P. (2003). Double-blind crossover study with dolasetron mesilate, a 5-HT3 receptor antagonist in cerebellar syndrome secondary to multiple sclerosis. J Neurol, 250(10), 1190-1194. doi:10.1007/s00415-003-0176-3 [doi] MS, Parkinson's disease and physiotherapy. (2002). Drug and Therapeutics Bulletin, 40(5), 38- 40. doi:10.1136/dtb.2002.40538 Muthuraman, M., Raethjen, J., Koirala, N., Anwar, A. R., Mideksa, K. G., Elble, R., . . . Deuschl, G. (2018). Cerebello-cortical network fingerprints differ between essential, Parkinson’s and mimicked tremors. Brain, 141(6), 1770-1781. doi:10.1093/brain/awy098 Naderi, F., Javadi, S. A., Motamedi, M., & Sahraian, M. A. (2012). The efficacy of primidone in reducing severe cerebellar tremors in patients with multiple sclerosis. Clin Neuropharmacol, 35(5), 224-226. doi:10.1097/WNF.0b013e31826249bb Nakamura, R., Kamakura, K., Tadano, Y., Hosoda, Y., Nagata, N., Tsuchiya, K., . . . Shibasaki, H. (1993). MR imaging findings of tremors associated with lesions in cerebellar outflow tracts: report of two cases. Mov Disord, 8(2), 209-212. doi:10.1002/mds.870080218 Natrajan, M. (2015). Retinoid X receptor activation reverses age-related deficiencies in myelin debris phagocytosis and CNS remyelination. Nave, K.-A., & Trapp, B. D. (2008). Axon-Glial Signaling and the Glial Support of Axon Function. Annual Review of Neuroscience, 31(1), 535-561. doi:10.1146/annurev.neuro.30.051606.094309 Novakova, L., Axelsson, M., Malmeström, C., Imberg, H., Elias, O., Zetterberg, H., . . . Lycke, J. (2018). Searching for neurodegeneration in multiple sclerosis at clinical onset: Diagnostic value of biomarkers. PLoS One, 13(4), e0194828. doi:10.1371/journal.pone.0194828 Novellino, F., Cherubini, A., Chiriaco, C., Morelli, M., Salsone, M., Arabia, G., & Quattrone, A. (2013). Brain iron deposition in essential tremor: a quantitative 3- Tesla magnetic resonance imaging study. Mov Disord, 28(2), 196-200. doi:10.1002/mds.25263

Novotna, M., Paz Soldán, M. M., Abou Zeid, N., Kale, N., Tutuncu, M., Crusan, D. J., . . . Kantarci, O. H. (2015). Poor early relapse recovery affects onset of progressive disease course in multiple sclerosis. Neurology, 85(8), 722-729. doi:10.1212/WNL.0000000000001856 Patrikios, P., Stadelmann, C., Kutzelnigg, A., Rauschka, H., Schmidbauer, M., Laursen, H., . . . Lassmann, H. (2006). Remyelination is extensive in a subset of multiple sclerosis patients. Brain, 129(Pt 12), 3165-3172. doi:10.1093/brain/awl217 Perera, T., Yohanandan, S. A., Thevathasan, W., Jones, M., Peppard, R., Evans, A. H., . . . McDermott, H. J. (2016). Clinical validation of a precision electromagnetic tremor measurement system in participants receiving deep brain stimulation for essential tremor. Physiol Meas, 37(9), 1516-1527. doi:10.1088/0967-3334/37/9/1516 Perera, T., Yohanandan, S. A. C., McKay, C. M., & McDermott, H. J. (2015). Evaluation of an objective tremor assessment system for movement disorders research [online]. Paper presented at the Australian Biomedical Engineering Conference (ABEC 2015), Barton, ACT. https://search.informit.com.au/documentSummary;dn=730379337328993;res=IE LENG Peterson, J. W., Bo, L., Mork, S., Chang, A., & Trapp, B. D. (2001). Transected neurites, apoptotic neurons, and reduced inflammation in cortical multiple sclerosis lesions. Ann Neurol, 50(3), 389-400. Pittock, S. J., McClelland, R. L., Mayr, W. T., Rodriguez, M., & Matsumoto, J. Y. (2004). Prevalence of tremor in multiple sclerosis and associated disability in the Olmsted County population. Mov Disord, 19(12), 1482-1485. doi:10.1002/mds.20227 Polman, C. H., Reingold, S. C., Banwell, B., Clanet, M., Cohen, J. A., Filippi, M., . . . Wolinsky, J. S. (2011). Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol, 69(2), 292-302. doi:10.1002/ana.22366 Popescu, B. F., & Lucchinetti, C. F. (2012). Pathology of demyelinating diseases. Annu Rev Pathol, 7, 185-217. doi:10.1146/annurev-pathol-011811-132443 Raju, S. S., Niranjan, A., Monaco, E. A., 3rd, Flickinger, J. C., & Lunsford, L. D. (2018). Stereotactic radiosurgery for medically refractory multiple sclerosis-related tremor. J Neurosurg, 128(4), 1214-1221. doi:10.3171/2017.1.Jns162512 Ransohoff, R. M., & Engelhardt, B. (2012). The anatomical and cellular basis of immune surveillance in the central nervous system. Nat Rev Immunol, 12(9), 623-635. doi:10.1038/nri3265

Reddy, H., Narayanan, S., Arnoutelis, R., Jenkinson, M., Antel, J., Matthews, P. M., & Arnold, D. L. (2000). Evidence for adaptive functional changes in the cerebral cortex with axonal injury from multiple sclerosis. Brain, 123, 2314-2320. Reddy, H., Narayanan, S., Woolrich, M., Mitsumori, T., Lapierre, Y., Arnold, D. L., & Matthews, P. M. (2002). Functional brain reorganization for hand movement in patients with multiple sclerosis: defining distinct effects of injury and disability. Brain, 125(Pt 12), 2646-2657. Rice, G. P., Lesaux, J., Vandervoort, P., Macewan, L., & Ebers, G. C. (1997). Ondansetron, a 5-HT3 antagonist, improves cerebellar tremor. J Neurol Neurosurg Psychiatry, 62(3), 282-284. Richard, I., Guglielmi, M., Boisliveau, R., Perrouin-Verbe, B., Mauduyt de la Grève, I., Bioulac, B., . . . Mathé, J.-F. (1997). Load compensation tasks evoke tremor in cerebellar patients: the possible role of long latency stretch reflexes. Neuroscience Letters, 234(2–3), 99-102. doi:http://dx.doi.org/10.1016/S0304-3940(97)00681-2 Riise, T., Gronning, M., Fernandez, O., Lauer, K., Midgard, R., Minderhoud, J. M., . . . Aarli, J. A. (1992). Early prognostic factors for disability in multiple sclerosis, a European multicenter study. Acta Neurol Scand, 85(3), 212-218. Rinker, J. R., Salter, A. R., Walker, H., Amara, A., Meador, W., & Cutter, G. R. (2015). Prevalence and characteristics of tremor in the NARCOMS multiple sclerosis registry: a cross-sectional survey. BMJ Open, 5(1). doi:10.1136/bmjopen-2014- 006714 Rovini, E., Maremmani, C., & Cavallo, F. (2017). How Wearable Sensors Can Support Parkinson's Disease Diagnosis and Treatment: A Systematic Review. Front Neurosci, 11, 555. doi:10.3389/fnins.2017.00555 Sadiq, S. A., & Poopatana, C. A. (2007). Intrathecal baclofen and morphine in multiple sclerosis patients with severe pain and spasticity. J Neurol, 254(10), 1464-1465. doi:10.1007/s00415-007-0566-z [doi] Schoonheim, M. M. (2017). Functional reorganization is a maladaptive response to injury - Commentary. Mult Scler, 23(2), 194-196. doi:10.1177/1352458516677593 Schoonheim, M. M., Hulst, H. E., Brandt, R. B., Strik, M., Wink, A. M., Uitdehaag, B. M., . . . Geurts, J. J. (2015). Thalamus structure and function determine severity of cognitive impairment in multiple sclerosis. Neurology, 84(8), 776-783. doi:10.1212/wnl.0000000000001285

Schroeder, A., Linker, R. A., Lukas, C., Kraus, P. H., & Gold, R. (2010). Successful treatment of cerebellar ataxia and tremor in multiple sclerosis with topiramate: a case report. Clin Neuropharmacol, 33(6), 317-318. doi:10.1097/WNF.0b013e3181f84a39 Schuurman, P. R., Bosch, D. A., Bossuyt, P. M., Bonsel, G. J., van Someren, E. J., de Bie, R. M., . . . Speelman, J. D. (2000). A comparison of continuous thalamic stimulation and thalamotomy for suppression of severe tremor. N Engl J Med, 342(7), 461-468. doi:10.1056/nejm200002173420703 Sechi, G. P., Zuddas, M., Piredda, M., Agnetti, V., Sau, G., Piras, M. L., . . . Rosati, G. (1989). Treatment of cerebellar tremors with carbamazepine: a controlled trial with long-term follow-up. Neurology, 39(8), 1113-1115. Sharifi, S., Nederveen, A. J., Booij, J., & van Rootselaar, A.-F. (2014). Neuroimaging essentials in essential tremor: A systematic review. NeuroImage : Clinical, 5, 217-231. doi:10.1016/j.nicl.2014.05.003 Simone, I. L., Carrara, D., Tortorella, C., Liguori, M., Lepore, V., Pellegrini, F., . . . Livrea, P. (2002). Course and prognosis in early-onset MS: comparison with adult-onset forms. Neurology, 59(12), 1922-1928. Sohn, W. J., Niu, C. M., & Sanger, T. D. (2015). Increased long-latency reflex activity as a sufficient explanation for childhood hypertonic dystonia: a neuromorphic emulation study. J Neural Eng, 12(3), 036010. doi:10.1088/1741-2560/12/3/036010 Stacy, M. A., Elble, R. J., Ondo, W. G., Wu, S.-C., & Hulihan, J. (2007). Assessment of interrater and intrarater reliability of the Fahn–Tolosa–Marin Tremor Rating Scale in essential tremor. Movement Disorders, 22(6), 833-838. doi:10.1002/mds.21412 Steinman, L. (2008). Nuanced roles of cytokines in three major human brain disorders. The Journal of Clinical Investigation, 118(11), 3557-3563. doi:10.1172/JCI36532 Stewart, T., Spelman, T., Havrdova, E., Horakova, D., Trojano, M., Izquierdo, G., . . . Kalincik, T. (2017). Contribution of different relapse phenotypes to disability in multiple sclerosis. Mult Scler, 23(2), 266-276. doi:10.1177/1352458516643392 Thompson, A. J., Banwell, B. L., Barkhof, F., Carroll, W. M., Coetzee, T., Comi, G., . . . Cohen, J. A. (2018). Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria. The Lancet Neurology, 17(2), 162-173. doi:https://doi.org/10.1016/S1474-4422(17)30470-2

Tomassini, V., Matthews, P. M., Thompson, A. J., Fuglo, D., Geurts, J. J., Johansen-Berg, H., . . . Palace, J. (2012). Neuroplasticity and functional recovery in multiple sclerosis. Nat Rev Neurol, 8(11), 635-646. doi:10.1038/nrneurol.2012.179 Trapp, B. D., Peterson, J., Ransohoff, R. M., Rudick, R., Mork, S., & Bo, L. (1998). Axonal transection in the lesions of multiple sclerosis. N Engl J Med, 338(5), 278- 285. doi:10.1056/nejm199801293380502 Trapp, B. D., & Stys, P. K. (2009). Virtual hypoxia and chronic necrosis of demyelinated axons in multiple sclerosis. Lancet Neurol, 8(3), 280-291. doi:10.1016/s1474- 4422(09)70043-2 Trouillas, P., Brudon, F., & Adeleine, P. (1988). Improvement of cerebellar ataxia with levorotatory form of 5-hydroxytryptophan. A double-blind study with quantified data processing. Arch Neurol, 45(11), 1217-1222. Van der Walt, A., Buzzard, K., Sung, S., Spelman, T., Kolbe, S. C., Marriott, M., . . . Evans, A. (2015). The occurrence of dystonia in upper-limb multiple sclerosis tremor. Mult Scler. doi:10.1177/1352458515577690 Van Der Walt, A., Sung, S., Spelman, T., Marriott, M., Kolbe, S., Mitchell, P., . . . Butzkueven, H. (2012). A double-blind, randomized, controlled study of botulinum toxin type A in MS-related tremor. Neurology, 79(1), 92-99. doi:10.1212/WNL.0b013e31825dcdd9 Waxman, S. G. (2000). Multiple sclerosis as a neuronal disease. Arch Neurol, 57(1), 22-24. Wekerle, H. (2017). B cells in multiple sclerosis. Autoimmunity, 50(1), 57-60. doi:10.1080/08916934.2017.1281914 WHO. (2013). Atlas of MS 2013: Mapping Multiple Sclerosis Around the World. Retrieved from http://www.msif.org/about-ms/publications-and-resources/ Yap, L., Kouyialis, A., & Varma, T. R. (2007). Stereotactic neurosurgery for disabling tremor in multiple sclerosis: thalamotomy or deep brain stimulation? Br J Neurosurg, 21(4), 349-354. doi:10.1080/02688690701544002 Zambonin, J. L., Zhao, C., Ohno, N., Campbell, G. R., Engeham, S., Ziabreva, I., . . . Mahad, D. J. (2011). Increased mitochondrial content in remyelinated axons: implications for multiple sclerosis. Brain, 134(Pt 7), 1901-1913. doi:10.1093/brain/awr110

Project Design

To help understand the research chapters 3-6, we will first clarify the projects of which this thesis is part of.

2.1. Chapter 3 Chapter 3, is part of a phase 2 randomized controlled clinical trial with crossover to study the efficacy of BoNT-A in MS related tremor (Van Der Walt et al., 2012). The study was registered on www.clinicaltrials.gov as NCT01018485. Funding for this project was provided by the Box Hill MS Research Unit. Recruitment was completed between 2009 and 2010.

2.2. Chapters 4-6 As mentioned previously, these chapters of current thesis were part of a larger project called TREMTOX (ACTRN12617000379314). This project started in August 2015 and I joined the project in November 2015. Recruitment for this study closed in July 2018. Analyses for the other aims in the project are thought to continue to 2019.

2.0.1. Investigators Principal investigator: Dr. Anneke van der Walt

Associate investigators: Dr. Andrew Evans, A/Prof Helmut Butzkueven, Dr. Simon Sung, Dr. Thushara Perera, Dr. Scott Kolbe, Dr. Mark Marriot, A/Prof Lynette Kiers, Dr. Adam Vogel, Ms. Camille O’ Shanahan, Dr. Gustavo Noffs and Ms Pippa Iva.

2.0.2. Overall study aims This research project aimed to perform a detailed study of phenomenological, electrophysiological and neuroimaging characteristics of MS tremor patients before and after BoNT-A or placebo treatment. To achieve this, the following three aims were designed:

1. Determine the efficacy and safety of BoNT-A treatment for arm tremor in MS in a study of 60 MS tremor patients. 2. Perform a neuroimaging study comparing MRI features between MS patients with and without tremor in order to define the underlying pathophysiology of tremor subtypes. 3. Assess the correlation between baseline covariates and response to BoNT-A treatment using data obtained in aims 1 and 2.

2.0.3. Participants recruitment The project recruited 59 adults below the age of sixty-five years with confirmed MS and upper limb tremor in a randomised, double blind, placebo-controlled (RCT) study. Participants were recruited from multiple hospitals in and around Melbourne. Tremor patients will be studied at baseline using an extensive protocol before being randomized to receive BoNT-A or placebo treatment for their upper limb tremor. Follow-up assessments were performed at 6, 12 and 24 weeks after treatment. For schedule of assessments, see Table 1. Follow-up measures to ensure patient retention in the study included phone and mail contact in the week prior to scheduled appointments. Patients will be offered open label ongoing treatment through the Botox clinic at Royal Melbourne Hospital at the conclusion of the study. Of all MS tremor patients, 43 were MRI compatible and underwent an MRI brain at baseline and week 6.

In addition, we recruited 27 MS participants without upper limb tremor. Each MS control was matched for age, EDSS and disease duration on a one on one basis with an MRI compatible MS tremor participant. MS controls underwent the same baseline visit as the MS tremor participants apart from the injections. Furthermore, 15 healthy controls were

recruited that were matched for age with the MRI compatible MS tremor participants. Healthy controls also underwent a baseline visit, excluding the EDSS and injections, and they also had a second MRI brain scan at week 6.

2.0.4. Ethics The study was approved by the Melbourne Health Human Research Ethics Committee. All of the participants gave informed voluntary written consent.

Participants 101

Healthy MS controls 86 15

MS control MS tremor 27 59

MRI MRI compatible incompatible 43 16

BoNT-A Placebo 21 22

Figure 1. Participant groups overview. Colour-coded in red are the groups that have been included in this thesis.

2.0.5. Inclusion criteria 1. a definite diagnosis of RRMS or SPMS 2. aged between 18 and 65 years 3. no other neurological disease to explain presence of tremor 4. no MS relapse or treatment with corticosteroids in the 3 months prior to enrolment 5. normal or near-normal upper limb strength (Medical Research Council (MRC) score > 4+/5).

2.0.6. Exclusion criteria 1. Treatment of tremor with BoNT-A of any brand within the 6 months prior to enrolment 2. A known contraindication to BoNT-A injection 3. Pregnancy, breastfeeding or unwillingness to use adequate contraception. 4. Inability to cease other tremor treatments (wash-out period of 4 weeks for any tremor reducing agents) 5. A MS relapse during the study that affects upper limb function will be a withdrawal criterion.

2.0.7. Contribution As part of this thesis, all imaging has been collected by me. Furthermore, I was the main contact person for the participants in to scheduling their visits, reminders about their appointments and any other questions they have regarding the study. Finally, in a few participants I have assisted with their clinical assessment.

Table 1. Visit protocol. * MS tremor, † MS controls, ‡ healthy controls. Abbreviations: RCT, randomised, double blind, placebo-controlled study; EDSS, expanded disability status scale. Colour-coded in the red are the parts of assessments that have been used in this thesis.

Study RCT Open label Baseline Visit 1 Visit 2 Visit 3 Visit 4 Visit 5 Visits Week Week Week Week Week Week 0 6±2 12±2 24±2 32±2 48±2 Medical history, MS history and * † ‡

General progression Inclusion/exclusion * criteria BoNT-A/placebo * * injections EDSS * † * * Cerebellar assessment * † ‡ * * * * *

Standardized video * † ‡ * * * * * assessment Tremor clinical assessment and * † ‡ * * * * * Assessment functional tasks Dystonia assessment * † ‡ * * * * * 51

Electrophysiological * † ‡ * * * * * measurements Weakness assessment * * * * * * Adverse events * * * * * * assessment Quality of Life * † ‡ * * * * * questionnaires MRI Brain * † ‡ * ‡

2.3. References Van Der Walt, A., Sung, S., Spelman, T., Marriott, M., Kolbe, S., Mitchell, P., . . . Butzkueven, H. (2012). A double-blind, randomized, controlled study of botulinum toxin type A in MS-related tremor. Neurology, 79(1), 92-99. doi:10.1212/WNL.0b013e31825dcdd9

52 3

Pilot Study

Tremor in multiple sclerosis is associated with cerebello- thalamic pathology. Frederique Boonstra, MSc. a, g, Grace Florescu, MDDS b, g, Andrew Evans, PhD d, Chris Steward, PhD b, c, Peter Mitchell, PhD b, Patricia Desmond, PhD b, Brad Moffat, PhD a, Helmut Butzkueven, PhD d,e,f, Scott Kolbe, PhD a, Anneke van der Walt, PhD d, e a. Department of Anatomy and Neuroscience, University of Melbourne, Australia b. Department of Radiology, Royal Melbourne Hospital, University of Melbourne, Australia c. Department of Medicine, Royal Melbourne Hospital, Australia d. Department of Neurology, Royal Melbourne Hospital, Australia e. Melbourne Brain Centre at Royal Melbourne Hospital, Department of Medicine, University of Melbourne, Australia. f. Multiple Sclerosis Unit, Box Hill Hospital, Box Hill, Australia. g. Authors contributed equally

Published in the journal of neural transmission on 2nd of November 2017

3.1. Abstract Background: Tremor in people with multiple sclerosis (MS) is a frequent and debilitating symptom with a relatively poorly understood pathophysiology. Objective: To determine the relationship between clinical tremor severity and structural magnetic resonance imaging (MRI) parameters. Methods: Eleven patients with clinically definite MS and right-sided upper limb tremor were studied. Tremor severity was assessed using the Bain score (overall severity, writing, and Archimedes spiral drawing). Cerebellar dysfunction was assessed using the Scale for the Assessment and Rating of Ataxia (SARA). Dystonia was assessed using the Global Dystonia Scale (GDS) adapted for upper limb. For all subjects, volume was calculated for the thalamus from T1-weighted volumetric scans using Freesurfer. Superior cerebellar peduncle (SCP) cross-sectional areas were measured manually. The presence of lesions was visually determined and the lesion volumes were calculated by the lesion growth algorithm as implemented in the Lesion Segmentation Toolbox. Results: Right thalamic volume negatively correlated with Bain tremor severity score (r=- 0.65, p=0.03). Left thalamic volume negatively correlated with general Bain tremor severity score (r=-0.65, p=0.03) and the Bain writing score (r=-0.65, p=0.03). Right SCP area negatively correlated with Bain writing score (r=-0.69, p=0.02). Finally, Bain Archimedes score was significantly higher in patients with lesions in the contralateral thalamus. Whole brain lesion load showed no relationship with tremor severity. Conclusions: These results implicate degeneration of key structures within the cerebello- thalamic pathway as pathological substrates for tremor in MS patients.

3.2. Introduction Tremor in multiple sclerosis (MS) is a common and disabling symptom. Epidemiological data suggests that tremor affects 26-58% of people with MS and can significantly impact their quality of life (Alusi, Worthington, Glickman, & Bain, 2001; Pittock, McClelland, Mayr, Rodriguez, & Matsumoto, 2004). MS tremor most commonly affects the upper limbs, but has also been observed in the legs, head and trunk (Alusi, Glickman, Aziz, & Bain, 1999). MS tremor has postural components and intention components (Koch, Mostert, Heersema, & De Keyser, 2007), and these features can limit patients’ daily functions such that those affected are more likely to be unemployed or retired as a result of their disability (Pittock et al., 2004).

The pathophysiology of MS-related tremor remains poorly understood, as the multifocal nature of MS lesions makes it difficult to identify the precise neuroanatomical substrates of tremor (Polman et al., 2011). Essential tremor (ET) has been more extensively investigated. In ET patients, Contarino et al. (2012) reported a correlation between tremulous movement measured with EEG and fMRI activation in the cerebellum and thalamus. Furthermore, diffusion studies in ET have shown abnormalities in the white matter in the cerebellum, cerebellar peduncles, brainstem, and cerebellar hemispheres. Together, studies on the pathophysiology in ET suggests an important role for the cerebellum and the cerebello-thalamo-cortical pathways (Bucher, Seelos, Dodel, Reiser, & Oertel, 1997; Colebatch, Findley, Frackowiak, Marsden, & Brooks, 1990; Deuschl, Raethjen, Lindemann, & Krack, 2001; Louis et al., 2002).

Without an understanding of the pathophysiology of MS tremor, development of adequate treatment options is challenging. In MS, cerebellar injury was traditionally thought to be the main pathological driver of MS tremor due to the predominant presence of intention tremor and the correlation between the severity of tremor and dysmetria, dysarthria and dysdiadochokinesia (Diener & Dichgans, 1992). However, more recently, careful clinical evaluation of patients with MS tremor has highlighted the clinical commonalities with essential tremor and questioned the relevance of purely cerebellar features on clinical examination (Van der Walt et al., 2015; Van Der Walt et al., 2012). These two previous case studies observed the presence of lesions in the cerebellar peduncles (superior and middle) in patients with MS tremor as well as cerebellum (Fahn, 1986; Nakamura et al., 1993). Furthermore, animal studies investigating tremor pathogenesis and surgical treatment with deep brain stimulation (DBS) targeting the thalamic nuclei in MS tremor highlight an interaction between the cerebellum and thalamus (Deuschl & Bergman, 2002; Koch et al., 2007; Yap, Kouyialis, & Varma, 2007). Together, findings in Essential Tremor and MS indicated that tremor in people with MS could result from damage to the cerebello- thalamic connections via the superior cerebellar peduncle (SCP).

For the justification of a future larger study in MS patient with tremor, we conducted a pilot study that aimed to assess the hypothesis that damage to the cerebellar tracts rather than cerebellar hemispheres, specifically the cerebello-thalamic tract, is related to tremor severity in patients with clinically definite MS and unilateral upper limb tremor. We

hypothesized that atrophy of structures in the tract (the SCP and thalami) indicative of neuro-axonal degeneration within these structures would correlate with the severity of tremor. This study might provide the necessary support for future larger studies.

3.3. Methods

Participants Fourteen patients with clinically definite MS (Polman et al., 2005) and predominantly unilateral MS-related tremor were prospectively identified from the Royal Melbourne Hospital MS Clinic from 2009 to 2010 prior to enrolment in a phase 2 clinical trial examining the efficacy of onabotulinumtoxin-a in MS-related tremor (Van Der Walt et al., 2012). Left-hand dominant participants (n=3) were excluded from MRI analysis to focus on lateralization of the tremor pathogenesis in a clearly lateralized group of greatest size. At the time of testing, no patients were on tremor modulating treatment. The project was approved by the Royal Melbourne Hospital Human Research Ethics Committee and informed consent was obtained from each participant prior to participation.

Clinical assessments Detailed phenotypical data were collected for each patient, including demographic data, EDSS (Expanded Disability Severity Scale) (Kurtzke, 1983), and cerebellar dysfunction scores (Scale for the Assessment and Rating of Ataxia (SARA)) (Schmitz-Hubsch et al., 2006). A movement disorder neurologist completed a clinical tremor assessment as has been previously described (Van der Walt et al., 2015). Tremor in the affected limb was rated using the Bain score (Bain et al., 1993). The Bain score is a subjective rating scale from 0 to 10 (0=no tremor and 10=extremely severe) for overall tremor severity, tremor during handwriting and tremor during drawing of an Archimedes spiral on a pre-drawn pattern. Dystonia (Van der Walt et al., 2015) in the tremor-affected arm was assessed using the Global Dystonia Scale (GDS) in the upper limb only (range 0 to 10).

Neuroimaging Imaging was performed on a 3 Tesla MRI system (Siemens, Erlangen, Germany). and included a 3D MPRAGE T1-weighted scan (TR=1900, TE=2.26, TI=900, FOV= 224x256mm2, BW=200, flip angle=9, slice thickness=1mm and matrix size=256x215mm) and 3D SPACE FLAIR scan (TR=6000, TE=405.0, TI=2100, FOV=224x256, BW=700, 56

flip angle=120, slice thickness=1 and matrix size=256x222). Regions of interest in the cerebello-thalamic tract included the thalamus and the SCP. T1-weighted MRI scans were processed using Freesurfer (version 5.3) and the standard processing pipeline to obtain volume of the thalamus. The area of the SCP was manually labelled using OsiriX Lite v7.0.4 (Geneva, Switzerland) by two independent raters GF and FB (intraclass correlation = 0.729). The axis in OsiriX were aligned perpendicular and parallel to the cerebellar stalk (see figure 1 for a schematic overview of the alignment of the SCP) and then a closed polygon shape was drawn by the raters. The borders of the SCP were defined as by Akhlaghi et al. (2011). Finally, lesions have been segmented by the lesion growth algorithm (Schmidt et al., 2012) as implemented in the LST toolbox version 2.0.15 (www.statisticalmodelling.de/lst.html) for SPM. The lesion growth algorithm is able to segment T2-hyperintense lesions from a combination of T1 and FLAIR images. Default settings have been used and this technique have been validated in MS patients (Schmidt et al., 2012). For lesion localization, a neuro-radiologist visually analysed the scans and determined the presence or absence of any lesions within the thalamus and whole cerebellum (including SCP) for each hemisphere.

Figure 1. Alignment and manual labelling of the SCP. Step 1 in sagittal view: move the intersection of the axis to the middle of the cerebellar stalk and rotate the blue axis parallel to and the purple axis perpendicular to the cerebellar stalk. Step 2 and step 3: rotate the orange axis parallel to the either left or right cerebellar stalk. Step 4: use a closed polygon shape to map the area of the SCP

Statistical Analyses All statistical analyses were performed using SPSS 23 (IBM). The thalamic volume, SCP area and total lesion load were corrected for intracranial volume (ICV) and age by linear regression. Correlations between tremor scores and corrected imaging parameters were done using non-parametric spearman’s correlation and Mann-Whitney U tests as clinical scales are ordinal rather than continual. P-values below 0.05 were considered significant without correction for multiple comparisons as this is a preliminary exploratory study with low subject numbers.

3.4. Results

Subject demographic and disease data are reported in Table 1. Overall, patients were moderately disabled (EDSS 3 to 5.5), and had a mild to moderate tremor severity (mean Bain tremor severity score 3.21). Half of our patients had a lesion(s) within the ipsilateral and/or contralateral cerebellum/SCP (54.5% and 45.5% respectively), 27.3% of the patients had a lesion(s) in the ipsilateral thalamus and 18.2% had a lesion(s) in the contralateral thalamus.

Table 1 Patient demographics. Abbreviations: EDSS, expended disability status score; SARA, Scale for the Assessment and Rating of Ataxia; GDS, global dystonia scale (upper limb only); ICV, intracranial volume; SCP, superior cerebellar peduncle. Cohort n=11 Age, years (mean, sd) 54.1 (+/-11.4) Gender, male (%) 36.4 Disease duration, years (mean, sd) 16.3 (+/-8.7) EDSS score (median, iqr) 4.5 (3-5.5) SARA score (mean, sd) 11.6 (7.7) Bain tremor severity score (median, iqr) 3 (2-3) Bain writing score (median, iqr) 2 (1-4) Bain Archimedes spiral score (median, iqr) 1 (1-4) GDS (mean, sd) 3.1 (+/-1.2) ICV, x105 mm3 (mean, sd) 14.5 (+/-1.5) Thalamic volume, right mm3 (mean, sd) 6058 (+/-916) Thalamic volume, left mm3 (mean, sd) 6865 (+/-1267) SCP area, right mm2 (mean, sd) 22.4 (+/-3.3) SCP area, left mm2 (mean, sd) 22.4 (+/-2.7) Lesion load, ml (mean, sd) 14.2 (+/-10.9) Lesions in Cerebellum/SCP, right (%) 54.5 Lesions in Cerebellum/SCP, left (%) 45.5 Lesions in Thalamus, right (%) 27.3 Lesions in Thalamus, left (%) 18.2

Correlation results are summarized in Table 2. There was a significant negative correlation between right thalamic volume and Bain tremor severity score (r=-0.647, p=0.031). Left thalamic volume inversely correlated with both the overall Bain tremor severity score (r=- 59

0.647, p=0.031) and the Bain writing score (r=-0.650, p=0.030). In addition, our data showed a significant negative correlation between right SCP area and Bain Writing score (r=-0.693, p=0.018). Finally, loss of volume in the left SCP area was associated with a higher upper limb GDS (r=-0.697, p=0.017). There were no significant correlations between whole brain lesion load and any of the tremor severity scores; however, patients with lesion(s) within the contralateral thalamus had an increased Bain Archimedes score (U=0.5, p=0.036) compared to patients without lesion(s) within the contralateral thalamus. None of the imaging parameters correlated with the SARA score.

Table 2 Correlation between tremor and dystonia severity scores and MRI measures. * and bold text indicates significance p<0.05. Abbreviation: SARA, Scale for the Assessment and Rating of Ataxia; GDS, Global Dystonia Scale (upper limb only); SCP, superior cerebellar peduncle; r, Spearman Rho; U, Mann-Whitney U. Bain Bain Bain SARA GDS (r tremor writing Archimedes score (r value, p severity score (r spiral score value, p value) score (r value, p (r value, p value value, p value) value) value) Thalamic volume, right r=-0.647 r=-0.593 r=-0.507 r=-0.427 r=-0.524 p=0.031* p=0.055 p=0.112 p=0.190 p=0.098 Thalamic volume, left r=-0.647 r=-0.650 r=-0.507 r=-0.309 r=-0.557 p=0.031* p=0.030* p=0.112 p=0.255 p=0.075* SCP area, right r=-0.338 r=-0.693 r=-0.578 r=-0.154 r=-0.509 p=0.309 p=0.018* p=0.062 p=0.670 p=0.110 SCP area, left r=-0.162 r=-0.497 r=-0.382 r=0.191 r=-0.697 p=0.635 p=0.120 p=0.246 p=0.574 p=0.017* Lesions

Whole brain (volume) r=0.206 r=0.363 r=0.406 r=-0.136 r=0.106 p=0.543 p=0.272 p=0.215 p=0.689 p=0.757 Cerebellar/SCP, right U=13.5 U=6.0 U=7.5 U=12.0 U=11.0 (presence) p=0.792 p=0.126 p=0.177 p=0.662 p=0.537 Cerebellar/SCP, left U=15.0 U=14.0 U=12.5 U=9.0 U=14.5 (presence) p=1.00 p=0.931 p=0.662 p=0.329 p=0.931

Thalamus, right U=9.0 U=10.0 U=7.0 U=11.0 U=8.0 (presence) p=0.630 p=0.776 p=0.376 p=0.921 p=0.497 Thalamus, left U=3.0 U=2.0 U=0.5 U=9.0 U=2.0 (presence) p=0.218* p=0.145 p=0.036* p=1.00 p=0.145

3.5. Discussion This study investigated the involvement of the cerebello-thalamic tract in the pathogenesis of unilateral MS upper limb tremor. The thalamus was of particular interest as thalamic atrophy had been correlated with worse prognosis in MS patients (Rocca et al., 2010), is present in the very early stages of MS (Azevedo et al., 2015; Bergsland et al., 2012), and has been shown to allow differentiation from other MRI white matter pathologies (Solomon, Watts, Dewey, & Reich, 2017). We found that contra- and ipsilateral thalamic volumes and ipsilateral SCP area inversely correlated with increased unilateral tremor severity scores. These regions of volume loss are in line with the predicted neuroanatomy of the cerebello- thalamic tract. The lack of significant correlation between neuroimaging parameters and cerebellar dysfunction measured by the SARA score, likely reflects the effects of gait and unaffected upper limb scores when calculating the total SARA score.

Our findings are consistent with involvement of the cerebello-thalamic tract in MS tremor pathogenesis. Traditional views of MS tremor emphasize a central role for the cerebellum in MS tremor due to the predominant presence of intention tremor and the correlation between the severity of tremor and dysmetria, dysarthria and dysdiadochokinesia. However, animal studies of tremor pathogenesis and surgical treatment with deep brain stimulation (DBS) targeting the thalamic nuclei in MS tremor highlight an interaction between the cerebellum, cortex, basal ganglia and thalamus (Deuschl & Bergman, 2002; Koch et al., 2007; Yap et al., 2007). Clinical neuroimaging studies of MS tremor are, however, rare. Two case reports have both associated an acute SCP demyelinating lesion with acute tremor onset (Fahn, 1986; Nakamura et al., 1993). Combined with a positive clinical trial with onabotulinumtoxin-a, a treatment typically used in focal dystonias, (Van Der Walt et al., 2012), these studies provide further support for tract-based pathogenesis of tremor. Finally, Feys et al. (2005) demonstrated a link between MS tremor severity and T2 lesion load in the contralateral pons but not with the lesion load in the cerebellum.

We found no evidence of a correlation between the tremor severity scores and the overall lesion load in the brain in mild-to moderate unilateral MS tremor. Interestingly, the presence of lesions within in contralateral thalamus correlated to an increase in tremor severity in unilateral MS tremor. These findings support the lateralization and localization of tremor pathophysiology. Consistent with this, Feys et al. (2005) studied the relationship between lesion load and tremor severity in a significantly disabled cohort (EDSS from 6-8) where 93% of participants had bilateral tremor. The authors found no correlation between overall lesion load and tremor severity; however, they showed the importance of lesion topography and tremor severity. Specifically, they found a correlation between lesion load in the contralateral pons and tremor severity. Furthermore, the correlation was related to the degree of bilateral tremor. The pons is known to contains nuclei that are involved in a variety of functions that are commonly symptomatic in MS, including pyramidal function, bladder control, equilibrium and facial sensation (Brodal, 2014; Saladin Kenneth, 2007) and a correlation between pontine lesion load and severe bilateral tremor scores are therefore not surprising. Here, we further highlight the importance of lesion topography and further study into lesion topography and tremor severity in unilateral tremor is warranted.

Our findings were demonstrated in a small sample size limiting our ability to correct for EDSS, disease duration and other factors. Furthermore, the lack of controls and cross- sectional nature of the study hinders any cause-and-effect analyses. Despite these limitations, this study is hypothesis generating in an area that is currently poorly understood.

In conclusion, we have found significant correlations between functional measures of tremor and brain abnormalities as indicated by MRI. The data support the feasibility of further, larger studies using more advanced imaging techniques (e.g. tractography, DWI functional MRI) to better define tremor pathogenesis in MS.

3.6. Author’s roles Boonstra F contributed to data analyses, data interpretation and preparation of manuscript. Florescu G contributed to data analyses, data interpretation and preparation of manuscript. Evans A contributed to study design, data interpretation and preparation of manuscript. Steward C has worked on data analyses and preparation of manuscript.

Mitchell P contributed to MRI methodology, data analyses, data interpretation and preparation of manuscript. Desmond P contributed to preparation of manuscript. Moffat B contributed to preparation of manuscript. Butzkueven H contributed to study design, clinical recruitment, data interpretations and preparation of manuscript. Kolbe S contributed to study design, MRI methodology, data analysis and interpretation and preparation of manuscript. Van der Walt A contributed to study design, data collection, data interpretation and preparation of manuscript.

3.7. Financial Disclosures of all authors F. Boonstra has nothing to declare G. Florescu has nothing to declare C. Steward has nothing to declare A. Evans has received honoraria from Novartis for giving presentations and providing consultancy services. He has participated in scientific advisory board meetings for Novartis, UCB Pharma, Allergan, and Boehringer Ingelheim. He has received conference travel support from Boehringer Ingelheim. H. Butzkueven has received consulting fees from Genzyme, Biogen, Novartis, Merck and Oxford PharmaGenesis and grant/research support from Biogen, Novartis, Merck and Genzyme P. Mitchell has nothing to declare P. Desmond has nothing to declare B. Moffat receives a fellowship from the National Institutes of Health S. Kolbe receives grant income from the National Health and Medical Research Council of Australia and has received honoraria from Novartis A van der Walt has received travel support from Biogen Idec, Novartis, Teva, Merck and serves on several advisory boards. She receives grant and fellowship funding from the National Health and Medical Research Council of Australia.

3.8. References Akhlaghi, H., Corben, L., Georgiou-Karistianis, N., Bradshaw, J., Storey, E., Delatycki, M. B., & Egan, G. F. (2011). Superior cerebellar peduncle atrophy in Friedreich's 63

ataxia correlates with disease symptoms. Cerebellum, 10(1), 81-87. doi:10.1007/s12311-010-0232-3 Alusi, S. H., Glickman, S., Aziz, T. Z., & Bain, P. G. (1999). Tremor in multiple sclerosis. J Neurol Neurosurg Psychiatry, 66(2), 131-134. Alusi, S. H., Worthington, J., Glickman, S., & Bain, P. G. (2001). A study of tremor in multiple sclerosis. Brain, 124(Pt 4), 720-730. Azevedo, C. J., Overton, E., Khadka, S., Buckley, J., Liu, S., Sampat, M., . . . Pelletier, D. (2015). Early CNS neurodegeneration in radiologically isolated syndrome. Neurology® Neuroimmunology & Neuroinflammation, 2(3), e102. doi:10.1212/NXI.0000000000000102 Bain, P. G., Findley, L. J., Atchison, P., Behari, M., Vidailhet, M., Gresty, M., . . . Marsden, C. D. (1993). Assessing tremor severity. J Neurol Neurosurg Psychiatry, 56(8), 868-873. Bergsland, N., Horakova, D., Dwyer, M. G., Dolezal, O., Seidl, Z. K., Vaneckova, M., . . . Zivadinov, R. (2012). Subcortical and cortical gray matter atrophy in a large sample of patients with clinically isolated syndrome and early relapsing-remitting multiple sclerosis. AJNR Am J Neuroradiol, 33(8), 1573-1578. doi:10.3174/ajnr.A3086 Brodal, P. (2014). Pontine Nuclei. In Encyclopedia of the Neurological Sciences (Second Edition) (pp. 938-940). Oxford: Academic Press. Bucher, S. F., Seelos, K. C., Dodel, R. C., Reiser, M., & Oertel, W. H. (1997). Activation mapping in essential tremor with functional magnetic resonance imaging. Annals of Neurology, 41(1), 32-40. doi:10.1002/ana.410410108 Colebatch, J. G., Findley, L. J., Frackowiak, R. S., Marsden, C. D., & Brooks, D. J. (1990). Preliminary report: activation of the cerebellum in essential tremor. Lancet, 336(8722), 1028-1030. Contarino, M. F., Groot, P. F. C., van der Meer, J. N., Bour, L. J., Speelman, J. D., Nederveen, A. J., . . . van Rootselaar, A.-F. (2012). Is There a Role for Combined EMG-fMRI in Exploring the Pathophysiology of Essential Tremor and Improving Functional Neurosurgery? PLoS One, 7(10), e46234. doi:10.1371/journal.pone.0046234 Deuschl, G., & Bergman, H. (2002). Pathophysiology of nonparkinsonian tremors. Mov Disord, 17 Suppl 3, S41-48. Deuschl, G., Raethjen, J., Lindemann, M., & Krack, P. (2001). The pathophysiology of tremor. Muscle Nerve, 24(6), 716-735.

Diener, H. C., & Dichgans, J. (1992). Pathophysiology of cerebellar ataxia. Mov Disord, 7(2), 95-109. doi:10.1002/mds.870070202 Fahn, S. (1986). What is it? Case 1 1986. Presentation of a case. Movement Disorders, 1(4), 275-280. doi:10.1002/mds.870010408 Feys, P., Maes, F., Nuttin, B., Helsen, W., Malfait, V., Nagels, G., . . . Liu, X. (2005). Relationship between multiple sclerosis intention tremor severity and lesion load in the brainstem. Neuroreport, 16(12), 1379-1382. Koch, M., Mostert, J., Heersema, D., & De Keyser, J. (2007). Tremor in multiple sclerosis. J Neurol, 254(2), 133-145. doi:10.1007/s00415-006-0296-7 [doi] Kurtzke, J. F. (1983). Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology, 33(11), 1444-1452. Louis, E. D., Shungu, D. C., Chan, S., Mao, X., Jurewicz, E. C., & Watner, D. (2002). Metabolic abnormality in the cerebellum in patients with essential tremor: a proton magnetic resonance spectroscopic imaging study. Neurosci Lett, 333(1), 17-20. Nakamura, R., Kamakura, K., Tadano, Y., Hosoda, Y., Nagata, N., Tsuchiya, K., . . . Shibasaki, H. (1993). MR imaging findings of tremors associated with lesions in cerebellar outflow tracts: report of two cases. Mov Disord, 8(2), 209-212. doi:10.1002/mds.870080218 Pittock, S. J., McClelland, R. L., Mayr, W. T., Rodriguez, M., & Matsumoto, J. Y. (2004). Prevalence of tremor in multiple sclerosis and associated disability in the Olmsted County population. Mov Disord, 19(12), 1482-1485. doi:10.1002/mds.20227 Polman, C. H., Reingold, S. C., Banwell, B., Clanet, M., Cohen, J. A., Filippi, M., . . . Wolinsky, J. S. (2011). Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol, 69(2), 292-302. doi:10.1002/ana.22366 Polman, C. H., Reingold, S. C., Edan, G., Filippi, M., Hartung, H. P., Kappos, L., . . . Wolinsky, J. S. (2005). Diagnostic criteria for multiple sclerosis: 2005 revisions to the "McDonald Criteria". Ann Neurol, 58(6), 840-846. doi:10.1002/ana.20703 Rocca, M. A., Mesaros, S., Pagani, E., Sormani, M. P., Comi, G., & Filippi, M. (2010). Thalamic Damage and Long-term Progression of Disability in Multiple Sclerosis. Radiology, 257(2), 463-469. doi:10.1148/radiol.10100326 Saladin Kenneth, S. (2007). Anatomy & physiology the unity of form and function. Dubuque, IA: McGraw-Hill. Schmidt, P., Gaser, C., Arsic, M., Buck, D., Forschler, A., Berthele, A., . . . Muhlau, M. (2012). An automated tool for detection of FLAIR-hyperintense white-matter

lesions in Multiple Sclerosis. NeuroImage, 59(4), 3774-3783. doi:10.1016/j.neuroimage.2011.11.032 Schmitz-Hubsch, T., du Montcel, S. T., Baliko, L., Berciano, J., Boesch, S., Depondt, C., . . . Fancellu, R. (2006). Scale for the assessment and rating of ataxia: development of a new clinical scale. Neurology, 66(11), 1717-1720. doi:10.1212/01.wnl.0000219042.60538.92 Solomon, A. J., Watts, R., Dewey, B. E., & Reich, D. S. (2017). MRI evaluation of thalamic volume differentiates MS from common mimics. Neurology - Neuroimmunology Neuroinflammation, 4(5). doi:10.1212/nxi.0000000000000387 Van der Walt, A., Buzzard, K., Sung, S., Spelman, T., Kolbe, S. C., Marriott, M., . . . Evans, A. (2015). The occurrence of dystonia in upper-limb multiple sclerosis tremor. Mult Scler, 21(14), 1847-1855. doi:10.1177/1352458515577690 Van Der Walt, A., Sung, S., Spelman, T., Marriott, M., Kolbe, S., Mitchell, P., . . . Butzkueven, H. (2012). A double-blind, randomized, controlled study of botulinum toxin type A in MS-related tremor. Neurology, 79(1), 92-99. doi:10.1212/WNL.0b013e31825dcdd9 Yap, L., Kouyialis, A., & Varma, T. R. (2007). Stereotactic neurosurgery for disabling tremor in multiple sclerosis: thalamotomy or deep brain stimulation? Br J Neurosurg, 21(4), 349-354. doi:10.1080/02688690701544002

66 4

Validation Study

Novel functional MRI task for studying the neural correlates of upper limb tremor Frederique M C Boonstra1, Thushara Perera4,5, Gustavo Noffs2,3, Cassandra Marotta3,6, Adam P. Vogel3,4,6,7, Andrew H Evans2, Helmut Butzkueven2,8,10, Bradford A Moffat1, Anneke van der Walt2,8,10, Scott C Kolbe1,9 1. Department of Medicine, University of Melbourne, Australia 2. Department of Neurology, Royal Melbourne Hospital, Australia 3. Centre for Neuroscience of Speech, University of Melbourne, Victoria, Australia 4. The Bionics Institute, East Melbourne, Australia 5. Department of Medical Bionics, University of Melbourne, Australia 6. Redenlab, Australia 7. Department of Neurodegeneration, Hertie Institute for Clinical Brain Research, University of Tübingen, Germany 8. Melbourne Brain Centre, Department of Medicine, University of Melbourne, Australia 9. Florey Institute of Neuroscience and Mental Health, Australia h. Department of Neuroscience, Central Clinical School, Monash University, Australia

Published in Frontiers in Neurology on 2nd of July 2018

4.1. Abstract Introduction: Tremor of the upper limbs is a disabling symptom that is present during several neurological disorders and is currently without treatment. Functional MRI (fMRI) is an essential tool to investigate the pathophysiology of tremor and aid the development of treatment options. However, no adequately or standardised protocols for fMRI exists at present. Here we present a novel, online available fMRI task that could be used to assess the in vivo pathology of tremor. Objective: This study aims to validate the tremor-evoking potential of the fMRI task in a small group of tremor patients outside the scanner and assess the reproducibility of the fMRI task related activation in healthy controls. Methods: Twelve HCs were scanned at two time points (baseline and after 6-weeks). There were two runs of multi-band fMRI and the tasks included a ‘brick-breaker’ joystick game. The game consisted of three conditions designed to control for most of the activation related to performing the task by contrasting the conditions: WATCH (look at the game without moving joystick), MOVE (rhythmic left/right movement of joystick without game), and PLAY (playing the game). Task fMRI was analysed using FSL FEAT to determine clusters of activation during the different conditions. Maximum activation within the clusters was used to assess the ability to control for task related activation and reproducibility. Four tremor patients have been included to test ecological and construct validity of the joystick task by assessing tremor frequencies captured by the joystick. Results: In HCs the game activated areas corresponding to motor, attention and visual areas. Most areas of activation by our game showed moderate to good reproducibility (intraclass correlation coefficient (ICC) 0.531 to 0.906) with only inferior parietal lobe activation showing poor reproducibility (ICC 0.446). Furthermore, the joystick captured significantly more tremulous movement in tremor patients compared to HCs (p=0.01) during PLAY, but not during MOVE. Conclusion: Validation of our novel task confirmed tremor-evoking potential and reproducibility analyses yielded acceptable results to continue further investigations into the pathophysiology of tremor. The use of this technique in studies with tremor patient will no doubt provide significant insights into the treatment options.

4.2. Introduction Effective and efficient use of hands and arms is essential for undertaking daily activities (e.g. writing or consuming food and drink). Tremulous movements of the upper limb while 68

performing activities of daily living, specifically action tremors, are a symptom of a variety of neurological disorders, including Essential Tremor (ET) and multiple sclerosis (MS). These movements vary in severity but even when mild, can interfere with daily activities and be perceived by the patient as severely embarrassing (Lorenz, Schwieger, Moises, & Deuschl, 2006; Louis & Rios, 2009). Ameliorating tremor is therefore a priority for patients and clinicians treating these diverse diseases. While several treatments are in routine use, medical therapy often fails to provide adequate control (Schneider & Deuschl, 2015). The pathogenesis of tremors remains incompletely understood, further limiting the development of treatment options.

Action tremor encompasses tremors involve in movement, posture against gravity, and movement towards a target (intention tremor). These types of tremor arise from problems in parts of the brain that control and integrate movement. In ET and MS, most studies into the pathophysiology of tremor use surgical, post-mortem, neurophysiological, or animal studies. The use of magnetic resonance imaging (MRI) has been limited to structural brain imaging or basal network activity using resting-state functional MRI (fMRI). These studies have implicated the cerebellum, thalamus, red nucleus and cortex (Benninger, Thees, Kollias, Bassetti, & Waldvogel, 2009; Boonstra et al., 2017; Helmich, Janssen, Oyen, Bloem, & Toni, 2011; Jia, Jia-Lin, Qin, Qing, & Yan, 2011; Klein et al., 2011; Saini et al., 2012). Few studies have investigated the pathogenesis and relevant functional neuroanatomy of tremulous movement using fMRI.

Assessing in vivo brain activation via fMRI, especially during a movement/task, provides for a unique opportunity to better understand the pathophysiology of neurological upper limb function. However, the utility of fMRI in this field is under-explored. The most commonly used fMRI tasks to examine action tremor are the finger-tapping movement, finger-flexion and hand gripping task (Bucher, Seelos, Dodel, Reiser, & Oertel, 1997; Buijink et al., 2015; Buma, Raemaekers, Kwakkel, & Ramsey, 2015; Lewis et al., 2011; Lewis et al., 2007; Manganotti et al., 2010; Tessa et al., 2013). These simple tasks may not capture the full extent of the tremor that usually involve oscillations around several joints, including the shoulder joint, elbow, wrist and fingers, and the tasks are often not complex enough to represent typical, everyday manual movements during which the upper limb dysfunction is most noticeable. Furthermore, tasks are often not standardised, with variable study designs in the absence of proper control tasks to separate the activity related to performing the task from

activity related to the upper limb dysfunction within the patients (Sharifi, Nederveen, Booij, & van Rootselaar, 2014). Reproducibility is essential for the implementation of any proposed task in longitudinal studies, particularly when attempting to study the effect of treatment. It is important to study reproducibility to determine measurement error and sensitivity prior to application in patient groups, however, limited data is available about variability and reproducibility of tasks used to examine action tremor (McKinsey, Moritz, Meyerand, & Tome, 2010; Yoo et al., 2007).

We developed a novel, online available task that is dynamic enough to elicit a tremor independently of the upper limb area where the tremor is evident. In this study, we aimed to assess the feasibility of this task, potentially useful for studying the neural correlates of action tremor in the MRI scanner. To this end, we aimed to show proof of validity that our task evokes tremor in patients with action tremor. In addition, we examine the brain networks activated by a novel fMRI task and test the reproducibility of these activations in a group of healthy individuals. Developing a task that is able to reproducibly activate brain regions associated with upper limb tremor is an essential first step in exploring the brain regions involved in tremor and testing putative neural effects of treatments that ameliorate tremor.

4.3. Methods

Participants Twelve healthy control (HC) participants (45.4±15.3y, 75% female) underwent testing at baseline and 6-week follow-up that included performance of a joystick task both outside and inside the MRI scanner. In order to validate that the joystick task was able to evoke tremulous motion, we recruited four patients with multiple sclerosis and unilateral right-sided upper limb tremor (49.3±4.3y, 75% female, median Expanded Disability Status Scale (EDSS) 5.25, interquartile range (IQR) 3.75 (Kurtzke, 1983); and median Bain Tremor Rating 4.5, IQR 4 (Alusi, Worthington, Glickman, Findley, & Bain, 2000; Bain et al., 1993)). The patients performed the joystick task with the affected hand outside of the scanner. The study was approved by the local Human Research Ethics Committee and all participants provided voluntary written consent.

MRI acquisition Healthy controls underwent 3T MRI (MAGNETOM TrioTim, Siemens, Medical Systems, Erlangen, Germany) to acquire: (a) T1-weighted volumetric sequence (TR = 11.0ms, TE =

5.0ms, FOV = 1536*1536mm2, matrix = 208*256, slice thickness = 8.0mm, flip angle = 7.0°), and (b) two runs of gradient echo planar imaging fMRI (TR = 1.5ms, TE = 33ms, FOV = 204*204mm2, matrix = 104*104, slice thickness = 2mm, flip angle = 85.0°, volumes = 200, GRAPPA acceleration factor of 2, multi-band slice acceleration factor of 3). We used a MRI compatible joystick (FOJ-2B-10B, NAtA technologies) with 67Hz sampling frequency of x and y joystick positions (figure 1). To facilitate use of the joystick inside the MRI bore, the handle was shortened (Figure 1B). During imaging, the joystick was stabilised on the subject’s torso using a wedge-shaped foam pillow that allowed the subject to lie in a more comfortable position.

Figure 1. The joystick used outside of the scanner (A)and inside the scanner (B). The right figure shows the stick that we have shortened to facilitate use inside the scanner.

Joystick task The joystick task was purpose-written in C# (Visual Studio 2010, Microsoft Corporation, Washington, USA) and incorporates a game adapted from “Brick Ball Game” (Gajjar, 2014). The game timing was controlled by the MRI triggers. A compiled version is available for download from GitHub (Boonstra, 2017). The task is based on the retro arcade game Breakout (Atari, Inc.). The objective is to manipulate a paddle along the bottom of the screen to bounce a ball onto bricks. Bricks break and disappear from the screen when the ball collides with them. We modified this game in three ways: 1) sufficient rows of bricks were added to the top of the screen such that the player would never be able to break all of them within a single trial, 2) gameplay continues even after a player has failed to manoeuvre the 71

paddle to correctly bounce the ball, and 3) two balls (white and red) were introduced, with one being a distractor. Points 1 and 2 above allow the game to continue without pauses and remain consistent across trials.

The game was a block design with three conditions: PLAY (playing the game), WATCH (watching a simulated game without moving the joystick to control for visual attention related activity) and MOVE (moving the joystick left and right across the entire screen at a self- timed pace to control for basic manual motor activity). Each task block commenced with the display of an instruction screen indicating which condition was to follow for 5 seconds. Each condition was presented three times for 30 seconds, and the condition order was randomised.

This task should provide the complexity that allows a tremor to appear in patients with neurological tremor. Importantly, the participant’s arm movement was not restricted during our study and we did not explicitly instruct participants to move the joystick using wrist alone. Tremor originating in the forearm and upper-arm will mechanically couple through the wrist joint and be measurable with the joystick (i.e. the hand is the end-effector and its position is influenced by any movement of connected joints: elbow and shoulder). The task was performed during fMRI scanning by HC subjects, and the start of the game was triggered by a pulse sent from the MRI scanner. The game was performed under the same conditions outside the scanner by all subjects using the full-sized joystick shown in figure 1A.

Joystick task analysis During performance of the task, the position of on-screen objects, time timestamps for each MRI trigger and joystick movements were recorded. The joystick data were pass-band filtered (3 to 10Hz) to attenuate non-tremulous activity using custom Matlab scripts. The average power of the remaining frequency band (expressed in decibels, dB) was used to represent the amplitude of tremor-like movement. To validate that the task was able to elicit tremulous movements in patients with a diagnosed neurological tremor, we measured the amount of tremor-like movement in four patients with MS. As the patients performed the joystick task outside of the scanner, we compared this with the amount of tremor-like movement in HC when they performed the task outside the scanner. Differences in tremor-like movement were calculated using one-way ANOVA. Finally, we assessed whether there was a change in tremor-like movement over time in HCs.

Imaging analysis Functional MRI analyses were performed using FEAT v6.00 (FSL, FMRIB, Oxford, UK). Raw fMRI scans were pre-processed to correct for head motion, spatially smoothed (4mm extent threshold) and registered to the main structural image using boundary based registration and then to standard MNI space using FNIRT nonlinear registration.

Given that two conditions in the task involved hand movements that could induce head motion, we investigated the degree of head motion in all subjects. FSL motion outlier was used to detect slices within the data that was corrupted by large movements, with threshold set at the 75th percentile + 1.5 times the interquartile range. The confound matrix resulting from motion artefact detection was subsequently used in a general linear model (GLM) to remove the effects of these slices on the analysis.

Multi-level GLM analysis was used to identify regions of significant activation at baseline in HC subjects for the following contrasts: 1) PLAY > WATCH (motor component) 2) PLAY > MOVE (visual-attention component) 3) PLAY > WATCH + MOVE. Run-level environment variables included the time-course for each condition and associated temporal dispersion derivatives, and head translation and rotation parameters. Relative movement was calculated as the relative difference in head position and rotation volume to volume. Second-level analyses were used to combine the two joystick runs performed for each subject. Finally, third-level analyses assessed common areas of brain activation across all HC subjects.

Reproducibility To assess reproducibility, we compared the peak activation within clusters identified during baseline with data from the 6-week follow-up. From the third-level GLM analyses, we extracted the significant group level clusters for all three contrasts at baseline. These clusters were used to identify the maximum z-stat within each cluster for each HC at both baseline and at the 6-week scans. Reproducibility was quantified using two separate measures: intraclass correlation (ICC; based on a mean of measures, absolute agreement, 2-way mixed- effect model (Koo & Li, 2016)) for a measure of repeatability and the average coefficient of variation (CoV) to quantify the amount of variation useful for future treatment studies. For

the ICC, we compared the maximum z-stat for all the clusters for each HC between baseline and the 6-week time point. For the CoV we calculated the temporal standard deviation for the baseline and 6-week time points for each HC and divided this by the mean of both time points. The final CoV per cluster was the average of the CoV for all HCs.

4.4. Results

Head movement The relative motion and the number of slices removed for each subject are shown in supplementary table 1. At baseline, the median of relative movement was 0.105mm and the maximum numbers of slices removed was 7.5% (30 out of 400 slices). At 6 weeks, the average median of relative movement was 0.102mm and the maximum number of slices removed was 7% (28 out of 400 slices).

Validation of tremulous motion during task performance in tremor patients Average power in the tremor-like frequency band for joystick motion during MOVE was not significantly different in HC (10.8 dB) compared to tremor patients (14.9 dB) (F = 0.08, p = 0.787). However, during PLAY, the amount of tremor-like movement was significantly higher in tremor affected MS patients (36.2 dB) than for HC (27.9 dB) (F = 9.53, p = 0.010) (Figure 2). Finally, the amount of tremor-like movement was consistent over time in HC with no significant change at baseline versus 6-week follow-up during PLAY (t = 1.528, p = 0.155) and MOVE (t = 0.598 p = 0.5618).

Figure 2. The tremor-like movement during MOVE (top) PLAY (bottom) for both healthy controls and tremor patients. Each dot represents the average tremor-like movement per time point for the healthy controls (left) and tremor patients (right) during the PLAY condition.

Task analyses The statistical maps for PLAY > WATCH, PLAY > MOVE and PLAY > WATCH + MOVE are presented in figure 3 and table 1. At baseline, we found four significant (z-stat

>2.3) clusters during PLAY + WATCH. Specifically, activation cluster 1.1 and 1.2 were part of the contralateral and ipsilateral motor cortex respectively. Activation cluster 1.3 and 1.4 were part of the ipsilateral and contralateral cerebellum respectively.

Figure 3. Statistical parametric maps for the PLAY > WATCH (top), PLAY > MOVE (middle) and PLAY > WATCH + MOVE (bottom) contrasts, z-stat threshold of >2.3. Images are presented according to radiological convention (R = right, L = left).

During PLAY > MOVE we found four significant (z-score > 2.3) clusters. Specifically, the first activation cluster 2.1 was part of the visual attention network. Activation cluster 2.2 and 2.3 were part of the ipsilateral and contralateral frontal operculum cortex. Finally, activation cluster 2.4 was within the contralateral motor cortex.

Table 1. Significant clusters and their corresponding activation peaks for the PLAY > WATCH, PLAY > MOVE and PLAY > WATCH +MOVE contrasts x, y, z, the average MNI coordinates of the maximum activation peak for all HC for each cluster; ICC, intraclass correlation; CoV, coefficient of variation; R, right hemisphere; L, left hemisphere.

Cluster Voxels Anatomical region MNI coordinates ICC CoV # (mm) x y z PLAY > WATCH + MOVE 1 10964 Motor cortex L/R -11 -18 63 0.850 12.9% Superior parietal lobe 2 5740 Cerebellum L/R -22 -65 -9 0.888 11.3% 3 673 Visual cortex R 47 -67 4 0.689 25.6% 4 376 Inferior parietal R 63 -30 28 0.446 19.5% lobe PLAY > WATCH 1 2781 Motor cortex L -24 -14 78 0.823 19.2% 2 2393 Motor cortex R 46 -2 74 0.624 18.3% 3 968 Cerebellum R 28 -42 -10 0.881 21.0% 4 478 Cerebellum L -14 -40 -14 0.531 29.3% PLAY > MOVE 1 37689 Visual-attention L 50 8 70 0.906 8.6% network 2 626 Frontal operculum R 54 32 24 0.692 23.4% Cortex 3 498 Frontal operculum L -20 22 20 0.536 23.0% Cortex 4 317 Motor cortex L -44 12 42 0.743 20.6%

Finally, we found four significant (z-score > 2.3) clusters during the PLAY > WATCH + MOVE condition. Specifically, the first activation cluster 3.1 was part of the contralateral and ipsilateral motor cortex that includes the contralateral precentral and postcentral gyrus, and it extended towards the superior parietal lobe. The second activation cluster 3.2 includes

ipsilateral and contralateral cerebellum. The third activation cluster 3.3 was in the ipsilateral visual cortex. The fourth activation cluster 3.4 was within the inferior parietal lobe.

Reproducibility Reproducibility measures, ICC and CoV, between baseline and 6-week follow-up was calculated for all clusters and are shown in table 1. The ICC ranged from 0.437 to 0.888 and the CoV ranged from 9.2% to 28.2%.

4.5. Discussion This study aimed to validate a novel fMRI task potentially useful to study the brain activation correlates of upper limb tremor. Using a test condition (PLAY) and two control conditions (MOVE and WATCH) we were able to differentially identify regions involved in motor and visual attention aspects of the task. Motor regions included pre- and postcentral gyrus and the cerebellum. These regions of activation are similar to previous upper limb motor studies in healthy controls that indicate involvement of the premotor cortex, supplementary motor cortex, primary motor cortex and the cerebellum (Cowper-Smith, Lau, Helmick, Eskes, & Westwood, 2010; Dhamala et al., 2003). Visual regions included occipital cortex, occipital pole and lingual gyrus. These regions are known to play a role in visual processing (Bisley, 2011). Evidence suggests that tremors can originate after an insult or injury to the cerebello- thalamo-cortical tract (Lewis et al., 2011; Luo et al., 2017; Sen, Kawaguchi, Truong, Lewis, & Huang, 2010). Our task showed activation within areas of both ipsilateral and contralateral cerebellum that are known to be functionally linked to the somatomotor and premotor cortex (Buckner, Krienen, Castellanos, Diaz, & Yeo, 2011). However, no thalamic activations were detected indicating insignificant role of the thalamus during this task. Taken together, we believe that our work demonstrates an upper limb motor task that successfully activates the motor network, including areas hypothesised to be involved in action tremor pathophysiology.

We have shown that all brain areas activated by our task have moderate to good reliability, with the exception of the inferior parietal lobe (Portney & Watkins, 2000). Specifically, we found good reproducibility of the fMRI activation in the contralateral motor cortex and ipsilateral cerebellum, which indicates a strong activation within these brain areas. Our ICC within the motor cortex is consistent with ICC found by non-tremor focused upper-limb

imaging studies (Caceres, Hall, Zelaya, Williams, & Mehta, 2009; Kimberley, Khandekar, & Borich, 2008; Kristo et al., 2014). Furthermore, other studies that have used the CoV to assess the reproducibility of BOLD signals have also shown similar results (Gaxiola-Valdez & Goodyear, 2012; Tjandra et al., 2005); specifically, Tjandra et al. (2005) examined the CoV for the primary motor cortex and found a group mean of 24%. Reproducibility is essential to facility longitudinal use of the joystick task in future studies.

Importantly, patients with diagnosed upper-limb tremor displayed a significant increase in tremulous movements during the task compared to HCs, validating the ability of the task to elicit tremulous movements in future clinical applications. Tremor is one of the most common movement disorders and is a symptom of many neurological disorders including PD, ET and MS (Bötzel, Tronnier, & Gasser, 2014; WHO, 2013). In MS, tremor often presents as an action tremor with a postural and intention tremor component (Van der Walt et al., 2015). There have been no task fMRI studies in MS tremor patients and tasks used in fMRI studies of other neurological diseases causing upper limb dysfunction lack the complexity to elicit tremulous movement and ability to assess tremor in the whole upper limb. For example, finger-tapping has been used in PD tremor and ET but a tremor is not produced during the task, possibly due to the low muscle activation (Buijink et al., 2015). MS patients with tremor experienced more tremor-like movement while playing the game compared to controls, indicating our task successfully evokes tremulous movement. Examining the brain activity while patients are experiencing tremulous movement would be the most informative about the pathophysiology of intention tremors in neurological disorders, including MS patients.

Understanding the pathophysiology of PD and ET have led to improved and targeted treatments for these disorders such as deep brain stimulation (DBS) of the sub-thalamic area. DBS in this specific area disrupts pathological circuits in the cerebello-thalamo-cortical and pallido-thalamo-cortical pathways (Gallay, Jeanmonod, Liu, & Morel, 2008). The knowledge of this motor circuitry is currently based on animal and surgical studies (Gallay et al., 2008; Kobayashi et al., 2003; Magnin, Jeanmonod, Morel, & Siegemund, 2001; Obeso et al., 2000; Tasker, Lang, & Lozano, 1997). The value of fMRI is the non-invasive method to image motor circuitry and networks in vivo. To maximize its use, experiments need to be designed meticulously and validation and reproducibility assessments are essential.

Two limitations of our study are the low number of subjects and having only two different time points to assess reproducibility. Increasing the sample size of our study could provide more power to assess the reproducibility of the joystick task. By providing more time points we could better assess a learning effect over time. Further studies in larger cohorts and subsequent time points will allow a more thorough investigation of the reproducibility and ability to control for task related activation.

To conclude, our study introduced a novel task for fMRI studies that focuses on neurological upper limb dysfunction. Reproducibility shows acceptable results, and importantly our task is properly controlled and showed ecological validity. Future studies using this joystick task in patients with upper-limb tremors will further elucidate the pathophysiology of tremor.

4.6. Acknowledgement & funding This project was funded by an NHMRC Project Grant (1085461 CIA Van der Walt). F. Boonstra has nothing to declare. Bionics Institute receives Operational Infrastructure Support from the Victorian Government G. Noffs has nothing to declare. C. Marotta has nothing to declare. A/Prof Vogel receives funding from the National Health and Medical Research Council, Australia (Career Development Fellowship ID 1082910), received funding from the Alexander von Humboldt Foundation and receives institutional support from The University of Melbourne. A. Evans has received honoraria from Novartis for giving presentations and providing consultancy services. He has participated in scientific advisory board meetings for Novartis, UCB Pharma, Allergan, and Boehringer Ingelheim. He has received conference travel support from Boehringer Ingelheim. Prof Helmut Butzkueven is supported by an NHRMC Practitioner Fellowship 1080518 B. Moffat receives a fellowship from the National Institutes of Health. A/Prof Van der Walt is supported by an NHMRC Early Career Fellowship (1123330). S. Kolbe receives grant income from the National Health and Medical Research Council of Australia and has received honoraria from Novartis.

4.7. References Alusi, S., Worthington, J., Glickman, S., Findley, L., & Bain, P. (2000). Evaluation of three different ways of assessing tremor in multiple sclerosis. J Neurol Neurosurg Psychiatry, 68(6), 756-760. doi:10.1136/jnnp.68.6.756 Bain, P. G., Findley, L. J., Atchison, P., Behari, M., Vidailhet, M., Gresty, M., . . . Marsden, C. D. (1993). Assessing tremor severity. J Neurol Neurosurg Psychiatry, 56(8), 868-873. Benninger, D. H., Thees, S., Kollias, S. S., Bassetti, C. L., & Waldvogel, D. (2009). Morphological differences in Parkinson’s disease with and without rest tremor. Journal of Neurology, 256(2), 256-263. Bisley, J. W. (2011). The neural basis of visual attention. J Physiol, 589(Pt 1), 49-57. doi:10.1113/jphysiol.2010.192666 Boonstra, F. (2017). Joystick-fMRI. Boonstra, F., Florescu, G., Evans, A., Steward, C., Mitchell, P., Desmond, P., . . . van der Walt, A. (2017). Tremor in multiple sclerosis is associated with cerebello-thalamic pathology. J Neural Transm (Vienna), 124(12), 1509-1514. doi:10.1007/s00702-017- 1798-4 Bötzel, K., Tronnier, V., & Gasser, T. (2014). The Differential Diagnosis and Treatment of Tremor. Deutsches Ärzteblatt International, 111(13), 225-236. doi:10.3238/arztebl.2014.0225 Bucher, S. F., Seelos, K. C., Dodel, R. C., Reiser, M., & Oertel, W. H. (1997). Activation mapping in essential tremor with functional magnetic resonance imaging. Annals of Neurology, 41(1), 32-40. doi:10.1002/ana.410410108 Buckner, R. L., Krienen, F. M., Castellanos, A., Diaz, J. C., & Yeo, B. T. T. (2011). The organization of the human cerebellum estimated by intrinsic functional connectivity. Journal of Neurophysiology, 106(5), 2322-2345. doi:10.1152/jn.00339.2011 Buijink, A. W., Broersma, M., van der Stouwe, A. M., van Wingen, G. A., Groot, P. F., Speelman, J. D., . . . van Rootselaar, A. F. (2015). Rhythmic finger tapping reveals cerebellar dysfunction in essential tremor. Parkinsonism Relat Disord, 21(4), 383-388. doi:10.1016/j.parkreldis.2015.02.003 Buma, F. E., Raemaekers, M., Kwakkel, G., & Ramsey, N. F. (2015). Brain Function and Upper Limb Outcome in Stroke: A Cross-Sectional fMRI Study. PLoS One, 10(10), e0139746. doi:10.1371/journal.pone.0139746

Caceres, A., Hall, D. L., Zelaya, F. O., Williams, S. C., & Mehta, M. A. (2009). Measuring fMRI reliability with the intra-class correlation coefficient. NeuroImage, 45(3), 758- 768. doi:10.1016/j.neuroimage.2008.12.035 Cowper-Smith, C. D., Lau, E. Y. Y., Helmick, C. A., Eskes, G. A., & Westwood, D. A. (2010). Neural Coding of Movement Direction in the Healthy Human Brain. PLoS One, 5(10), e13330. doi:10.1371/journal.pone.0013330 Dhamala, M., Pagnoni, G., Wiesenfeld, K., Zink, C. F., Martin, M., & Berns, G. S. (2003). Neural correlates of the complexity of rhythmic finger tapping. NeuroImage, 20(2), 918-926. doi:https://doi.org/10.1016/S1053-8119(03)00304-5 Gajjar, A. (2014). Brick Ball Game: WPF# Game. Retrieved from https://www.codeproject.com/Articles/470039/Brick-Ball-Game-WPF-Csharp- Game Gallay, M. N., Jeanmonod, D., Liu, J., & Morel, A. (2008). Human pallidothalamic and cerebellothalamic tracts: anatomical basis for functional stereotactic neurosurgery. Brain Structure & Function, 212(6), 443-463. doi:10.1007/s00429-007-0170-0 Gaxiola-Valdez, I., & Goodyear, B. G. (2012). Origins of intersubject variability of blood oxygenation level dependent and arterial spin labeling fMRI: implications for quantification of brain activity. Magn Reson Imaging, 30(10), 1394-1400. doi:http://dx.doi.org/10.1016/j.mri.2012.05.002 Helmich, R. C., Janssen, M. J. R., Oyen, W. J. G., Bloem, B. R., & Toni, I. (2011). Pallidal dysfunction drives a cerebellothalamic circuit into Parkinson tremor. Annals of Neurology, 69(2), 269-281. doi:10.1002/ana.22361 Jia, L., Jia-Lin, S., Qin, D., Qing, L., & Yan, Z. (2011). A diffusion tensor imaging study in essential tremor. J Neuroimaging, 21(4), 370-374. doi:10.1111/j.1552- 6569.2010.00535.x Kimberley, T. J., Khandekar, G., & Borich, M. (2008). fMRI reliability in subjects with stroke. Experimental Brain Research, 186(1), 183-190. doi:10.1007/s00221-007-1221-8 Klein, J. C., Lorenz, B., Kang, J. S., Baudrexel, S., Seifried, C., van de Loo, S., . . . Hilker, R. (2011). Diffusion tensor imaging of white matter involvement in essential tremor. Hum Brain Mapp, 32(6), 896-904. doi:10.1002/hbm.21077 Kobayashi, K., Katayama, Y., Kasai, M., Oshima, H., Fukaya, C., & Yamamoto, T. (2003). Localization of thalamic cells with tremor-frequency activity in Parkinson's disease and essential tremor. Acta Neurochir Suppl, 87, 137-139.

Koo, T. K., & Li, M. Y. (2016). A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research. Journal of Chiropractic Medicine, 15(2), 155-163. doi:10.1016/j.jcm.2016.02.012 Kristo, G., Rutten, G. J., Raemaekers, M., de Gelder, B., Rombouts, S. A., & Ramsey, N. F. (2014). Task and task-free FMRI reproducibility comparison for motor network identification. Hum Brain Mapp, 35(1), 340-352. doi:10.1002/hbm.22180 Kurtzke, J. F. (1983). Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology, 33(11), 1444-1452. Lewis, M. M., Du, G., Sen, S., Kawaguchi, A., Truong, Y., Lee, S., . . . Huang, X. (2011). Differential involvement of striato- and cerebello-thalamo-cortical pathways in tremor-and akinetic/rigid-predominant Parkinson's disease. Neuroscience, 177, 230- 239. doi:10.1016/j.neuroscience.2010.12.060 Lewis, M. M., Slagle, C. G., Smith, A. B., Truong, Y., Bai, P., McKeown, M. J., . . . Huang, X. (2007). Task specific influences of Parkinson's disease on the striato- thalamo-cortical and cerebello-thalamo-cortical motor circuitries. Neuroscience, 147(1), 224-235. doi:10.1016/j.neuroscience.2007.04.006 Lorenz, D., Schwieger, D., Moises, H., & Deuschl, G. (2006). Quality of life and personality in essential tremor patients. Mov Disord, 21(8), 1114-1118. doi:10.1002/mds.20884 Louis, E. D., & Rios, E. (2009). Embarrassment in essential tremor: prevalence, clinical correlates and therapeutic implications. Parkinsonism Relat Disord, 15(7), 535-538. doi:10.1016/j.parkreldis.2008.10.006 Luo, C., Song, W., Chen, Q., Yang, J., Gong, Q., & Shang, H. F. (2017). White matter microstructure damage in tremor-dominant Parkinson's disease patients. Neuroradiology, 59(7), 691-698. doi:10.1007/s00234-017-1846-7 Magnin, M., Jeanmonod, D., Morel, A., & Siegemund, M. (2001). Surgical control of the human thalamocortical dysrhythmia. Thalamus & Related Systems, 1(1), 81-89. doi:http://dx.doi.org/10.1016/S1472-9288(01)00002-4 Manganotti, P., Acler, M., Formaggio, E., Avesani, M., Milanese, F., Baraldo, A., . . . Fiaschi, A. (2010). Changes in cerebral activity after decreased upper-limb hypertonus: an EMG-fMRI study. Magn Reson Imaging, 28(5), 646-652. doi:10.1016/j.mri.2009.12.023

McKinsey, R. D., Moritz, C. H., Meyerand, M. E., & Tome, W. A. (2010). Assessment of multiple task activation and reproducibility in patients with benign and low-grade neoplasm. Technol Cancer Res Treat, 9(4), 319-326. doi:10.1177/153303461000900402 Obeso, J. A., Rodriguez-Oroz, M. C., Rodriguez, M., Lanciego, J. L., Artieda, J., Gonzalo, N., & Olanow, C. W. (2000). Pathophysiology of the basal ganglia in Parkinson's disease. Trends in Neurosciences, 23, S8-S19. doi:http://dx.doi.org/10.1016/S1471- 1931(00)00028-8 Portney, L. G., & Watkins, M. P. (2000). Foundations of clinical research: application to practice. New Jersey: Prentice Hall. Saini, J., Bagepally, B. S., Bhatt, M. D., Chandran, V., Bharath, R. D., Prasad, C., . . . Pal, P. K. (2012). Diffusion tensor imaging: tract based spatial statistics study in essential tremor. Parkinsonism Relat Disord, 18(5), 477-482. doi:10.1016/j.parkreldis.2012.01.006 Schneider, S. A., & Deuschl, G. (2015). Medical and surgical treatment of tremors. Neurol Clin, 33(1), 57-75. doi:10.1016/j.ncl.2014.09.005 Sen, S., Kawaguchi, A., Truong, Y., Lewis, M. M., & Huang, X. (2010). Dynamic changes in cerebello-thalamo-cortical motor circuitry during progression of Parkinson's disease. Neuroscience, 166(2), 712-719. doi:10.1016/j.neuroscience.2009.12.036 Sharifi, S., Nederveen, A. J., Booij, J., & van Rootselaar, A.-F. (2014). Neuroimaging essentials in essential tremor: A systematic review. NeuroImage : Clinical, 5, 217-231. doi:10.1016/j.nicl.2014.05.003 Tasker, R. R., Lang, A. E., & Lozano, A. M. (1997). Pallidal and Thalamic Surgery for Parkinson's Disease. Experimental Neurology, 144(1), 35-40. doi:http://dx.doi.org/10.1006/exnr.1996.6385 Tessa, C., Diciotti, S., Lucetti, C., Baldacci, F., Cecchi, P., Giannelli, M., . . . Mascalchi, M. (2013). fMRI changes in cortical activation during task performance with the unaffected hand partially reverse after ropinirole treatment in de novo Parkinson's disease. Parkinsonism Relat Disord, 19(2), 265-268. doi:http://dx.doi.org/10.1016/j.parkreldis.2012.07.018 Tjandra, T., Brooks, J. C. W., Figueiredo, P., Wise, R., Matthews, P. M., & Tracey, I. (2005). Quantitative assessment of the reproducibility of functional activation measured with BOLD and MR perfusion imaging: Implications for clinical trial design. NeuroImage, 27(2), 393-401. doi:http://dx.doi.org/10.1016/j.neuroimage.2005.04.021

Van der Walt, A., Buzzard, K., Sung, S., Spelman, T., Kolbe, S. C., Marriott, M., . . . Evans, A. (2015). The occurrence of dystonia in upper-limb multiple sclerosis tremor. Mult Scler, 21(14), 1847-1855. doi:10.1177/1352458515577690 WHO. (2013). Atlas of MS 2013: Mapping Multiple Sclerosis Around the World. Retrieved from http://www.msif.org/about-ms/publications-and-resources/ Yoo, S.-S., O’Leary, H. M., Lee, J.-H., Chen, N.-K., Panych, L. P., & Jolesz, F. A. (2007). Reproducibility of trial-based functional MRI on motor imagery. International Journal of Neuroscience, 117(2), 215-227. doi:10.1080/00207450600582546

4.8. Supplementary table

Supplementary table 1. For each subject, the relative motion and number of slices removed due to excessive motion from the 400 slices dataset.

Baseline 6 weeks Relative Removed Relative Removed motion slices motion slices Subject 1 Median 0.052 11 0.056 12 Minimum 0.007 0.012 Maximum 0.415 0.312 Subject 2 Median 0.090 4 0.086 6 Minimum 0.006 0.013 Maximum 0.332 0.511 Subject 3 Median 0.111 18 0.121 7 Minimum 0.023 0.020 Maximum 1.157 1.077 Subject 4 Median 0.154 21 0.128 16 Minimum 0.014 0.013 Maximum 1.285 0.875 Subject 5 Median 0.055 10 0.062 20 Minimum 0.006 0.007 Maximum 0.557 2.350 Subject 6 Median 0.070 18 0.053 9

Minimum 0.012 0.001 Maximum 0.231 0.181 Subject 7 Median 0.152 19 0.175 22 Minimum 0.009 0.031 Maximum 0.858 0.719 Subject 8 Median 0.077 19 0.088 18 Minimum 0.012 0.015 Maximum 0.532 0.393 Subject 9 Median 0.063 19 0.066 5 Minimum 0.014 0.011 Maximum 0.166 0.500 Subject Median 0.107 20 0.124 28 10 Minimum 0.014 0.011 Maximum 0.741 0.810 Subject Median 0.169 18 0.141 19 11 Minimum 0.023 0.021 Maximum 0.674 1.001 Subject Median 0.160 30 0.121 23 12 Minimum 0.033 0.020 Maximum 1.194 0.609

Pathophysiology Study

Functional neuroplasticity in response to cerebello-thalamic injury underpins the clinical presentation of tremor in multiple sclerosis Frederique M. C. Boonstra, MSc 1, Gustavo Noffs, MD 2,3, Thushara Perera, PhD 4,5, Vilija G. Jokubaitis, PhD 6, Adam P. Vogel, PhD 3,4,7,8, Bradford A. Moffat, PhD 1, Helmut Butzkueven, PhD 6, Andrew Evans, MD 2,4, Anneke van der Walt, PhD 2,4,6,10 and Scott C. Kolbe, PhD 1,9,10 1. Department of Medicine and Radiology, University of Melbourne 2. Department of Neurology, Royal Melbourne Hospital, Australia 3. Centre for Neuroscience of Speech, University of Melbourne, Victoria, Australia 4. The Bionics Institute, Australia 5. Department of Medical Bionics, University of Melbourne, Australia 6. Department of Neuroscience, Central Clinical School, Monash University, Australia 7. Department of Neurodegeneration, Hertie Institute for Clinical Brain Research, University of Tübingen, Germany 8. Redenlab, Victoria, Australia 9. Florey Institute of Neuroscience and Mental Health, Australia 10. Authors contributed equally published in Multiple Sclerosis Journal on … January 2019

5.1. Abstract Background: Tremor is present in almost half of multiple sclerosis (MS) patients. The lack of understanding of its pathophysiology is hampering progress in development of treatments. Objectives: To clarify the structural and functional brain changes associated with the clinical phenotype of upper limb tremor in people with MS. Methods: Fifteen healthy controls (46.1±15.4y), 27 MS participants without tremor (46.7±11.6y) and 42 with tremor (46.6±11.5y) were included. Tremor was quantified using the Bain score (0 to 10) for overall severity, handwriting and Archimedes spiral drawing. Functional MRI activations were compared between participants groups during performance of a joystick task designed to isolate tremulous movement. Inflammation and atrophy of cerebello-thalamo-cortical brain structures were quantified. Results: Tremor participants were found to have atrophy of the cerebellum and thalamus, and higher ipsilateral cerebellar lesion load compared to participants without tremor (p<0.020). We found higher ipsilateral activation in the inferior parietal lobule, the premotor cortex, and supplementary motor area in MS tremor participants compared to MS participants without tremor during the joystick task. Finally, stronger activation in those areas was associated with lower tremor severity. Conclusions: Subcortical neurodegeneration and inflammation along the cerebello- thalamo-cortical and cortical functional neuroplasticity contribute to the severity of tremor in MS.

5.2. Introduction

Multiple sclerosis (MS) is the leading non-traumatic neurological disorder in young adults, with 2.3 million people worldwide affected (Browne et al., 2014). Tremor is reported in approximately 45% of people with MS, primarily affecting the upper limbs (Rinker et al., 2015). It is defined as an “involuntary, rhythmic, oscillatory movement of a body part” (Deuschl, Bain, & Brin, 1998). MS tremor often manifests during: 1) postures held against gravity (postural tremor); 2) voluntary movement (kinetic tremor); and 3) visually-guided target-oriented movement (intention tremor) (Deuschl et al., 1998). The presence of tremor significantly impacts patients’ quality of life, with more than half of patients with mild tremor being unemployed (Julian, Vella, Vollmer, Hadjimichael, & Mohr, 2008; Koziarska et al., 2018; Rinker et al., 2015). MS tremor can occur at any stage of the disease but is most

often observed, and is most severe, in people with secondary-progressive MS (Pittock, McClelland, Mayr, Rodriguez, & Matsumoto, 2004; Rinker et al., 2015). Furthermore, there is currently no effective treatment available to help alleviate the severity of tremor in MS (Labiano-Fontcuberta & Benito-León, 2012).

There is limited understanding of the anatomical correlates of tremor in MS. Two case studies have reported MS patient with tremor had a lesion within the superior cerebellar peduncle (SCP), and thalamus and red nucleus respectively (Fahn, 1986; Nakamura et al., 1993). A later study by Feys et al. (Feys et al., 2005) demonstrated a link between tremor severity in MS patients and T2 lesion load in the contralateral pons. Finally, in a focussed study by our group we confirmed damage to the SCP, thalamus is associated with tremor severity (F. Boonstra et al., 2017). Taken together, these studies suggest that cerebello- thalamo-cortical (CTC) tract pathology is an important contributor to tremor in MS. However, these studies were low in sample size and limited to structural MRI. In contrast to structural MRI, functional MRI (fMRI) offers the unique ability to assess brain activation in vivo during tremulous movement and can therefore provide direct insights into tremor pathophysiology. To date, there have been no functional imaging studies of tremor in people with MS. However, fMRI studies of general hand function in people with MS have reported that increased activation of ipsilateral activation was associated with compensatory effects on hand function (Reddy et al., 2000; Reddy et al., 2002).

This study aimed to examine the functional correlates of tremor in MS. Specifically, the relative contributions of structural and functional brain changes to the presence and severity of tremor in patients with MS. We hypothesised that tremor is associated with pathology within the CTC tract based on our previous pilot results (F. Boonstra et al., 2017). Furthermore, we also hypothesised that functional reorganisation of sensorimotor brain networks might play a role in ameliorating the severity of tremor.

5.3. Methods

Study design This study is the part of a registered clinical trial studying the efficacy of onabotulinumtoxinA for MS-related tremor (ACTRN12617000379314).

Participants Sixty-nine participants with MS were included of which 42 participants presented with unilateral upper limb tremor and 27 presented without tremor. Detailed medical and tremor history were obtained for all participants. Inclusion criterium was normal or near- normal upper limb strength (Medical Research Council (MRC) score >4+/5). A qualified neurologist determined the Expanded Disability Status scale scores (EDSS) (Kurtzke, 1983) for all participants. Tremor components were defined according to the Consensus Statement of the Movement Disorder Society, 1998 (Deuschl et al., 1998) and assessed according to our previously described protocol (Van Der Walt et al., 2012). The Bain score (Bain et al., 1993) was used to rate tremor severity observed while writing a standardised sentence (“This is a sample of my best handwriting”) and drawing an Archimedes spiral on a pre-drawn pattern with the dominant hand (Alusi, Worthington, Glickman, & Bain, 2001). Finally, 15 healthy controls were recruited for this study and assessed using the same protocol.

Standard protocol approvals, registrations, and patient consents The study was approved by the Melbourne Health Human Research Ethics Committee. All of the participants gave informed voluntary written consent.

Imaging protocols Participants were imaged using a 3T MRI system (Trio, Siemens, Erlangen). Structural imaging included (1) high-resolution three-dimensional (3D) T1-weighted magnetization- prepared rapid acquisition with gradient echo (MPRAGE) scan with online motion correction (recovery time (TR) = 2530ms; echo time (TE) = 2.5ms; inversion time (TI) = 1260ms; field of view (FOV) = 176x256 mm; voxel size = 1.0x1.0x1.0 mm) and (2) high- resolution 3D T2-weighted double inversion recovery (DIR) sequence (TR = 7400ms; TE = 324ms; TI = 3000ms; flip angle = 120°; echo train length (ETL) = 625; FOV = 144x220 mm; voxel size = 1.0x1.0x1.0). Functional imaging included two runs (TR = 1.5ms, TE = 33ms, FOV = 2040x2040, matrix = 104x104, slice thickness = 2.5, flip angle = 85.0, multi- band slice acceleration factor = 3, volumes = 200). For, healthy controls only the T1- weighted MPRAGE scan was performed.

Joystick task The fMRI task performed is described in detail in Boonstra et al. (2018). Briefly, the task involved participating in a simple game using an MRI-compatible joystick. There were three conditions in the game (PLAY, WATCH and MOVE) and each condition started with an instruction screen indicating which condition would follow. During PLAY the participant manoeuvred a paddle under a ball to bounce it up. During WATCH the participant only watched a ball move along the screen. During MOVE the participant moved the joystick to the left and right of the screen continuously, while the ball was frozen on the screen. The PLAY condition includes visual tracking, attention and basic motor activity. Furthermore, this condition has shown to elicit tremulous movement in tremor participants compared to healthy controls (F. M. C. Boonstra et al., 2018). The WATCH condition required no motor activity but controlled for visual aspects and visual attention components of PLAY. In addition, the WATCH condition controls visual aspects. The MOVE condition required no visual attention but controlled for the baseline movement activity component of PLAY.

Joystick task analysis Analyses of the tremor-like movement measured using the joystick are described in Boonstra et al. (F. M. C. Boonstra et al., 2018). Briefly, we measured the average power of frequencies between 3-10Hz, expressed in decibels (dB), to represent the amplitude of tremor-like movement during the different conditions.

Functional imaging analyses Task fMRI was analysed using FSL FEAT 5.0.8 (Smith et al., 2004). The data were y- flipped for all participants with a left-handed tremor, resulting in images in which the hemisphere contralateral to the tremor-affected side was on the same side for each participant (Buma, Raemaekers, Kwakkel, & Ramsey, 2015; Lotze, Flor, Grodd, Larbig, & Birbaumer, 2001). Only the prefrontal cortex showed lateralization (see Supplementary Figure 1). However, we found no significant difference between the flipped and the un- flipped functional images. Possible bias was further limited by were near equal numbers of participants flipped in both the MS participants group with tremor and without tremor.

To create the best template for our group of MS participants, we created a study-specific template using Advanced Neuroimaging Tools (ANTs) (Avants et al., 2011). A nonlinear

registration was performed within FEAT to the main structural image and the ANTs template. Pre-processing of fMRI images included motion correction FMRIB Linear Registration Tool (MCFLIRT), high pass filtering (<0.01Hz), and spatial smoothing (4mm). FSL motion outlier detection was used to detect slices within the data that were corrupted by large movements defined as movement above the 75th percentile + 1.5 times the interquartile range. Finally, we added standard motion parameters. Run-level general linear models for the following contrasts were performed: PLAY > WATCH (motor component) PLAY > MOVE (visual attention component) PLAY > WATCH + MOVE (tremor component) For each contrast, second level analyses were used to combine the two task runs for each participant. Subsequently, third level analyses were used to compare each contrast between participants with MS with and without tremor. For PLAY > WATCH + MOVE we compared the main activation between participants with MS with tremor and without tremor to identify the areas related to tremor, while controlling for visual and motor components. The cluster(s) that showed a significant difference between participants with MS with and without tremor were selected (z-stat > 2.3, cluster size threshold p = 0.05). We measured the intensity of the voxel with the maximum activation within the cluster(s) for each subject, as the maximum is not biased towards the size of the cluster

Structural imaging analyses Structural MRI was analysed using FreeSurfer version 5.3 (http://surfer.nmr.mgh.harvard.edu/), Osirix Lite v9.0.2, and Lesion Segmentation Toolbox (LST) version 2.0.15 for Statistical Parametric Mapping (SPM) (Schmidt et al., 2012). Based on previous literature, we selected the thalamus, superior cerebellar peduncle (SCP) and cerebellum as structural region of interests (ROIs). In addition, cortical thickness was measured to examine its effect on any functional changes. FreeSurfer was used to determine the volumes of the thalamus and cerebellum and average cerebral cortical thickness from the MPRAGE images. The area of the SCP was manually labelled as described previously in Boonstra et al. (2017). Lesion load in MS participants within the cerebellum and the thalamus was derived from the DIR images using the lesion prediction algorithm within LST. Due to the size of the SCP, we did not separately calculated lesion load for this structure. To optimize the automatic lesions segmentation, we manually determined the optimal threshold for each participant and then corrected the final lesion

maps. See Supplementary Figure 2 for an example of the lesion segmentation. Lesion load within the thalamus and cerebellum was determined using the brain segmentation maps derived from the FreeSurfer analyses described above.

Statistical analyses Statistical analyses were performed in SPSS version 24. To test for group differences in clinical measures (Bain and EDSS scores), we used non-parametric Mann-Whitney tests. For structural measures, corrected for intracranial volume (ICV) by division, group differences were measured using a one-way ANOVA with Turkey post-hoc test. Group difference in lesion load, corrected for intracranial volume, was assessed using an independent t-test. For the MS participants, tremor-like movement was compared during the MOVE and PLAY conditions using two-way repeated measures ANOVA (within- subject factor: condition; between-subject factor: group). To examine the relationship between fMRI activation level and tremor severity within the participants with tremor, we used a linear regression with the Bain scores as predictors and the maximum activation within the cluster(s) and age as covariates. Finally, to determine the effect of overall disease severity on tremor-associated pathology, we used a Spearman’s rank correlation between all imaging parameters and general disease severity (EDSS) in tremor participants, while correcting for age and disease duration using regression. Correction of multiple comparisons using the false discovery rate was performed.

5.4. Results

Demographic and clinical data See table 1. Healthy controls had a mean age of 46.1, 60% was female and 80% performed the joystick task with the right hand. The participants without tremor had a mean age of 46.7, mean disease duration of 11.4 years, 55.4% had secondary progressive MS, 73.8% was female, 71.4% performed the joystick task with their right hand and a median EDSS of 3. The MS participants with tremor had a mean age of 46.6, mean disease duration of 14 years, 61.3% had secondary progressive MS, 77.8% was female, 74.1% had a right-sided tremor, mean tremor duration was 8.5 years and a median EDSS of 4. Furthermore, one participant with tremor presented with primary progressive MS. There were no significant differences between age, disease duration, disease course and EDSS. EDSS visual and sensory scores were mild and not different between the participants groups. The clinical

tremor measures, Bain severity, handwriting and Archimedes were significantly higher in the participants with tremor compared to without tremor (p<0.001).

Table 1. Demographic and clinical tremor data for multiple sclerosis participants with tremor and without tremor. * False discovery rate corrected p < 0.05. Abbreviations: MS, multiple sclerosis; sd, standard deviation; y, years; EDSS, expanded disability status scale; IQR, interquartile range. Healthy MS MS with p-value controls Without tremor tremor N 15 27 42 - Age mean(sd) 46.1 (15.4) 46.7 (11.6) 46.6 (11.5) 0.910 Disease duration, y mean(sd) - 11.4 (7.7) 14.0 (7.4) 0.168 Tremor duration, y mean(sd) - - 8.5 (8.0) - Affected hand right % 80 71.4 74.1 0.816 Gender female % 60 73.8 77.8 0.461 EDSS median (IQR) - 3 (2.5,5) 4 (2.75,6) 0.062 EDSS visual median (IQR) - 0 (0,1) 0 (0,1) 0.763 EDSS sensory median (IQR) - 1 (0,2) 1 (0,2) 0.794 Disease course, secondary 55.6% 61.9% 0.388 progressive % Bain severity median (IQR) - 0 (0,0) 3 (2,5) <0.001* Bain handwriting median (IQR) - 0 (0,0) 3 (2,5.5) <0.001* Bain Archimedes median (IQR) - 0 (0,0) 3 (2,4.5) <0.001*

Structural group differences There were significant structural differences between healthy controls, MS participants with and without tremor (Table 2 and Supplementary Figure 3). After correction for multiple comparisons, post hoc analyses revealed that for both the contralateral and ipsilateral cerebellum, MS participants with tremor had a smaller volume compared to healthy controls (p=0.002 and p=0.001) and MS participants without tremor (p<0.001 and p=0.001). Furthermore, the lesion load in the ipsilateral cerebellum was significantly higher (p=0.018) in MS participants with tremor compared to without tremor. Tremor participants had a smaller contralateral and ipsilateral thalamic volume than healthy

controls (p<0.001 and p=0.003) and MS participants without tremor (p=0.013 and p=0.014). No significant differences between tremor participants compared to healthy controls and MS participants without tremor were found for the volume of the SCP.

Table 2. Volumetric and lesion load data for multiple sclerosis participants with tremor and without tremor. * False daiscovery rate corrected p < 0.05. Abbreviations: MS, multiple sclerosis; ICV, intracranial volume; sd, standard deviation; SCP, superior cerebellar peduncle. Healthy MS without MS with p-value controls tremor tremor ICV mean(sd) 1.07*106 1.07*106 1.27*106 0.010* (2.49*105) (2.59*105) (3.31*105) Cortical thickness mm2 2.35 (0.07) 2.32 (0.12) 2.36 (0.15) 0.496 mean(sd) Cerebellar volume mm3 mean(sd) Contralateral 61070 (12536) 58810 (8950) 55065 (15886) <0.001* Ipsilateral 62589 (11585) 59219 (9823) 55377 (15698) <0.001* Thalamic volume mm3 mean(sd) Contralateral 7491 (1139) 6432 (796) 6294 (897) <0.001* Ipsilateral 6920 (649) 6311 (926) 6129 (889) 0.001* SCP area mm2 mean(sd) Contralateral 26.9 (4.65) 24.4 (5.08) 24.9 (4.56) 0.017 Ipsilateral 27.1 (4.33) 25.3 (5.19) 25.7 (4.91) 0.053 Total lesions mL(sd) - 14583 (18063) 16946 (12294) 0.901 Cerebellar lesions mL(sd) Contralateral - 4.4 (13.1) 40.1 (153) 0.257 Ipsilateral - 1.9 (5.4) 25.4 (67.8) 0.018* Thalamic lesions mL(sd) Contralateral - 10.4 (22.4) 21.7 (85.7) 0.613

Ipsilateral - 13.2 (24.3) 8.5 (22.7) 0.187

Tremulous movement during task MS participants with tremor showed an increase in tremor-like movement during the PLAY and MOVE conditions compared to participants without tremor (p<0.001). Furthermore, the amount of tremor-like movement within both groups was significantly higher in PLAY compared to MOVE (p<0.001; Figure. 1).

Figure 1. Tremor during fMRI task. The amount of tremor-like movement in decibel (dB) during the PLAY (left) and MOVE (right) condition for multiple sclerosis participants with tremor and without tremor. Points are overlayed a 95% confidence interval in dark grey and a 1 SD in light grey. * indicate significant difference p<0.001.

Functional group differences Five tremor participants were unable to complete the fMRI experiment due to practical limitations and were excluded from fMRI analyses. Specifically, we were unable to fit the

joystick inside the bore for two participants, and experienced software difficulties for the remaining three. The main effect for MS participants with and without tremor using the three contrasts is displayed in Figure. 2. The main effect using the contrast PLAY > WATCH showed activation of the motor cortex (Figure. 2A), PLAY > MOVE showed the visual and attention areas (Figure. 2B) and PLAY > WATCH + MOVE showed the attention areas remaining after subtracting the motor and visual component (Figure. 2C). Compared to patients without tremor, tremor patients showed greater activation within the ipsilateral inferior parietal lobule (Figure. 2D, cluster 1: 451 voxels; maximum/mean z-stat 2.99/1.30; Montreal Neurological Institute (MNI) coordinates of voxel with maximum z- stat 56, -56, 26), and the premotor/supplementary motor area (SMA) (Figure. 2E, cluster 2: 691 voxels; maximum/mean z-stat 3.10/1.12; MNI coordinates of voxel with maximum z- stat 36, 16, 52) when looking at PLAY > WATCH + MOVE.

Figure 2. Activation during fMRI task. A-C: Main effect for the three contrasts. Multiple sclerosis participants with tremor (red) and without tremor (blue), overlap is depicted in purple. D-E: The areas with higher activation in multiple sclerosis participants with tremor than without tremor. The areas are separated into cluster 1 (D) and cluster 2 (E). Abbreviations: ipsi, ipsilateral to the tremor-affected hand; contra, contralateral to the tremor affected hand.

Functional task activations and tremor severity We identified some associations of modest significance. Specifically, one unit increase in cluster 1 activation was associated with a drop of 0.40 in Bain severity (p=0.022), 0.59 in Bain handwriting: (p=0.018) and 0.43 in Bain Archimedes (p=0.052). Furthermore, one unit increase in cluster 2 activation was associated with a drop of 0.24 in Bain severity (p=0.065), 0.42 in Bain handwriting (p=0.018) and 0.29 in Bain Archimedes (p=0.069). Multiple comparisons corrected p=0.043.

MRI and general disability in patients with tremor In order to determine whether MRI variation was related to general disease progression rather than tremor-specific pathology, we determined the relationship between EDSS and the MRI parameters. No correlations between any of the structural MRI measures and EDSS were found. There was a moderate correlation between EDSS and the degree of activation within the ipsilateral SMA and premotor area (ρ=-0.379, p=0.028). However, this correlation did not remain significant after correction for multiple comparisons.

5.5. Discussion We examined the structural and functional brain changes related to the presence and severity of tremor in patients with MS. Two distinct features in the brains of patients with MS were found to be associated with the presence of moderate tremor: subcortical structural damage and cortical functional neuroplasticity. Subcortical structural damage presented as atrophy of the cerebellum and thalamus, and increased lesion load within the ipsilateral cerebellum. Functional neuroplasticity was observed as increased activation within the ipsilateral cortex; specifically, the premotor and supplementary motor area (SMA) and the inferior parietal lobule (IPL).

Neurophysiological models of MS tremor have emphasized the effect of demyelinating lesions on cortico-cerebello-cortical loops by which motor commands are projected from the motor cortex via the cortico-ponto-cerebellar pathway to the cerebellum, and return via the cerebello-thalamo-cortical pathway back to the motor cortex (Liu et al., 1997). Specifically, it has been proposed that MS tremor patients are more dependent on visual information and that their internal representation of the state of the limb (position and velocity), thought to be encoded by an internal model in the cerebellum (Miall, Weir, Wolpert, & Stein, 1993), has been impaired (Liu et al., 1997). Further evidence is provided by animal studies and surgical treatments that emphasize the SCP, thalamus, and basal ganglia in the pathophysiology of tremor in patients with MS (Carpenter & Hanna, 1962; Foote et al., 2006; Wilms, Sievers, & Deuschl, 1999; Yap, Kouyialis, & Varma, 2007). Furthermore, small cross-sectional structural imaging studies have indicated pathology of the SCP, pons, brainstem, internal capsule, thalamus, and red nucleus are associated with tremor pathogenesis (Anderson, Fox, & Miller, 2006; F. Boonstra et al., 2017; Feys et al., 2005; Nakamura et al., 1993). Our results support the involvement of neurodegeneration

and inflammatory damage along the cerebello-thalamic tract in the pathophysiology of tremor in patients with MS. We observed bilateral atrophy in these structures in patients with tremor compared to patients without tremor. Importantly, logistic regression showed that ipsilateral cerebellar atrophy and contralateral thalamic atrophy were significant, independent predictors of tremor status. This is consistent with the anatomy of the cerebello-thalamic tract that decussates between these structures. Additional inflammatory lesion load in the contralateral cerebellar hemispheres was found. We observed a correlation between tremor and bilateral cerebellar and thalamic atrophy. This finding is inconsistent with the notion that unilateral tremor is attributable to damage of the ipsilateral cerebellum and contralateral thalamus - consistent with the known anatomy of the cerebellar-thalamo-cortical network. One study found a correlation between unilateral tremor severity and contralateral pontine lesions (Feys et al., 2005). However, these findings are in a small cohort (N=14) of participants with bilateral MS tremor making it challenging to study tract lateralization. We attribute our finding to secondary trans- synaptic degeneration of contralateral cerebellar and thalamic neurons. Although it is not possible to test this directly using the cross-sectional data in this study, our findings are consistent with previous observations in MS (F. Boonstra et al., 2017) and in essential tremor (Broersma et al., 2016; Buijink et al., 2015; Neely et al., 2015). Furthermore, the pons is a structure that is thought to have connectivity across both hemispheres, which supports our hypothesis that interhemispheric connections contribute to bilateral atrophy. Taken together, these findings support the hypothesis that damage, including both atrophy and inflammation, in the cerebellum and thalamus are involved in the pathogenesis of tremor in MS patients.

The observed subcortical damage, combined with animal and treatment studies, suggests causality of tremor pathophysiology. Consistent with previous neurophysiological models of intention tremor in MS (Liu et al., 1997), the functional correlates reported here specifically link tremor during a visually guided task in MS to aberrant activity within a cortical region functionally linked to the cerebellum and thalamus. In addition to the more well-known connections between the premotor/SMA, and the cerebellum and thalamus in motor planning/control (Svoboda & Li, 2018; VanMeter et al., 1995), the IPL is a target of cerebellar efferents (Blakemore & Sirigu, 2003; Clower, West, Lynch, & Strick, 2001; Shum, Shiller, Baum, & Gracco, 2011) and through this it is thought to facilitate sensorimotor plasticity (Blakemore & Sirigu, 2003). Furthermore, connectivity mapping and functional

100

analyses have implicated regions of the IPL in both a lower-level, luminance-based and a higher-level, attention-based system for motion processing (Claeys, Lindsey, De Schutter, & Orban, 2003). While performing the task, we found that patients with tremor show increased activation in both ipsilateral IPL and premotor/SMA compared to patients without tremor. Furthermore, the amount of activation showed a moderate negative correlation with tremor severity in the tremor group. This finding suggests that stronger activation of this network, involving motor planning and sensorimotor integration areas, is associated with neural processes that attenuate the tremor, or when the activation is weaker, the tremor is more pronounced.

Failure of functional neuroplasticity could be due to an overall increase in disease burden. Indeed, we observed that ipsilateral thalamic volume and premotor/SMA activation lower with worsening of general disease severity. Functional neuroplasticity thought to compensate for abnormal or inadequate motor movement has also been shown in other neurological diseases. In Parkinson’s disease, an increased activation within the ipsilateral premotor cortex was associated with better motor performance (Mallol et al., 2007). Broersma et al. (2016) showed that increased activation in the primary motor cortex and cerebellum in essential tremor patients compared to healthy controls was associated with tremor. Cross-sectional studies are subject to antecedent-consequence bias, meaning that one cannot determine whether an observed difference is a cause or consequence of the disease. However, taken together, our results suggest that functional neuroplasticity within the ipsilateral cortex is a compensatory mechanism acting to alleviate the tremulous movement in patients that is invoked once a threshold for degeneration or demyelination within cerebello-thalamic pathways has been reached.

We showed that our joystick task was able to elicit tremulous movement in patients with clinical tremor. Patients in this study without clinical tremor still showed an increase in tremor-like movement during PLAY. Tremor-like movement is measured is the amount of movement with a frequency between the 3-10Hz. During the PLAY condition, there is more ‘jerky’ movement which increases the likelihood of falling within the tremor frequency band. Compared to the controlled condition MOVE, this could explain the increased amount of measured tremor-like movement in the MS controls while playing the game. However, there was significantly higher amount of tremor-like movement in the MS tremor participants.

101

Understanding the pathophysiology of tremor is essential for examining how drugs affect the clinical presentation of tremor. There is insufficient evidence supporting the efficacy of pharmaceutical treatment options for tremor patients with MS, while surgical treatment is very invasive and is a last resort option (Labiano-Fontcuberta & Benito-León, 2012). However, almost 50% of patients with MS with tremor use symptomatic medications (Meador, Salter, & II, 2016). Anticonvulsants are thought to be the most effective treatment (Meador et al., 2016); however, the mechanism of effect remains unclear as it is prescribed for a variety of MS symptoms other than tremor (Schapiro, 2002). Our study is the first in MS tremor research to use both structural and functional imaging. Of all the parameters, we found that when combined, the maximum activation within the ipsilateral premotor/SMA, contralateral cerebellar atrophy and ipsilateral cerebellar inflammation best predicts which patients with MS have tremor. Taken together, studies that examine the pathological processes for tremor in MS could help guide new treatments, understand treatment effect, and assist in patient selection for specific treatments (Koch, Mostert, Heersema, & De Keyser, 2007).

Longitudinal and treatment studies are warranted to better understand what leads to the failure of the functional neuroplasticity and to assess how treatments affect the tremor pathophysiology. Ultimately, new therapies can be introduced or targeted to tremor in patients with MS based on the in vivo structural and functional presentation of tremor in the brain.

5.6. Author contributions F. Boonstra has worked on study concept and design, data collection, data analyses, statistical analyses, data interpretation and critical revision of writing manuscript for intellectual content. G. Noffs has worked on study concept and design, data collection and critical revision of writing manuscript for intellectual content. T. Perera has worked on study concept and design and critical revision of writing manuscript for intellectual content. V. Jokubaitis has worked on data collection and critical revision of writing manuscript for intellectual content. A. Vogel, H. Butzkueven and A. Evans have worked on study concept and design, data interpretation and critical revision of writing manuscript for intellectual content. B. Moffat has worked on critical revision of writing manuscript for intellectual content. A. van der Walt has worked on study concept and design, data collection, data

102

interpretation and critical revision of writing manuscript for intellectual content. S. Kolbe has worked on study concept and design, data collection, data analyses, data interpretation and critical revision of writing manuscript for intellectual content.

5.7. Acknowledgement We would like to acknowledge the Murdoch Children’s Research Institute for providing access to the MRI suite and invaluable technical support. We would like to acknowledge C. Shanahan for her contributions to the project.

5.8. Disclosures F. Boonstra reports no disclosures. G. Noffs reports no disclosures. T. Perera reports no disclosures. V. Jokubaitis: received travel support from Biogen, Merck, and Sanofi, and received speaker’s honoraria from Biogen. She receives fellowship funding from Multiple Sclerosis Research Australia. A. Vogel: consults to Takeda and Pfizer. He is Chief Science Officer of Redenlab. He receives grant and fellowship funding from the National Health and Medical Research Council of Australia. B. Moffat: is a National Imaging Fellow via the NCRIS program (Government of Australia), he has/had research and clinical trial contracts with Siemens, Roche, Piramal, Jansen, GE and CSIRO. H. Butzkueven: Compensation for steering committee, advisory board and consultancy fees from Biogen, Merck, Roche, Novartis, Teva, Oxford Pharmagenesis; research support from Novartis, Biogen, Merck, NHMRC Australia, MS Research Australia, UK MS Trust, Monash University. A. Evans: received honoraria from Novartis for giving presentations and providing consultancy services. He has participated in scientific advisory board meetings for Novartis, UCB Pharma, Allergan, and Boehringer Ingelheim. He has received conference travel support from Boehringer Ingelheim. A. van der Walt: received travel support from Biogen Idec, Novartis, Teva, Merck, Sanofi and served on advisory boards for Biogen Idec, Novartis and Merck. She receives grant and fellowship funding from the National Health and Medical Research Council of Australia.

103

S. Kolbe: receives grant income from the National Health and Medical Research Council of Australia and has received honoraria from Novartis, Biogen and Merck

5.9. Study funding This project was funded by an NHMRC Project Grant (1085461 CIA Van der Walt). The Bionics Institute receives Operational Infrastructure Support from the Victorian Government.

5.10. References Alusi, S. H., Worthington, J., Glickman, S., & Bain, P. G. (2001). A study of tremor in multiple sclerosis. Brain, 124(Pt 4), 720-730. Anderson, V. M., Fox, N. C., & Miller, D. H. (2006). Magnetic resonance imaging measures of brain atrophy in multiple sclerosis. J Magn Reson Imaging, 23(5), 605-618. doi:10.1002/jmri.20550 [doi] Avants, B. B., Tustison, N. J., Song, G., Cook, P. A., Klein, A., & Gee, J. C. (2011). A Reproducible Evaluation of ANTs Similarity Metric Performance in Brain Image Registration. NeuroImage, 54(3), 2033-2044. doi:10.1016/j.neuroimage.2010.09.025 Bain, P. G., Findley, L. J., Atchison, P., Behari, M., Vidailhet, M., Gresty, M., . . . Marsden, C. D. (1993). Assessing tremor severity. J Neurol Neurosurg Psychiatry, 56(8), 868-873. Blakemore, S. J., & Sirigu, A. (2003). Action prediction in the cerebellum and in the parietal lobe. Exp Brain Res, 153(2), 239-245. doi:10.1007/s00221-003-1597-z Boonstra, F., Florescu, G., Evans, A., Steward, C., Mitchell, P., Desmond, P., . . . van der Walt, A. (2017). Tremor in multiple sclerosis is associated with cerebello-thalamic pathology. J Neural Transm (Vienna), 124(12), 1509-1514. doi:10.1007/s00702-017- 1798-4 Boonstra, F. M. C., Perera, T., Noffs, G., Marotta, C., Vogel, A. P., Evans, A. H., . . . Kolbe, S. C. (2018). Novel Functional MRI Task for Studying the Neural Correlates of Upper Limb Tremor. Frontiers in Neurology, 9(513). doi:10.3389/fneur.2018.00513 Broersma, M., van der Stouwe, A. M. M., Buijink, A. W. G., de Jong, B. M., Groot, P. F. C., Speelman, J. D., . . . Maurits, N. M. (2016). Bilateral cerebellar activation in unilaterally challenged essential tremor. NeuroImage: Clinical, 11, 1-9. doi:https://doi.org/10.1016/j.nicl.2015.12.011

104

Browne, P., Chandraratna, D., Angood, C., Tremlett, H., Baker, C., Taylor, B. V., & Thompson, A. J. (2014). Atlas of Multiple Sclerosis 2013: A growing global problem with widespread inequity. Neurology, 83(11), 1022-1024. doi:10.1212/WNL.0000000000000768 Buijink, A. W., Broersma, M., van der Stouwe, A. M., van Wingen, G. A., Groot, P. F., Speelman, J. D., . . . van Rootselaar, A. F. (2015). Rhythmic finger tapping reveals cerebellar dysfunction in essential tremor. Parkinsonism Relat Disord, 21(4), 383-388. doi:10.1016/j.parkreldis.2015.02.003 Buma, F. E., Raemaekers, M., Kwakkel, G., & Ramsey, N. F. (2015). Brain Function and Upper Limb Outcome in Stroke: A Cross-Sectional fMRI Study. PLoS One, 10(10), e0139746. doi:10.1371/journal.pone.0139746 Carpenter, M. B., & Hanna, G. R. (1962). Effects of thalamic lesions upon cerebellar dyskinesia in the rhesus monkey. J Comp Neurol, 119, 127-147. Claeys, K. G., Lindsey, D. T., De Schutter, E., & Orban, G. A. (2003). A higher order motion region in human inferior parietal lobule: evidence from fMRI. Neuron, 40(3), 631-642. Clower, D. M., West, R. A., Lynch, J. C., & Strick, P. L. (2001). The inferior parietal lobule is the target of output from the superior colliculus, hippocampus, and cerebellum. J Neurosci, 21(16), 6283-6291. Deuschl, G., Bain, P., & Brin, M. (1998). Consensus statement of the Movement Disorder Society on Tremor. Ad Hoc Scientific Committee. Mov Disord, 13, 2-23. Fahn, S. (1986). What is it? Case 1 1986. Presentation of a case. Movement Disorders, 1(4), 275-280. doi:10.1002/mds.870010408 Feys, P., Maes, F., Nuttin, B., Helsen, W., Malfait, V., Nagels, G., . . . Liu, X. (2005). Relationship between multiple sclerosis intention tremor severity and lesion load in the brainstem. Neuroreport, 16(12), 1379-1382. Foote, K. D., Seignourel, P., Fernandez, H. H., Romrell, J., Whidden, E., Jacobson, C., . . . Okun, M. S. (2006). Dual electrode thalamic deep brain stimulation for the treatment of posttraumatic and multiple sclerosis tremor. Neurosurgery, 58(4 Suppl 2), ONS-280-285; discussion ONS-285-286. doi:10.1227/01.neu.0000192692.95455.fd Julian, L. J., Vella, L., Vollmer, T., Hadjimichael, O., & Mohr, D. C. (2008). Employment in multiple sclerosis: Exiting and re-entering the work force. Journal of Neurology, 255(9), 1354-1360. doi:10.1007/s00415-008-0910-y

105

Koch, M., Mostert, J., Heersema, D., & De Keyser, J. (2007). Tremor in multiple sclerosis. J Neurol, 254(2), 133-145. doi:10.1007/s00415-006-0296-7 [doi] Koziarska, D., Król, J., Nocoń, D., Kubaszewski, P., Rzepa, T., & Nowacki, P. (2018). Prevalence and factors leading to unemployment in MS (multiple sclerosis) patients undergoing immunomodulatory treatment in Poland. PLoS One, 13(4), e0194117. doi:10.1371/journal.pone.0194117 Kurtzke, J. F. (1983). Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology, 33(11), 1444-1452. Labiano-Fontcuberta, A., & Benito-León, J. (2012). Understanding Tremor in Multiple Sclerosis: Prevalence, Pathological Anatomy, and Pharmacological and Surgical Approaches to Treatment. Tremor and Other Hyperkinetic Movements, 2, tre-02-109- 765-102. Liu, X., Miall, C., Aziz, T. Z., Palace, J. A., Haggard, P. N., & Stein, J. F. (1997). Analysis of action tremor and impaired control of movement velocity in multiple sclerosis during visually guided wrist-tracking tasks. Mov Disord, 12(6), 992-999. doi:10.1002/mds.870120624 Lotze, M., Flor, H., Grodd, W., Larbig, W., & Birbaumer, N. (2001). Phantom movements and pain An fMRI study in upper limb amputees. Brain, 124(11), 2268- 2277. doi:10.1093/brain/124.11.2268 Mallol, R., Barrós-Loscertales, A., López, M., Belloch, V., Parcet, M. A., & Ávila, C. (2007). Compensatory cortical mechanisms in Parkinson's disease evidenced with fMRI during the performance of pre-learned sequential movements. Brain Res, 1147, 265-271. doi:https://doi.org/10.1016/j.brainres.2007.02.046 Meador, W., Salter, A. R., & II, J. R. R. (2016). Symptomatic Management of Multiple Sclerosis–Associated Tremor Among Participants in the NARCOMS Registry. International Journal of MS Care, 18(3), 147-153. doi:10.7224/1537-2073.2015-008 Miall, R. C., Weir, D. J., Wolpert, D. M., & Stein, J. F. (1993). Is the cerebellum a smith predictor? J Mot Behav, 25(3), 203-216. doi:10.1080/00222895.1993.9942050 Nakamura, R., Kamakura, K., Tadano, Y., Hosoda, Y., Nagata, N., Tsuchiya, K., . . . Shibasaki, H. (1993). MR imaging findings of tremors associated with lesions in cerebellar outflow tracts: report of two cases. Mov Disord, 8(2), 209-212. doi:10.1002/mds.870080218 Neely, K. A., Kurani, A. S., Shukla, P., Planetta, P. J., Wagle Shukla, A., Goldman, J. G., . . . Vaillancourt, D. E. (2015). Functional Brain Activity Relates to 0-3 and 3-8 Hz

106

Force Oscillations in Essential Tremor. Cereb Cortex, 25(11), 4191-4202. doi:10.1093/cercor/bhu142 Pittock, S. J., McClelland, R. L., Mayr, W. T., Rodriguez, M., & Matsumoto, J. Y. (2004). Prevalence of tremor in multiple sclerosis and associated disability in the Olmsted County population. Mov Disord, 19(12), 1482-1485. doi:10.1002/mds.20227 Reddy, H., Narayanan, S., Arnoutelis, R., Jenkinson, M., Antel, J., Matthews, P. M., & Arnold, D. L. (2000). Evidence for adaptive functional changes in the cerebral cortex with axonal injury from multiple sclerosis. Brain, 123, 2314-2320. Reddy, H., Narayanan, S., Woolrich, M., Mitsumori, T., Lapierre, Y., Arnold, D. L., & Matthews, P. M. (2002). Functional brain reorganization for hand movement in patients with multiple sclerosis: defining distinct effects of injury and disability. Brain, 125(Pt 12), 2646-2657. Rinker, J. R., Salter, A. R., Walker, H., Amara, A., Meador, W., & Cutter, G. R. (2015). Prevalence and characteristics of tremor in the NARCOMS multiple sclerosis registry: a cross-sectional survey. BMJ Open, 5(1). doi:10.1136/bmjopen-2014- 006714 Schapiro, R. T. (2002). Pharmacologic Options for the Management of Multiple Sclerosis Symptoms. Neurorehabilitation and Neural Repair, 16(3), 223-231. doi:10.1177/154596802401105162 Schmidt, P., Gaser, C., Arsic, M., Buck, D., Forschler, A., Berthele, A., . . . Muhlau, M. (2012). An automated tool for detection of FLAIR-hyperintense white-matter lesions in Multiple Sclerosis. NeuroImage, 59(4), 3774-3783. doi:10.1016/j.neuroimage.2011.11.032 Shum, M., Shiller, D. M., Baum, S. R., & Gracco, V. L. (2011). Sensorimotor integration for speech motor learning involves the inferior parietal cortex. The European journal of neuroscience, 34(11), 1817-1822. doi:10.1111/j.1460-9568.2011.07889.x Smith, S. M., Jenkinson, M., Woolrich, M. W., Beckmann, C. F., Behrens, T. E., Johansen-Berg, H., . . . Matthews, P. M. (2004). Advances in functional and structural MR image analysis and implementation as FSL. NeuroImage, 23, S208-219. doi:10.1016/j.neuroimage.2004.07.051 Svoboda, K., & Li, N. (2018). Neural mechanisms of movement planning: motor cortex and beyond. Current Opinion in Neurobiology, 49, 33-41. doi:https://doi.org/10.1016/j.conb.2017.10.023

107

Van Der Walt, A., Sung, S., Spelman, T., Marriott, M., Kolbe, S., Mitchell, P., . . . Butzkueven, H. (2012). A double-blind, randomized, controlled study of botulinum toxin type A in MS-related tremor. Neurology, 79(1), 92-99. doi:10.1212/WNL.0b013e31825dcdd9 VanMeter, J. W., Maisog, J. M., Zeffiro, T. A., Hallett, M., Herscovitch, P., & Rapoport, S. I. (1995). Parametric Analysis of Functional Neuroimages: Application to a Variable-Rate Motor Task. NeuroImage, 2(4), 273-283. doi:https://doi.org/10.1006/nimg.1995.1035 Wilms, H., Sievers, J., & Deuschl, G. (1999). Animal models of tremor. Mov Disord, 14(4), 557-571. Yap, L., Kouyialis, A., & Varma, T. R. (2007). Stereotactic neurosurgery for disabling tremor in multiple sclerosis: thalamotomy or deep brain stimulation? Br J Neurosurg, 21(4), 349-354. doi:10.1080/02688690701544002

5.11. Supplementary figures

Supplementary Figure 1. Y-flipping of the functional images. Top row shows the activation for 10 participants who played the game with their right hand. Bottom row shows the activation for 10 participants who played the game with their left hand and subsequently had the images y-flipped.

108

Supplementary Figure 2. Lesion segmentation tool on a double inversion recovery image. The extend of lesion segmentation using the lesion prediction algorithm within the lesion segmentation tool software.

109

Supplementary Figure 3. Atrophy measures in all three groups. The mean volumetric measures for the cerebellum (top), thalamus (middle) and superior cerebellar peduncle (bottom) in healthy controls, multiple sclerosis participants without tremor and multiple sclerosis participants with tremor. Abbreviations: SCP, superior cerebellar peduncle.

110

Treatment Effect Study

OnabotulinumtoxinA treatment for MS-tremor modifies fMRI tremor response in central sensory-motor integration areas 1. Frederique M. C. Boonstra, MSc 1, Andrew Evans, MD 2,4, Gustavo Noffs, MD 2,3, Thushara Perera, PhD 4,5, Vilija Jokubaitis 6, Jim Stankovich 6, Adam P. Vogel, PhD 3,4,7,8, Bradford A. Moffat, PhD 1, Helmut Butzkueven, PhD 6, Scott C. Kolbe, PhD 1,9,* and Anneke van der Walt, PhD 2,4,6,* 2. Department of Medicine and Radiology, University of Melbourne 3. Department of Neurology, Royal Melbourne Hospital, Australia 4. Centre for Neuroscience of Speech, University of Melbourne, Victoria, Australia 5. The Bionics Institute, Australia 6. Department of Medical Bionics, University of Melbourne, Australia 7. Department of Neuroscience, Central Clinical School, Monash University, Australia 8. Department of Neurodegeneration, Hertie Institute for Clinical Brain Research, University of Tübingen, Germany 9. Redenlab, Victoria, Australia 10. Florey Institute of Neuroscience and Mental Health, Australia *Authors contributed equally

Accepted in Multiple Sclerosis Journal and ‘in press’

111

6.1. Abstract Background: Treatment of tremor in multiple sclerosis (MS) is an unmet need. OnabotulinumtoxinA (BoNT-A) has shown promising results; however, little is known regarding its central effects. Objective: To identify effects of BoNT-A on neural plasticity in MS and upper-limb tremor using functional MRI. Methods: Forty-three MS participants were randomized to receive intramuscular injections of placebo (n=22) or BoNT-A (n=21). Tremor was quantified using the Bain score (0-10) for severity, handwriting and Archimedes drawing at baseline, 6 weeks and 12 weeks. Functional MRI activation within two previously identified clusters, ipsilateral inferior parietal cortex (IPL) and supplementary motor cortex (SMC) of compensatory activity, was measured at baseline and 6 weeks. Results: Treatment with BoNT-A resulted in improved tremor severity -0.79 (p=0.007) and handwriting -0.53 (p=0.014) scores over 12 weeks. The BoNT-A group showed a reduction in activation within the IPL (p=0.034), but not in the SMC. The change in IPL activation correlated with the reduction in tremor severity from baseline to 12 weeks (β =0.608; p=0.015) in the BoNT-A group. No fMRI changes were seen in the placebo treated group. Conclusion: Reduction in MS-tremor severity after intramuscular injection with BoNT-A is associated with changes in brain activity in sensorimotor integration regions.

6.2. Introduction Tremor is a common symptom in multiple sclerosis (MS) with a prevalence between 45% and 46.8% (Rinker et al., 2015). There are currently no validated treatments for MS tremor; however, evidence for positive treatment effects using intramuscular onabotulinumtoxinA (BoNT-A) have been observed in Parkinson’s disease, essential tremor, spasticity and dystonia (Brin et al., 2001; Foley et al., 2013; Marvulli et al., 2012; Samotus, Kumar, Rizek, & Jog, 2018; Trosch & Pullman, 1994). Recently, Van der Walt et al. (Van Der Walt et al., 2012) performed a double-blinded, randomized, controlled crossover study of targeted BoNT-A injections in 33 tremor-affected limbs of MS patients. They reported significant improvement (p<0.001) following BoNT-A treatment compared with placebo in overall tremor severity as well as writing and drawing of an Archimedes spiral at 6 weeks and 12 weeks. Side-effects emerged as muscle weakness in 42.2% of patients treated with BoNT-A as opposed to 6.1% for placebo (p<0.001). However, weakness was generally mild (just detectable) to moderate (still able to use limb) and on average, resolved within 2 weeks.

112

Studies in primary dystonia have shown that, in addition to a local effect, BoNT-A seems to modulate brain function. Specifically, pre- and post BoNT-A imaging studies have shown that BoNT-A treatment normalizes disease-related changes in muscle activity leading to improved proprioception and enduring neural plasticity (Blood et al., 2006; Delnooz, Pasman, Beckmann, & van de Warrenburg, 2013; Opavsky, Hlustik, Otruba, & Kanovsky, 2011; Walsh & Hutchinson, 2007). BoNT-A is an enzyme that inhibits the release of acetylcholine by preventing the fusion of synaptic vesicles with the membrane (Tighe & Schiavo, 2013). As such, signal transmission between the nerve and muscle is blocked, with resultant reduced muscle activation. To-date, no studies have investigated neural activation changes in people with MS-associated tremor treated with intramuscular BoNT-A. Such information will elucidate mechanisms of action for BoNT-A in the treatment of MS-associated tremor. It may also assist in identifying patients best suited for this treatment approach.

We have previously characterized a network of functional brain regions associated with tremor in MS (F. M. C. Boonstra, Noffs, et al., 2018). Our results provided evidence that tremor severity reflects an interaction between damage (atrophy) to the cerebellar-thalamic pathway, and functional neuroplasticity within cortical sensory-motor planning and integration areas. This study aimed to expand upon our previous work to investigate whether treatment with BoNT-A modulates neural activity within these brain regions. We hypothesised that intramuscular BoNT- A treatment would ameliorate tremor and reduce functional activations within previously identified compensatory brain regions compared to placebo treatment. Furthermore, we hypothesized that participants with more severe pathology along the cerebello-thalamo-cortical tract will be less responsive to treatment.

6.3. Methods

Study design This study forms part of an ongoing randomized, controlled trial of the efficacy and safety of BoNT-A in MS-related arm tremor (ACTRN12617000379314) that is an extension of our previous cross-sectional study (F. M. C. Boonstra, Noffs, et al., 2018). Clinical and tremor assessments were performed at baseline 6 weeks and 12 weeks. MRI scans were performed prior to randomization and at 6 weeks following treatment. The study was approved by the Melbourne Health Human Research Ethics Committee. All of the participants gave informed voluntary written consent.

113

Participants We recruited 43 participants with relapsing-remitting and secondary-progressive multiple sclerosis and unilateral upper limb tremor (March 2016 – April 2018). Exclusion criteria included treatment with BoNT-A in the 6 months prior to recruitment, contraindication to BoNT-A injections, inability to cease other anti-tremor medications at least 1 week before baseline, or upper limb weakness (strength less than 4.5/5) (Matthews, 1977). Forty participants completed the longitudinal 6 weeks and 12 weeks assessment.

Randomization and BoNT-A injections Randomization was performed at baseline on a 1:1 basis by an independent third party. An expert movement disorder neurologist administered targeted injections of BoNT-A or placebo after a careful clinical assessment of the tremor pattern. A maximum dose of 150 IU BoNT-A or normal saline placebo was injected at the baseline visit. An unblinded pharmacist prepared in equal volumes 100 units of BoNT-A diluted per 2 mL 0.9% sterile saline or 2 mL 0.9% sterile saline syringes. To optimise targeted treatment, injections were guided using the Dantec Clavis handheld electromyogram (EMG) and stimulation device (Natus Neurology).

Clinical assessments Disease and tremor history were recorded at baseline and disease severity measured using the Expanded Disability Status Scale (EDSS) (Kurtzke, 1983) by a qualified neurologist.

Strength measurements To determine possible weakness as a result of BoNT-A injections, we assessed participants strength at baseline, 6 weeks and 12 weeks. Strength was assessed using Medical Research Council (MRC) ratings (Matthews, 1977). Adverse events were documented 1 week following baseline treatments using a standardized interview. Reported weakness was then assessed clinically and followed until resolution.

Tremor measurement At each visit, the patient sat comfortably and performed a standardized tremor assessment which was video recorded for later review (Bain et al., 1993). Tremor was recorded at rest, posture against gravity, in the “bat-wing” position and with finger-nose-testing. Tremor was rated using the Bain score (where 0=no tremor and 10=severe tremor) for overall tremor severity, writing a

114

standardised sentence (“This is a sample of my best handwriting”), and Archimedes spiral drawing on a pre-drawn pattern (Alusi, Worthington, Glickman, & Bain, 2001). Video recordings and handwriting sheets were relabelled, intermixed and rated by a blinded trained investigator. Intra-rater reliability was measured by rating videos of 10 randomly selected participants twice, 2 times apart.

Imaging protocols All patients underwent imaging using a 3T MRI system (Trio, Siemens, Erlangen) at baseline and 6 weeks. Baseline protocol included: (a) high-resolution 3D T1-weighted MPRAGE scan with online motion correction (TR=2530ms; TE=2.5ms; TI=1260ms; FOV=176x256mm; voxel size=1.0x1.0x1.0mm), (b) high resolution 3D T2-weighted double inversion recovery (DIR) sequence (TR=7400ms; TE=324ms; TI=3000ms; flip angle=120°; ETL=625; FOV=144x220mm; voxel size=1.0x1.0x1.0), and (c) two runs of functional MRI (fMRI; TR = 1.5ms, TE = 33ms, FOV = 2040*2040, matrix = 104*104, slice thickness = 2.5, flip angle = 85.0, multi-band slice acceleration factor = 3, volumes = 200). At 6 weeks, DIR and fMRI sequences were repeated.

Joystick fMRI task The fMRI task performed is described in detail by Boonstra et al. (F. M. C. Boonstra, Perera, et al., 2018). The task involves playing a joystick video game with three conditions: PLAY, during which the participants plays the game using the joystick; MOVE, move the joystick left to right; and WATCH, observe the game while holding the joystick still. The condition PLAY was the active condition that has been shown to elicit tremulous movement in MS-related tremor. MOVE and WATCH were two conditions that control for basic movement and visual attention respectively. Furthermore, in healthy controls, functional network activation during PLAY > WATCH + MOVE was found to be reproducible with no network changes detected over 6 weeks.

Imaging analyses Baseline, structural MRI was analysed using FreeSurfer version 5.3 (http://surfer.nmr.mgh.harvard.edu/), Osirix Lite v9.0.2, and Lesion Segmentation Toolbox (LST) version 2.0.15 for SPM (Schmidt et al., 2012). We measured the volume of the thalamus and cerebellum. The area of the SCP was manually labelled as described previously in Boonstra et al. (F. Boonstra et al., 2017). Lesion load was derived from the DIR images using the lesion

115

prediction algorithm within LST. To optimize the automatic lesion segmentation, we manually determined the optimal threshold for each participant and then corrected the final lesion maps. Lesion load within the thalamus and cerebellum was determined using brain segmentation maps derived from the FreeSurfer analyses described above. The longitudinal DIR image was used to determine the development of additional lesions at 6 weeks compared to baseline.

FSL FEAT 5.0.8 (Smith et al., 2004) was used for all fMRI analyses. At baseline, we identified two regions of functional abnormal activation: the ipsilateral inferior parietal lobule (IPL: 451 voxels; maximum/mean z-stat 2.99/1.30; MNI coordinates of voxel with maximum z-stat 56, - 56, 26), and the premotor/supplementary motor cortex (premotor/SMC: 691 voxels; maximum/mean z-stat 3.10/1.12; MNI coordinates of voxel with maximum z-stat 36, 16, 52) (F. M. C. Boonstra, Noffs, et al., 2018). Featquery was used to estimate the maximum activation within both clusters at baseline and 6 weeks.

Statistical analyses Statistical analyses were performed in SPSS version 24. Continuous variables were assessed for normality. Quantitative variables were summarized using mean and standard deviation (sd). Baseline group differences for age, affected hand, gender, disease duration, EDSS and disease course we performed using an independent samples t-test. Cohen's κ was used to determine the intra-rater reliability for clinical scores. Treatment effects were assessed by parametric paired t-test for both the BoNT-A group and the placebo group from baseline to 6 weeks and from baseline to 12 weeks. Similarly, a paired t-test was performed to assess the functional activation changes from baseline to 6 weeks. To determine if the effect was stronger in the BoNT-A group compared to the placebo group, we used an independent t- test for both the 6 weeks and 12 weeks change in Bain score and the 6 weeks change in functional activation.

Adverse effects on muscle strength were measured using paired t-tests. We compared both the change from baseline to 6 weeks and baseline to 12 weeks for both the BoNT-A group and the placebo group. The dose response relationship was assessed using a Pearson’s rank correlation between BoNT-A dose injection at baseline and the change in strength at both 6 weeks and 12 weeks. We examined the relationship between the change in functional activation and the change in all Bain scores at 6 weeks and 12 weeks using a linear regression. Finally, we performed a linear regression between baseline structural imaging parameters and the change in Bain scores at

116

both 6 weeks and 12 weeks to determine possible structural predispositions to treatment responsiveness while controlling for age. Significance was set at p<0.05, and adjustment of statistical significance for multiple comparisons was performed using the false discovery rate correction.

6.4. Results

Participants There were no significant differences in age, tremor-affected upper limb, gender, disease course and EDSS between the placebo group and the BoNT-A group (Table 1). There was a significant difference in disease duration with the placebo group having a slightly longer disease duration.

Table 1. Participants’ demographics and disease characteristics at baseline. * indicated uncorrected significance p<0.05. Abbreviations: sd, standard deviation; EDSS, expanded disability status scale. Placebo group Botox group p

N 22 21

Age, years (sd) 47.0 (9.35) 45.9 (13.5) 0.956

Affected hand, right % 77.3 66.7 0.450

Gender, female % 77.3 71.4 0.670

Disease duration, years (sd) 16.6 (7.10) 11.1 (7.55) 0.020*

Disease course, % 0.704 Relapsing-remitting MS 31.8 42.9 Secondary progressive MS 68.2 52.4 Primary progressive MS 0.0 4.8

EDSS, mean (sd) 3.95 (1.71) 4.41 (1.76) 0.394

Effects of BoNT-A treatment on MS-related tremor Ratings of the Bain scores showed a moderate to strong intra-rater reliability, κ = 0.723, p < 0.001. Tremor was rated as mild to moderate in severity (Bain scores 1 to 4 out of 10) in all participants. Outcomes are reported as the mean change (with sd) at 6 weeks and 12 weeks compared to baseline in both the placebo and the BoNT-A group (Table 2). There were no

117

significant differences in any Bain scores between the groups at baseline. From baseline to 6 weeks, there was a decrease in overall tremor severity in the treatment group compared to placebo for handwriting (p=0.049). From baseline to 12 weeks, there was a significant improvement in overall tremor severity (p=0.007) and handwriting (p=0.014) in the BoNT-A treatment group, while a trend for improved Archimedes drawing (p=0.069) was observed. No significant clinical improvements were noted in the placebo group apart from a trend for a lower tremor severity score (p=0.055) from baseline to 12 weeks. Furthermore, there were no significant differences between the two groups in change of the Bain scores.

Table 2. Baseline strength, tremor severity scores and fMRI activation with posttreatment change at 6 and 12 weeks. The 6 weeks and 12 weeks measures are the change compared to baseline. Tremor was measured using the Bain scores. * indicates uncorrected significance p<0.05 Abbreviations: sd, standard deviation; IPL, inferior parietal lobule; SMC, supplementary motor cortex. Placebo group BoNT-A group Baselin D 6 p D 12 p Baselin D 6 p D 12 p Visit e Week Week e Week Week -0.05 -0.03 -0.11 -0.05 Strength, mean 4.93 0.28 0.35 4.99 0.029 (0.20 (0.16 (0.20 (0.12 0.070 (sd) (0.11) 5 7 (0.03) * ) ) ) ) Tremor -0.25 -0.60 -0.26 -0.79 2.60 0.34 0.05 3.89 0.007 severity, mean (1.16 (1.31 (1.05 0.287 (1.13 (2.30) 9 5 (3.25) * (sd) ) ) ) ) -0.10 -0.25 -0.37 -0.53 Handwriting, 2.80 0.73 0.36 3.95 0.049 0.014 (1.29 (1.21 (0.76 (0.84 mean (sd) (2.59) 3 7 (3.91) * * ) ) ) ) -0.35 -0.25 -0.47 -0.63 Archimedes, 2.75 0.29 0.41 3.74 (1.46 (1.33 (1.26 0.120 (1.42 0.069 mean (sd) (2.45) 7 2 (3.09) ) ) ) ) -0.89 -1.68 3.96 0.21 5.04 0.034 IPL, mean (sd) (2.76 (2.99 (1.58) 8 (2.24) * ) ) 0.89 -1.25 Premotor/SMC 5.20 0.11 6.66 (2.14 (4.02 0.218 , mean (sd) (2.06) 9 (3.28) ) )

118

A significant decrease in muscle strength was noted in the BoNT-A-treated group at 6 weeks compared to baseline (p=0.029). Strength had recovered close to baseline by 12 weeks with no inter-group difference detectable. Furthermore, there was a correlation between the dose of BoNT-A injection at baseline and the reduction in strength at 6 weeks (r=-0.471, p=0.042) and 12 weeks (r=-0.511, p=0.025). These modestly significant muscle strength comparisons did not survive correction for multiple testing.

MRI structural and functional measures and tremor improvement after BoNT-A There was a significant difference in lesions within the ipsilateral cerebellum at baseline between the groups (supplementary table 1). However, we found a negative association between the lesion load in the ipsilateral cerebellum at baseline and the change in tremor severity from baseline to 6 weeks (β=-0.515, p=0.032) and handwriting from baseline to 6 weeks (β =-0.583, p=0.014). Results are reported in supplementary table 2. There was no relationship between the baseline functional activation in the IPL and premotor/SMC and the change in Bain scores.

Functional neuroplasticity in response to BoNT-A Two functional activation clusters previously identified as being significantly associated with tremor were analysed for potential treatment related changes. In the BoNT-A treatment group, there was a significant decrease in activity within IPL (p=0.034) at 6 weeks compared to baseline, but not within premotor/SMC (p=0.218). Neither IPL nor the premotor/SMC changed in activity from baseline to 6 weeks in the placebo group (p=0.218 and p=0.119). There was no significant difference between both groups and the change in functional activation.

There was a significant association between the change in cluster 1 activation and change in overall tremor severity from baseline to 12 weeks (β =0.608, p=0.015) (see figure 1) in patients that received BoNT-A only. No other associations were found between changes in cluster activation and changes in Bain scores.

119

Figure 1. The relation between the change in tremor and the change in functional activation within the ipsilateral inferior parietal lobule (IPL). The top image indicates the location of the IPL cluster. The bottom image depicts the relationship between the change in IPL activation from baseline to 6 weeks (y-axis) and the change in tremor severity measured using the Bain scores from baseline to 12 weeks (x-axis). 6.5. Discussion

120

This is the first study to examine the effect of BoNT-A on the pathophysiology of tremor in people with MS. To this end, we performed a longitudinal study in which participants who underwent fMRI before and after receiving BoNT-A or placebo injections. We found a reduction in clinical tremor severity after injection at 6 weeks, with maximum effect at 12 weeks, confirming results of our previous work that reported reduced tremor severity after intramuscular BoNT-A treatment (Van Der Walt et al., 2012). Mild weakness that was not clinically disabling was a common side effect at 6 weeks after receiving BoNT-A and normalised by 12 weeks.

Analysis of longitudinal fMRI data, corrected for treatment-related weakness, demonstrated a significant reduction in neural activation within the ipsilateral inferior parietal lobule (IPL) after BoNT-A injection that was not found in participants receiving placebo. The IPL plays an important role in sensorimotor adaptation and is a target of cerebellar efferents (Blakemore & Sirigu, 2003; Clower, West, Lynch, & Strick, 2001; Shum, Shiller, Baum, & Gracco, 2011). Furthermore, it is suggested the interaction between IPL and cerebellum facilitates sensorimotor plasticity (Blakemore & Sirigu, 2003). Our previous work confirms the importance of this region of interest in tremor pathophysiology. In particular, we demonstrated that activation in the ipsilateral IPL occurs more strongly in MS patients with tremor when compared to matched MS- controls without tremor (F. M. C. Boonstra, Noffs, et al., 2018). Activation in the IPL was greatest in tremor affected pwMS with low tremor scores or milder tremor, raising the possibility that the increased IPL activation represents neural compensation acting to reduce tremor severity. This hypothesis is further supported by the current study, where reduced IPL activation was significantly associated with reduced tremor severity after BoNT-A treatment. Considering our observations in the context of neural compensation, the reduction in tremor severity associated with BoNT-A treatment would lead to a reduced requirement for neural compensation and thus reduced activation.

These findings are further supported by previous studies that suggest modulation of neural organisation occurs due to peripheral alterations in muscle spindle activation (Blood et al., 2006; Currà, Trompetto, Abbruzzese, & Berardelli, 2004; Delnooz et al., 2013; Opavsky et al., 2011; Walsh & Hutchinson, 2007). Specifically, by blocking muscle spindle activity with BoNT-A, the afferent input to the central nervous system is altered which is thought to be responsible for functional changes in motor mechanisms (Filippi, Errico, Santarelli, Bagolini, & Manni, 1993; Rosales, Arimura, Takenaga, & Osame, 1996). We observed no change in premotor/SMC

121

activation, an area involved in motor planning (Halsband, Ito, Tanji, & Freund, 1993) that may suggest that BoNT-A induces changes in integration areas more strongly in pwMS and tremor. However, the exact network modulatory effects due to BoNT-A remains unclear.

We found no evidence that structural MRI changes can be predictive of treatment responsiveness to BoNT-A. Rather, we report a negative association between the lesion load within the ipsilateral cerebellum and improved tremor scores following BoNT-A. This observation did not survive multiple comparisons correction and is likely a type I error.

In our previously described randomized controlled cross-over study of BoNT-A efficacy in MS- tremor (8) clinical effect remained significant 12 weeks post treatment. The effect of BoNT-A beyond its known pharmacodynamics profile has been demonstrated in studies of sudomotor function, bladder function and spasticity where persistent benefit for up to 6 months is well described (Birklein, Eisenbarth, Erbguth, & Winterholler, 2003; Safarpour, Mousavi, & Jabbari, 2017). The pathophysiology of this prolonged clinical effect, particular in the motor system, has been postulated to be due to central neuroplastic effects but this has not been objectively demonstrated until now. Longitudinal fMRI at further time-points, and until clinical effect declines, would be needed to fully elucidate the relationship between central activation patterns and clinical changes.

The clinical results reported here are part of a larger ongoing study and it is likely that the trends for clinical improvement in tremor severity, handwriting and Archimedes spiral drawing will be clarified when analysed in a larger cohort. A significant proportion of BoNT-A treated patients experienced mild weakness at 6 weeks which may have affected dexterity and tasks such as writing and Archimedes drawing. This is supported by improvement in tremor severity at 12 weeks when weakness resolved. Future studies that could greatly benefit from blinded rating of videos, which is a common tool employed in interventional movement disorder studies. The addition of paraclinical measures of tremor reduction such as accelerometry may be more sensitive and should be considered for future trials (Perera et al., 2018). Furthermore, the study cohort is small and limits our ability to evaluate differential effects and activation patterns in more severe compared to milder tremor. Understanding these differences would be valuable as it may provide information on network ability to compensate, or not, for more severe physical dysfunction.

122

To conclude, we provide evidence that peripherally administered BoNT-A exerts a central neuroplastic effect by reducing neural activation patterns in tremor-affected pwMS that is associated with a positive clinical response to treatment at 6 and 12 weeks. These results provide valuable insights into tremor pathophysiology in MS and BoNT-A efficacy. Further studies are needed to determine if baseline fMRI activation patterns could be predictive of BoNT-A treatment response in upper limb MS tremor.

6.6. Authors contributions F. Boonstra has worked on study concept and design, data collection, data analyses, statistical analyses, data interpretation and critical revision of writing manuscript for intellectual content. A. Evans has worked on study concept and design, data interpretation and critical revision of writing manuscript for intellectual content. G. Noffs has worked on study concept and design, data collection and critical revision of writing manuscript for intellectual content. T. Perera has worked on study concept and design and critical revision of writing manuscript for intellectual content. V. Jokubaitis has worked on data collection and critical revision of writing manuscript for intellectual content. J. Stankovich has worked on statistical analyses and critical revisions of the manuscript for intellectual content. A. Vogel, H. Butzkueven has worked on study concept and design, data interpretation and critical revision of writing manuscript for intellectual content. B. Moffat has worked on critical revision of writing manuscript for intellectual content. S. Kolbe has worked on study concept and design, data collection, data analyses, data interpretation and critical revision of writing manuscript for intellectual content. A. van der Walt has worked on study concept and design, data collection, data interpretation and critical revision of writing manuscript for intellectual content.

6.7. Acknowledgement We would like to acknowledge the Murdoch Children’s Research Institute for providing access to the MRI suite and invaluable technical support. We would like to acknowledge C. Shanahan for her contributions to the project.

6.8. Disclosures F. Boonstra reports no disclosures. A. Evans: received honoraria from Novartis for giving presentations and providing consultancy services. He has participated in scientific advisory board meetings for Novartis, UCB Pharma,

123

Allergan, and Boehringer Ingelheim. He has received conference travel support from Boehringer Ingelheim. G. Noffs reports no disclosures. T. Perera reports no disclosures. V. Jokubaitis: received travel support from Biogen, Merck, and Sanofi, and received speaker’s honoraria from Biogen. She receives fellowship funding from Multiple Sclerosis Research Australia. J. Stankovich reports no disclosures. A. Vogel: consults to Takeda and Pfizer. He is Chief Science Officer of Redenlab. He receives grant and fellowship funding from the National Health and Medical Research Council of Australia. B. Moffat: is a National Imaging Fellow via the NCRIS program (Government of Australia), he has/had research and clinical trial contracts with Siemens, Roche, Piramal, Jansen, GE and CSIRO. H. Butzkueven: Compensation for steering committee, advisory board and consultancy fees from Biogen, Merck, Roche, Novartis, Teva, Oxford Pharmagenesis; research support from Novartis, Biogen, Merck, NHMRC Australia, MS Research Australia, UK MS Trust, Monash University. S. Kolbe: receives grant income from the National Health and Medical Research Council of Australia and has received honoraria from Novartis, Biogen and Merck A. van der Walt: received travel support from Biogen Idec, Novartis, Teva, Merck, Sanofi and served on advisory boards for Biogen Idec, Novartis and Merck. She receives grant and fellowship funding from the National Health and Medical Research Council of Australia.

6.9. Study funding This project was funded by an NHMRC Project Grant (1085461 CIA Van der Walt) and Fellowship (1135683 Vogel). The Bionics Institute receives Operational Infrastructure Support from the Victorian Government.

6.10. References Alusi, S. H., Worthington, J., Glickman, S., & Bain, P. G. (2001). A study of tremor in multiple sclerosis. Brain, 124(Pt 4), 720-730. Bain, P. G., Findley, L. J., Atchison, P., Behari, M., Vidailhet, M., Gresty, M., . . . Marsden, C. D. (1993). Assessing tremor severity. J Neurol Neurosurg Psychiatry, 56(8), 868-873.

124

Birklein, F., Eisenbarth, G., Erbguth, F., & Winterholler, M. (2003). Botulinum toxin type B blocks sudomotor function effectively: a 6 month follow up. J Invest Dermatol, 121(6), 1312-1316. doi:10.1046/j.1523-1747.2003.12620.x Blakemore, S. J., & Sirigu, A. (2003). Action prediction in the cerebellum and in the parietal lobe. Exp Brain Res, 153(2), 239-245. doi:10.1007/s00221-003-1597-z Blood, A. J., Tuch, D. S., Makris, N., Makhlouf, M. L., Sudarsky, L. R., & Sharma, N. (2006). White matter abnormalities in dystonia normalize after botulinum toxin treatment. Neuroreport, 17(12), 1251-1255. doi:10.1097/01.wnr.0000230500.03330.01 Boonstra, F., Florescu, G., Evans, A., Steward, C., Mitchell, P., Desmond, P., . . . van der Walt, A. (2017). Tremor in multiple sclerosis is associated with cerebello-thalamic pathology. J Neural Transm (Vienna), 124(12), 1509-1514. doi:10.1007/s00702-017-1798-4 Boonstra, F. M. C., Noffs, G., Perera, T., Jokubaitis, V. G., Vogel, A. P., Moffat, B. A., . . . Kolbe, S. C. (2018). Functional neuroplasticity in response to cerebello-thalamic injury underpins the clinical presentation of tremor in multiple sclerosis. Manuscript sumbitted for publication. Boonstra, F. M. C., Perera, T., Noffs, G., Marotta, C., Vogel, A. P., Evans, A. H., . . . Kolbe, S. C. (2018). Novel Functional MRI Task for Studying the Neural Correlates of Upper Limb Tremor. Frontiers in Neurology, 9(513). doi:10.3389/fneur.2018.00513 Brin, M. F., Lyons, K. E., Doucette, J., Adler, C. H., Caviness, J. N., Comella, C. L., . . . Koller, W. C. (2001). A randomized, double masked, controlled trial of botulinum toxin type A in essential hand tremor. Neurology, 56(11), 1523-1528. Clower, D. M., West, R. A., Lynch, J. C., & Strick, P. L. (2001). The inferior parietal lobule is the target of output from the superior colliculus, hippocampus, and cerebellum. J Neurosci, 21(16), 6283-6291. Currà, A., Trompetto, C., Abbruzzese, G., & Berardelli, A. (2004). Central effects of botulinum toxin type A: Evidence and supposition. Movement Disorders, 19(S8), S60-S64. doi:doi:10.1002/mds.20011 Delnooz, C. C., Pasman, J. W., Beckmann, C. F., & van de Warrenburg, B. P. (2013). Task-free functional MRI in cervical dystonia reveals multi-network changes that partially normalize with botulinum toxin. PLoS One, 8(5), e62877. doi:10.1371/journal.pone.0062877 Filippi, G. M., Errico, P., Santarelli, R., Bagolini, B., & Manni, E. (1993). Botulinum A toxin effects on rat jaw muscle spindles. Acta Otolaryngol, 113(3), 400-404.

125

Foley, N., Pereira, S., Salter, K., Fernandez, M. M., Speechley, M., Sequeira, K., . . . Teasell, R. (2013). Treatment with botulinum toxin improves upper-extremity function post stroke: a systematic review and meta-analysis. Arch Phys Med Rehabil, 94(5), 977-989. doi:10.1016/j.apmr.2012.12.006 Halsband, U., Ito, N., Tanji, J., & Freund, H. J. (1993). The role of premotor cortex and the supplementary motor area in the temporal control of movement in man. Brain, 116 ( Pt 1), 243-266. Kurtzke, J. F. (1983). Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology, 33(11), 1444-1452. Marvulli, R., Ianieri, G., Megna, G., Lancioni, G., Saggini, R., Ranieri, M., . . . Megna, M. (2012). Botulinum toxin type A in cervical dystonia. Int J Immunopathol Pharmacol, 25(1 Suppl), 23s-28s. Matthews, W. B. (1977). Aids to the examination of the peripheral nervous system. J Neurol Sci, 33(1), 299. doi:10.1016/0022-510X(77)90205-2 Opavsky, R., Hlustik, P., Otruba, P., & Kanovsky, P. (2011). Sensorimotor network in cervical dystonia and the effect of botulinum toxin treatment: a functional MRI study. J Neurol Sci, 306(1-2), 71-75. doi:10.1016/j.jns.2011.03.040 Perera, T., Lee, W. L., Yohanandan, S. A. C., Nguyen, A. L., Cruse, B., Boonstra, F. M. C., . . . van der Walt, A. (2018). Validation of a precision tremor measurement system for multiple sclerosis. J Neurosci Methods. doi:10.1016/j.jneumeth.2018.09.022 Rinker, J. R., Salter, A. R., Walker, H., Amara, A., Meador, W., & Cutter, G. R. (2015). Prevalence and characteristics of tremor in the NARCOMS multiple sclerosis registry: a cross-sectional survey. BMJ Open, 5(1). doi:10.1136/bmjopen-2014-006714 Rosales, R. L., Arimura, K., Takenaga, S., & Osame, M. (1996). Extrafusal and intrafusal muscle effects in experimental botulinum toxin-A injection. Muscle Nerve, 19(4), 488-496. doi:10.1002/(SICI)1097-4598(199604)19:4<488::AID-MUS9>3.0.CO;2-8 Safarpour, Y., Mousavi, T., & Jabbari, B. (2017). Botulinum Toxin Treatment in Multiple Sclerosis—a Review. Current Treatment Options in Neurology, 19(10), 33. doi:10.1007/s11940-017-0470-5 Samotus, O., Kumar, N., Rizek, P., & Jog, M. (2018). Botulinum Toxin Type A Injections as Monotherapy for Upper Limb Essential Tremor Using Kinematics. Can J Neurol Sci, 45(1), 11-22. doi:10.1017/cjn.2017.260

126

Schmidt, P., Gaser, C., Arsic, M., Buck, D., Forschler, A., Berthele, A., . . . Muhlau, M. (2012). An automated tool for detection of FLAIR-hyperintense white-matter lesions in Multiple Sclerosis. NeuroImage, 59(4), 3774-3783. doi:10.1016/j.neuroimage.2011.11.032 Shum, M., Shiller, D. M., Baum, S. R., & Gracco, V. L. (2011). Sensorimotor integration for speech motor learning involves the inferior parietal cortex. The European journal of neuroscience, 34(11), 1817-1822. doi:10.1111/j.1460-9568.2011.07889.x Smith, S. M., Jenkinson, M., Woolrich, M. W., Beckmann, C. F., Behrens, T. E., Johansen- Berg, H., . . . Matthews, P. M. (2004). Advances in functional and structural MR image analysis and implementation as FSL. NeuroImage, 23, S208-219. doi:10.1016/j.neuroimage.2004.07.051 Tighe, A. P., & Schiavo, G. (2013). Botulinum neurotoxins: mechanism of action. Toxicon, 67, 87-93. doi:10.1016/j.toxicon.2012.11.011 Trosch, R. M., & Pullman, S. L. (1994). Botulinum toxin a injections for the treatment of hand tremors. Movement Disorders, 9(6), 601-609. doi:doi:10.1002/mds.870090604 Van Der Walt, A., Sung, S., Spelman, T., Marriott, M., Kolbe, S., Mitchell, P., . . . Butzkueven, H. (2012). A double-blind, randomized, controlled study of botulinum toxin type A in MS-related tremor. Neurology, 79(1), 92-99. doi:10.1212/WNL.0b013e31825dcdd9 Walsh, R., & Hutchinson, M. (2007). Molding the sensory cortex: Spatial acuity improves after botulinum toxin treatment for cervical dystonia. Movement Disorders, 22(16), 2443-2446. doi:doi:10.1002/mds.21759

6.11. Supplementary tables

Supplementary Table 1. Volumetric and lesion load data for tremor patient that received placebo and BoNT-A. * indicates uncorrected significance p<0.05. Abbreviations: BoNT-A, botulinum toxin a; ICV, intracranial volume; sd, standard deviation; SCP, superior cerebellar peduncle. Placebo BoNT-A p-value ICV mean(sd) 1.25*106 (3.13*105) 1.28*106 (3.49*105) 0.788 Cerebellar volume mm3 mean(sd) Contralateral 53640 (15510) 56846 (16155) 0.543 Ipsilateral 53694 (15332) 57441 (15897) 0.458 Thalamic volume mm3 mean(sd)

127

Contralateral 6200 (686) 6442 (1085) 0.887 Ipsilateral 6045 (828) 6178 (961) 0.861 SCP area mm2 mean(sd) Contralateral 25.1 (3.69) 24.58 (5.35) 0.542 Ipsilateral 25.75 (3.43) 25.43 (6.12) 0.511 Total lesions mL mean(sd) 17420 (13707) 16152 (10759) 0.982 Cerebellar lesions mL mean(sd) Contralateral 24.67 (47.59) 53.62 (210.94) 0.546 Ipsilateral 44.14 (91.47) 5.38 (8.36) 0.046* Thalamic lesions mL mean(sd) Contralateral 29.48 (118.81) 12.86 (21.12) 0.707 Ipsilateral 10.29 (29.65) 6.38 (12.03) 0.797

128

Supplementary Table 2. Regression between baseline structural and functional measures and change in Bain scores. * indicates uncorrected significance p<0.05. Abbreviations: β, standardized coefficient; SCP, superior cerebellar peduncle; IPL, inferior parietal lobule; SMC, supplementary motor cortex Bain severity Bain handwriting Bain Archimedes D 6 Week D 12 Week D 6 Week D 12 Week D 6 Week D 12

Week Cerebellar volume β=-0.066 β=-0.096 β=0.287 β=0.381 Β=0.297 β=0.324 Contralateral p=0.792 p=0.699 p=0.248 p=0.112 p=0.216 p=0.185 β=-0.038 β=-0.096 β=0.183 β=0.352 β=0.284 β=0.302 Ipsilateral p=0.880 p=0.698 p=0.468 p=0.145 p=0.238 p=0.219 Thalamic volume β=0.038 β=-0.138 β=0.006 β=0.203 β=0.059 β=0.083 Contralateral p=0.881 p=0.578 p=0.981 p=0.412 p=0.811 p=0.743 β=-0.066 β=-0.246 β=0.103 β=0.207 β=0.107 β=0.197 Ipsilateral p=0.794 p=0.317 p=0.686 p=0.402 p=0.663 p=0.430 SCP area β=-0.036 β=-0.392 β=-0.282 β=-0.159 β=-0.054 β=-0.199 Contralateral p=0.886 p=0.100 p=0.257 p=0.522 p=0.827 p=0.423 β=-0.185 β=-0.346 β=-0.417 β=-0.186 β=-0.133 β=-0.174 Ipsilateral p=0.461 p=0.153 p=0.087 p=0.455 p=0.589 p=0.489 β=-0.327 β=0.089 β=-0.165 β=-0.004 β=-0.276 β=0.059 Total lesions p=0.185 p=0.723 p=0.517 p=0.987 p=0.257 p=0.817 Cerebellar lesions β=0.020 β=-0.011 β=-0.201 β=-0.085 β=-0.053 β=-0.031 Contralateral p=0.938 p=0.966 p=0.435 p=0.740 p=0.834 p=0.904 β=-515 β=-0.274 β=-0.583 β=-0.353 β=0.050 β=0.057 Ipsilateral p=0.032* p=0.273 p=0.014* p=0.154 p=0.844 p=0.824 Thalamic lesions β=-0.371 β=-0.186 β=-0.379 β=-0.245 β=-0.079 β=0.062 Contralateral p=0.131 p=0.457 p=0.126 p=0.324 p=0.749 p=0.809 β=-0.082 β=-0.252 β=0.170 β=0.352 β=0.182 β=0.124 Ipsilateral p=0.747 p=0.308 p=0.504 p=0.149 p=0.461 p=0.625

129

β=-0.216 β=-0.379 β=0.117 β=0.022 β=0.108 β=-0.265 IPL p=0.404 p=0.133 p=0.658 p=0.933 p=0.675 p=0.307 β=-0.081 β=-0.149 β=0.208 β=0.054 β=0.193 β=-0.036 Premotor/SMC p=0.757 p=0.572 p=0.432 p=0.838 p=0.454 p=0.892

130 7

General Discussion

Our understanding of the structural and functional neuroimaging characteristics of the pathophysiology of MS-related tremor remains incomplete and, furthermore, there is currently no effective non-invasive, long-term treatment option to help patients with tremor. Previous pathophysiology and treatment studies in MS tremor are generally underpowered, limited by design faults and none examine the central mechanisms of effect of potential treatments (Koch, Mostert, Heersema, & De Keyser, 2007). Focussed and extensive studies are needed to determine both the pathophysiology and explore new treatments for MS tremor.

This chapter summarises the main results and conclusions of chapters 3-5 of this thesis; integrates these findings into the literature; highlights the clinical and scientific relevance of the findings and finally, discusses the caveats of the methods used and suggests future studies that could be performed to investigate questions raised by these results.

7.1. Summary of findings Chapter 3 tested the hypothesis, based on clinical observations, treatment studies, and animal model studies (Alusi, Worthington, Glickman, & Bain, 2001; Bittar et al., 2005; Burrows, 1937), that pathology in the cerebello-thalamic tract is involved in the pathophysiology of MS-related tremor. We showed that tremor severity significantly correlated with the degree of atrophy of both the thalamus and the SCP, and lesion load within the contralateral thalamus, in MS patients with unilateral tremor. These findings indicate that the pathophysiology of tremor in MS extends beyond just cerebellar pathology 131

as was historically thought, and includes inflammation and atrophy of other structures along the cerebello-thalamo tract. The results of this study provided a strong hypothesis for testing in chapter 5.

In chapter 4, we introduced a novel fMRI task that aimed to produce tremulous movement for in vivo characterisation of functional activations associated with tremor. The fMRI task involved performance of a video game with an MRI compatible joystick, with tasks to control for visual tracking and basic manual motor function. We demonstrated moderate to excellent reproducibility in fMRI activation while performing the task between baseline to 6 weeks in healthy controls. Furthermore, there was significantly more tremulous movement (joystick movements in the frequency band 3-10 Hz) while playing the game in MS patients with tremor compared to healthy controls. Together, these results show that the task can evoke tremor in a controlled manner in the scanner and is reproducible enough to examine treatment effects in longitudinal studies. This chapter provides the methodology needed for the following chapters to examine in vivo tremor pathophysiology and examining changes over time as a result of BoNT-A intervention.

Chapter 5 aimed to determine the functional tremor network in MS. We applied the previously validated joystick fMRI and measured atrophy and lesion volumes in regions of interest within the cerebello-thalamic tract. We found increased atrophy and inflammation in the cerebellum and thalamus in MS patients with tremor compared to MS patients without tremor. Furthermore, we found increased ipsilateral activation within the IPL and the premotor/SMC in MS tremor patients compared to MS patients without tremor. Maximum activation within these areas negatively correlated with the tremor severity. Specifically, patients with a milder tremor showed higher activation, and patients with a more severe tremor show less activation. These findings suggest that functional neuroplasticity within the ipsilateral IPL and premotor cortex might provide compensatory processing that ameliorates tremor severity in MS patients.

In chapter 6, we examined the effect of BoNT-A on fMRI activity within the ipsilateral IPL and premotor cortex regions identified in chapter 5. MS patients that received intramuscular injection of BoNT-A showed a mild drop in tremor severity at 6 weeks after injection that became more pronounced at 12 weeks after injection. The patients also showed decreased activation within the ipsilateral IPL at 6 weeks after injection.

132

Interestingly, patients with a stronger reduction of tremor severity after BoNT-A injections showed a stronger reduction in activation, indicating that better response to treatment results in less need for compensatory functional activations. Furthermore, we found no MRI predictors of response to treatment. Together, these results indicate that BoNT-A treatment is associated with functional CNS changes that are likely to be the indirect result of changes in peripheral input.

7.2. Integration into the literature We found that the clinical presentation of tremor in MS is dependent on both structural and functional features of the disease, with functional changes acting to maintain function in the presences of degenerative structural changes. The possibility for functional neuroplasticity could be due to inflammation, which is an important characteristic of MS pathology (Amedei, Prisco, & D’Elios, 2012). Inflammation increases the excitability of neurons (Vezzani & Viviani, 2015), which could facilitate the reorganization of a functional networks involved in the tremor pathophysiology. Furthermore, this isn’t the first study to identify the brain’s ability to using abnormal activation to compensate for symptoms. In Parkinson’s disease, an increased activation within the ipsilateral premotor cortex correlates with better motor performance (Mallol et al., 2007). Furthermore, people after stroke that show increased amounts of ipsilateral activation have better motor recovery (Dodd, Nair, & Prabhakaran, 2017). In MS there are also a few studies examining the effect of brain activity on motor performance; however, whether these changes are compensatory (Rocca et al., 2010; Rocca et al., 2005; Rocca et al., 2002; Rocca et al., 2004) or maladaptive (Reddy et al., 2002) remains elusive. Direction of effect is thought to depend on lateralization, strength of connectivity, breadth or size of the activation (Peterson & Fling, 2018). This thesis indicates that functional activation has a compensatory effect; however, long-term studies that capture progression of tremor are needed to see how the exact relationship between functional activation and the clinical presentation of tremor.

Another interesting finding of this thesis is the effect of BoNT-A on the tremor network; specifically, with a reduction of the functional activation within the ipsilateral IPL which is associated with sensorimotor integration. There are a few studies examining the effect of BoNT-A on functional networks in other neurological disorders. Studies in cervical and focal dystonia showed that patients have normalization of the functional networks associated with the pathology (Byrnes et al., 1998; Delnooz, Pasman, Beckmann, & van de

133

Warrenburg, 2013; Thickbroom, Byrnes, Stell, & Mastaglia, 2003). Interestingly, Delnooz et al. (Delnooz et al., 2013) showed that similarly to MS tremor patients, BoNT-A treatment in dystonia patients effects the connectivity within the sensorimotor network. BoNT acts peripherally and effects by blockages of muscle contraction reducing muscle spindle afferents input (Urban & Rolke, 2004). Muscle spindles convey information to the CNS, including the cerebellum and the cerebral cortex (Stein, 2017); therefore, after intramuscular injections of BoNT-A, the altered peripheral input to the CNS could lead to the observed reorganisation (Currà, Trompetto, Abbruzzese, & Berardelli, 2004). The neurophysiological models of MS tremor have highlighted the role sensorimotor system; specifically, of the cerebello-thalamo-cortical loops (Liu et al., 1997) and a malfunctioning internal model in the cerebellum (Miall, Weir, Wolpert, & Stein, 1993). The nature of the peripheral input to the CNS and the neurophysiological models in MS tremor could explain why reorganisation occurs in regions of the sensorimotor system after BoNT-A treatment. To conclude, BoNT-A exert peripheral and central mechanisms; however, the central effects are thought to be indirect and facilitated by peripheral feedback.

7.3. Contributions to research and clinical treatment and management We introduced a novel joystick fMRI task that elicits tremor in MS patients. Previously used fMRI task to study upper limb tremor are inadequate to fully capture the complexity of tremor (Bucher, Seelos, Dodel, Reiser, & Oertel, 1997; Buijink et al., 2015; Buma, Raemaekers, Kwakkel, & Ramsey, 2015; Lewis et al., 2011; Lewis et al., 2007; Manganotti et al., 2010). Our joystick task is dynamic and provides adequate cognitive loading necessary to elicit tremulous movement in MS patients. Tremor originating anywhere in the upper limb will mechanically couple through the wrist joint and be measurable with our joystick. We made the joystick task accessible online, to allow other researches to use this task in their studies. This will ultimately provide more comparable findings and strengthen our research. Interestingly, this task is not limited to MS-related tremor. As mentioned in the background chapter of this thesis, there are different types of tremor and this task is suitable for any kinetic/action tremors of the upper limb. It would be of great interest to examine the differences and similarities in pathophysiology of tremor resulting from different pathological backgrounds using our joystick. To conclude, our joystick could benefit future studies in MS tremor or tremor associated with other neurological disorders.

134

This thesis is the first to comprehensively characterise the structural and functional aspects of the tremor network in MS patients. Getting a clear picture of the pathophysiology of tremor in MS is important for both science and clinic. First, there is the clear benefit of knowledge for its own sake. Most importantly, the structural pathology and functional neuroplasticity, provide targets for the development of treatments. One needs to understand the pathology before one can treat it. Furthermore, by examining the effects of drugs on the tremor network one might be able to identify treatment sensitive patients which is valuable information in the clinic. Finally, in addition to treatment, a adequate monitoring tool for tremor is currently lacking. The tremor network, specifically the functional aspects, provide a potential in vivo tool to monitor central effect during progression of tremor severity. To conclude, this thesis characterised the tremor network in MS which in will ultimately help MS patients deal with their tremor.

7.4. Limitations and Future directions The effect of BoNT-A beyond its known pharmacodynamics profile has been demonstrated in studies of sudomotor function, bladder function and spasticity where persistent benefit for up to 6 months is well described (Birklein, Eisenbarth, Erbguth, & Winterholler, 2003; Safarpour, Mousavi, & Jabbari, 2017). Our findings in chapter 6, indicate ability of BoNT-A to exert central effects; however, they are unable to confirm whether these changes are related to the prolonged effect of BoNT-A. To confirm the connection between the functional neuroplasticity and the duration of effect, a longitudinal imaging study including imaging at multiple time points over at least 6 months needs to be performed. This would provide essential information in understanding how BoNT-A exerts it effect and if there are any long-term central effects. Overall, longer-term studies are needed to identify all the effects associated with BoNT-A treatment.

Different chapters produce potentially contrasting results. Firstly, in chapter 3 we highlight that the volume of the SCP correlates with the tremor severity; however, in chapter 6 we don’t find a difference in SCP area between the MS patients with tremor compared to without tremor. This discrepancy might be due to the difference in cohort, as chapter 3 described a pilot study with 11 tremor subjects and chapter 5 is a larger study with 42 tremor subjects. The finding in chapter 3 was promising; however, as we weren’t able to replicate this in a larger sample it might have been a type I error. Secondly, in chapter 6 we showed a correlation between IPL activation and improvement following BoNT-A

135

treatment. However, the use of the atlases to name coordinates can cause confusion as labels span over distinct regions. The IPL is a large area with several sub-regions (Caspers et al., 2013). The IPL identified in chapter 4 is IPL-PF. The chapter 5 IPL cluster falls in the IPL-PGa, which is an area much more posterior and clearly distinct to the IPL-PF. Due to the different settings in which they are identified, they are likely different in functionality.

Chapter 1 describes the different types of tremor, and there are many more movement disorders that include high frequencies (Mehanna & Jankovic, 2013). Our joystick task analyses hyperkinetic movements within a frequency of 3-10Hz, that are exacerbated by targeted/visually guided movement (Alusi et al., 2001). Cerebellar ataxia is another hyperkinetic movement present in MS and has shown to be within that specific frequency band (Morales-Briceño, Fois, & Fung, 2018; Wilkins, 2017). Visual feedback is important in cerebellar ataxia, and the complexity of the task could result in worsening of cerebellar ataxia. To help specify the movement we assessed and prevent confusion, a spectral analysis examining the actually tremor peaks could have helped differentiate between these two types of movement. This was not performed in this thesis but would be of interest for future studies.

In addition to the pathophysiology of MS tremor, the current assessment of tremor in the clinic is limited by subjective rating scales. One of the sections in the first chapter of this thesis describes the potential of new advanced paraclinical assessment techniques of tremor, including our own TREMBAL device (Perera et al., 2018; Perera et al., 2016; Perera, Yohanandan, McKay, & McDermott, 2015). This technique is more sensitive, objective and quantifiable and should be considered for future pathophysiology studies of MS tremor (Perera et al., 2018). Increased sensitivity could potentially distinguish differences in the tremor network depending on tremor severity. Furthermore, treatment effect studies would greatly benefit from more quantifiable, objective and sensitive tools. An adequate tool would nullify the need for blinded ratings and randomisation, simplifying studies and providing bigger samples in the treatment group. Overall, there is a pressing need to implement the more advanced assessment techniques, like TREMBAL, to improve research.

Work presented in this thesis completes part of the imaging for the TremTox study. All participants underwent and extensive MRI not all of which sequences have been presented,

136

including diffusion-weighted imaging, resting-state fMRI and speech task fMRI. This provides a novel and deeply phenotyped dataset that provides a unique opportunity examine microstructural and connectivity changes associated with MS-tremor that can then be compared against the findings in this thesis. Studies in ET show altered resting-state patterns along the cerebello-thalamo-cortical tract and microstructural damage in the cerebellum (Wang et al., 2018; Yin, Lin, Li, Qian, & Mou, 2016), which indicates the potential that these techniques could have to further strengthen our current understanding of the tremor pathophysiology. Speech production and analysis in MS is an evolving and new research and the structural and functional imaging collected during my PhD will provide a basis for contributing to another cerebellar feature of MS (dysarthria). Taken together, much more is to be gained for the rich data set collected for the TremTox study.

Finally, the main aim of the TremTox study is to determine the efficacy of BoNT-A for the treatment of tremor in MS. The current thesis provides promising preliminary results, with an improvement in the tremor scores in the analysed subset of the total cohort. Furthermore, a previous smaller double-blind randomised, controlled study confirmed BoNT-A’s potential to effectively treat tremor (Van Der Walt et al., 2012). The efficacy study combined with the results from this thesis will generate class I evidence that is required to translate treatment of MS tremor with BoNT-A into clinical practice. Ultimately, we hope this will improve the quality of life for MS patients with tremor.

7.5. References Alusi, S. H., Worthington, J., Glickman, S., & Bain, P. G. (2001). A study of tremor in multiple sclerosis. Brain, 124(Pt 4), 720-730. Amedei, A., Prisco, D., & D’Elios, M. M. (2012). Multiple Sclerosis: The Role of Cytokines in Pathogenesis and in Therapies. International Journal of Molecular Sciences, 13(10), 13438-13460. doi:10.3390/ijms131013438 Birklein, F., Eisenbarth, G., Erbguth, F., & Winterholler, M. (2003). Botulinum toxin type B blocks sudomotor function effectively: a 6 month follow up. J Invest Dermatol, 121(6), 1312-1316. doi:10.1046/j.1523-1747.2003.12620.x Bittar, R. G., Hyam, J., Nandi, D., Wang, S., Liu, X., Joint, C., . . . Aziz, T. Z. (2005). Thalamotomy versus thalamic stimulation for multiple sclerosis tremor. J Clin Neurosci, 12(6), 638-642. doi:10.1016/j.jocn.2004.09.008

137

Bucher, S. F., Seelos, K. C., Dodel, R. C., Reiser, M., & Oertel, W. H. (1997). Activation mapping in essential tremor with functional magnetic resonance imaging. Annals of Neurology, 41(1), 32-40. doi:10.1002/ana.410410108 Buijink, A. W., Broersma, M., van der Stouwe, A. M., van Wingen, G. A., Groot, P. F., Speelman, J. D., . . . van Rootselaar, A. F. (2015). Rhythmic finger tapping reveals cerebellar dysfunction in essential tremor. Parkinsonism Relat Disord, 21(4), 383-388. doi:10.1016/j.parkreldis.2015.02.003 Buma, F. E., Raemaekers, M., Kwakkel, G., & Ramsey, N. F. (2015). Brain Function and Upper Limb Outcome in Stroke: A Cross-Sectional fMRI Study. PLoS One, 10(10), e0139746. doi:10.1371/journal.pone.0139746 Burrows, T. E. (1937). The Effects of Lesions of the Superior Cerebellar Peduncle in the Macacus rhesus Monkey. (Bull. Neur. Inst. New York, vol. v, p. 165, Aug., 1936.) Ferraro, A., and Barrera, S. E. Journal of Mental Science, 83(342), 115-115. doi:10.1192/bjp.83.342.115 Byrnes, M. L., Thickbroom, G. W., Wilson, S. A., Sacco, P., Shipman, J. M., Stell, R., & Mastaglia, F. L. (1998). The corticomotor representation of upper limb muscles in writer's cramp and changes following botulinum toxin injection. Brain, 121 ( Pt 5), 977-988. Caspers, S., Schleicher, A., Bacha-Trams, M., Palomero-Gallagher, N., Amunts, K., & Zilles, K. (2013). Organization of the human inferior parietal lobule based on receptor architectonics. Cereb Cortex, 23(3), 615-628. doi:10.1093/cercor/bhs048 Currà, A., Trompetto, C., Abbruzzese, G., & Berardelli, A. (2004). Central effects of botulinum toxin type A: Evidence and supposition. Movement Disorders, 19(S8), S60- S64. doi:doi:10.1002/mds.20011 Delnooz, C. C., Pasman, J. W., Beckmann, C. F., & van de Warrenburg, B. P. (2013). Task-free functional MRI in cervical dystonia reveals multi-network changes that partially normalize with botulinum toxin. PLoS One, 8(5), e62877. doi:10.1371/journal.pone.0062877 Dodd, K. C., Nair, V. A., & Prabhakaran, V. (2017). Role of the Contralesional vs. Ipsilesional Hemisphere in Stroke Recovery. Front Hum Neurosci, 11, 469. doi:10.3389/fnhum.2017.00469 Koch, M., Mostert, J., Heersema, D., & De Keyser, J. (2007). Tremor in multiple sclerosis. J Neurol, 254(2), 133-145. doi:10.1007/s00415-006-0296-7 [doi]

138

Lewis, M. M., Du, G., Sen, S., Kawaguchi, A., Truong, Y., Lee, S., . . . Huang, X. (2011). Differential involvement of striato- and cerebello-thalamo-cortical pathways in tremor-and akinetic/rigid-predominant Parkinson's disease. Neuroscience, 177, 230- 239. doi:10.1016/j.neuroscience.2010.12.060 Lewis, M. M., Slagle, C. G., Smith, A. B., Truong, Y., Bai, P., McKeown, M. J., . . . Huang, X. (2007). Task specific influences of Parkinson's disease on the striato- thalamo-cortical and cerebello-thalamo-cortical motor circuitries. Neuroscience, 147(1), 224-235. doi:10.1016/j.neuroscience.2007.04.006 Liu, X., Miall, C., Aziz, T. Z., Palace, J. A., Haggard, P. N., & Stein, J. F. (1997). Analysis of action tremor and impaired control of movement velocity in multiple sclerosis during visually guided wrist-tracking tasks. Mov Disord, 12(6), 992-999. doi:10.1002/mds.870120624 Mallol, R., Barrós-Loscertales, A., López, M., Belloch, V., Parcet, M. A., & Ávila, C. (2007). Compensatory cortical mechanisms in Parkinson's disease evidenced with fMRI during the performance of pre-learned sequential movements. Brain Res, 1147, 265-271. doi:https://doi.org/10.1016/j.brainres.2007.02.046 Manganotti, P., Acler, M., Formaggio, E., Avesani, M., Milanese, F., Baraldo, A., . . . Fiaschi, A. (2010). Changes in cerebral activity after decreased upper-limb hypertonus: an EMG-fMRI study. Magn Reson Imaging, 28(5), 646-652. doi:10.1016/j.mri.2009.12.023 Mehanna, R., & Jankovic, J. (2013). Movement disorders in multiple sclerosis and other demyelinating diseases. J Neurol Sci, 328(1-2), 1-8. doi:10.1016/j.jns.2013.02.007 Miall, R. C., Weir, D. J., Wolpert, D. M., & Stein, J. F. (1993). Is the cerebellum a smith predictor? J Mot Behav, 25(3), 203-216. doi:10.1080/00222895.1993.9942050 Morales-Briceño, H., Fois, A. F., & Fung, V. S. C. (2018). Chapter 18 - Tremor. In B. L. Day & S. R. Lord (Eds.), Handb Clin Neurol (Vol. 159, pp. 283-301): Elsevier. Perera, T., Lee, W. L., Yohanandan, S. A. C., Nguyen, A. L., Cruse, B., Boonstra, F. M. C., . . . van der Walt, A. (2018). Validation of a precision tremor measurement system for multiple sclerosis. J Neurosci Methods. doi:10.1016/j.jneumeth.2018.09.022 Perera, T., Yohanandan, S. A., Thevathasan, W., Jones, M., Peppard, R., Evans, A. H., . . . McDermott, H. J. (2016). Clinical validation of a precision electromagnetic tremor measurement system in participants receiving deep brain stimulation for essential tremor. Physiol Meas, 37(9), 1516-1527. doi:10.1088/0967-3334/37/9/1516

139

Perera, T., Yohanandan, S. A. C., McKay, C. M., & McDermott, H. J. (2015). Evaluation of an objective tremor assessment system for movement disorders research [online]. Paper presented at the Australian Biomedical Engineering Conference (ABEC 2015), Barton, ACT. https://search.informit.com.au/documentSummary;dn=730379337328993;res=IE LENG Peterson, D. S., & Fling, B. W. (2018). How changes in brain activity and connectivity are associated with motor performance in people with MS. NeuroImage: Clinical, 17, 153- 162. doi:https://doi.org/10.1016/j.nicl.2017.09.019 Reddy, H., Narayanan, S., Woolrich, M., Mitsumori, T., Lapierre, Y., Arnold, D. L., & Matthews, P. M. (2002). Functional brain reorganization for hand movement in patients with multiple sclerosis: defining distinct effects of injury and disability. Brain, 125(Pt 12), 2646-2657. Rocca, M. A., Ceccarelli, A., Rodegher, M., Misci, P., Riccitelli, G., Falini, A., . . . Filippi, M. (2010). Preserved brain adaptive properties in patients with benign multiple sclerosis. Neurology, 74(2), 142-149. doi:10.1212/WNL.0b013e3181c91a00 Rocca, M. A., Colombo, B., Falini, A., Ghezzi, A., Martinelli, V., Scotti, G., . . . Filippi, M. (2005). Cortical adaptation in patients with MS: a cross-sectional functional MRI study of disease phenotypes. The Lancet Neurology, 4(10), 618-626. doi:https://doi.org/10.1016/S1474-4422(05)70171-X Rocca, M. A., Falini, A., Colombo, B., Scotti, G., Comi, G., & Filippi, M. (2002). Adaptive functional changes in the cerebral cortex of patients with nondisabling multiple sclerosis correlate with the extent of brain structural damage. Annals of Neurology, 51(3), 330-339. doi:doi:10.1002/ana.10120 Rocca, M. A., Gallo, A., Colombo, B., Falini, A., Scotti, G., Comi, G., & Filippi, M. (2004). Pyramidal tract lesions and movement-associated cortical recruitment in patients with MS. NeuroImage, 23(1), 141-147. doi:10.1016/j.neuroimage.2004.05.005 Safarpour, Y., Mousavi, T., & Jabbari, B. (2017). Botulinum Toxin Treatment in Multiple Sclerosis—a Review. Current Treatment Options in Neurology, 19(10), 33. doi:10.1007/s11940-017-0470-5 Stein, J. (2017). Sensorimotor Control. In Reference Module in Neuroscience and Biobehavioral Psychology: Elsevier.

140

Thickbroom, G. W., Byrnes, M. L., Stell, R., & Mastaglia, F. L. (2003). Reversible reorganisation of the motor cortical representation of the hand in cervical dystonia. Mov Disord, 18(4), 395-402. doi:10.1002/mds.10383 Urban, P. P., & Rolke, R. (2004). Effects of botulinum toxin type A on vibration induced facilitation of motor evoked potentials in spasmodic torticollis. J Neurol Neurosurg Psychiatry, 75(11), 1541-1546. doi:10.1136/jnnp.2003.029215 Van Der Walt, A., Sung, S., Spelman, T., Marriott, M., Kolbe, S., Mitchell, P., . . . Butzkueven, H. (2012). A double-blind, randomized, controlled study of botulinum toxin type A in MS-related tremor. Neurology, 79(1), 92-99. doi:10.1212/WNL.0b013e31825dcdd9 Vezzani, A., & Viviani, B. (2015). Neuromodulatory properties of inflammatory cytokines and their impact on neuronal excitability. Neuropharmacology, 96(Pt A), 70-82. doi:10.1016/j.neuropharm.2014.10.027 Wang, P., Luo, X., Zhong, C., Yang, L., Guo, F., & Yu, N. (2018). Resting state fMRI reveals the altered synchronization of BOLD signals in essential tremor. J Neurol Sci, 392, 69-76. doi:10.1016/j.jns.2018.07.008 Wilkins, A. (2017). Cerebellar Dysfunction in Multiple Sclerosis. Frontiers in Neurology, 8, 312-312. doi:10.3389/fneur.2017.00312 Yin, W., Lin, W., Li, W., Qian, S., & Mou, X. (2016). Resting State fMRI Demonstrates a Disturbance of the Cerebello-Cortical Circuit in Essential Tremor. Brain Topogr, 29(3), 412-418. doi:10.1007/s10548-016-0474-6

141

Conclusion

Upper limb tremor in MS is a debilitating and embarrassing symptom. This thesis is the first to provide extensive the insight into the pathophysiology, that is needed to increase understanding and, subsequently, support the development of treatments. We found that the tremor network consists of structural and functional parameters. Specifically, atrophy and inflammation along the cerebello-thalamo-cortical tract is associated with tremor in MS. Furthermore, functional neuroplasticity, within motor planning and sensorimotor integration areas, is found to alleviate tremor severity. Together, both aspects of the tremor network impact the clinical presentation of tremor in MS patients.

The characterisation of the tremor network is important for the understanding and development of treatments. We found that intramuscular BoNT-A treatment affects the tremor network in MS; specifically, the functional neuroplasticity within the sensorimotor integration area. BoNT-A reduces tremor severity, and concurrently reduces functional neuroplasticity in the sensorimotor integration area. This suggests that due to the local tremor alleviating effect of BoNT-A, there is less need for the tremor alleviating activation of the tremor network. Taken together, the tremor network is instrumental in understanding and treating tremor in MS.

142

Appendices

2.1. Appendix 1. Bain Tremor Rating Score:

Rate Relevant Upper Limb RIGHT UL LEFT UL

143

Handwriting – affected limb

RUL _____ LUL _____

Archimedes Spiral Drawing – affected limb

RUL _____ LUL _____

144

2.2. Appendix 2. Bain sheet:

145

Minerva Access is the Institutional Repository of The University of Melbourne

Author/s: Boonstra, Frederique Maria Christina

Title: Tremor in multiple sclerosis: neuroimaging perspective

Date: 2019

Persistent Link: http://hdl.handle.net/11343/230596

File Description: Final thesis file

Terms and Conditions: Terms and Conditions: Copyright in works deposited in Minerva Access is retained by the copyright owner. The work may not be altered without permission from the copyright owner. Readers may only download, print and save electronic copies of whole works for their own personal non-commercial use. Any use that exceeds these limits requires permission from the copyright owner. Attribution is essential when quoting or paraphrasing from these works.