Parkinsonism & Related Disorders

@medicinejournal

Volume 46

January 2018

Editors-in-ChiEf

Vincenzo Bonifati Department of Clinical Genetics Erasmus University Medical Center Wytemaweg 80 Rotterdam, 3015 CN The Netherlands
E-mail: v.bonifati@erasmusmc.nl

Hubert H. Fernandez, MD
Center for Neurological Restoration
Cleveland Clinic Neurological Institute
9500 Euclid Avenue, S-3, Cleveland, Ohio, USA 44195 Tel: (216) 445 1108

assoCiatE Editors

Europe:

Vincenzo Bonifati

Department of Clinical Genetics Erasmus Medical Center,
Room EE-914
Dr. Molewaterplein 50, Rotterdam 3015 GE

The Netherlands
E-mail: v.bonifati@erasmusmc.nl

Managing Editor

Sue Calne C.M.

Editorial Office, 537 Bluff Place Kamloops, BC V2C 1S5, Canada E-mail: scalne@mail.ubc.ca

Editorial Board

F. Alarcon (Ecuador)
R. Alcalay (USA)
A. Antonini (Italy)
R. Bhidayasiri (Thailand) R.E. Burke (USA)

R Camicioli (Canada)
W.P. Cheshire (USA)
M. Coelho (Portugal)
M.F. Contarino (Netherlands) A. D’Abreu (USA)

S. K. Das (India)
M. Della Coletta (Brazil) A. Espay (USA)
G. Fabbrini (Italy)

Founding Editor
Donald B. Calne OC-FRSC

1995-2008

Middle East and Africa:

Jonathan Carr

Head Division of Neurology Tygerberg Hospital & University of Stellenbosch
Tygerberg 7505,
South Africa
E-mail: jcarr@sun.ac.za

S. Factor (USA)
A. Fasano (Canada)
S. Fujioka (Japan)
W. R. Galpern (USA)
D. Goldstein (USA)
A. Hassan (USA)
M. Heckman (USA)
T. Ikeuchi (Japan)
K. Jellinger (Austria)
P. Jenner (UK)
K.A. Josephs (USA)
J. M. Kim (South Korea) D. Koziorowski (Poland) E. C. Lai (USA)

formEr Editors-in-ChiEf

Fax: (216) 636 2989 E-mail: FERNANH@ccf.org

North and South America:

Robert L. Rodnitzky

Department of Neurology University of Iowa Hospital 200 Hawkins Drive
Iowa City, IA 52242, USA E-mail: robert-rodnitzky@uiowa.edu

M.F. Lew (USA)
S-Y. Lim (Malaysia)
T. Mestre (Canada)
H. Mizusawa (Japan)
M. J. Nirenberg (USA)
R. Pickering (UK)
A. Puschmann (Sweden)
J.F. Quinn (USA)
N. Quinn (UK)
A. Rajput (Canada)
S. Reich (USA)
I. Rektorova (Czech Republic) O. Riess (Germany)
M Rudzinska (Poland)

Asia and Oceania:

Eng-King Tan

Department of Neurology, Singapore General Hospital National Neuroscience Institute Outram road

Singapore 169608
Singapore Email:tan.eng.king@sgh.com.sg

S. A. Schneider (Germany) A. Schrag (UK)
L.M. Shulman (USA)
J. Sławek (Poland)

D. G. Standaert (USA)
A. J. Stoessl (Canada)
C. Tanner (USA)
D. Torres-Russotto (USA) D.D. Truong (USA)

E. Uc (USA)
A. Videnovic (USA) Y.R. Wu (Taiwan)
R. Yadav (India)
D. Zielonka (Poland)

Ronald F. Pfeiffer and Zbigniew K. Wszolek

2008-2017

Author inquiries: You can track your submitted article at http://www.elsevier.com/track-submission. You can track your accepted article at http://www.elsevier.com/trackarticle. You are also welcome to contact Customer Support via http://support.elsevier.com.

Publication information: Parkinsonism & Related Disorders (ISSN 1353-8020). For 2018, volumes 46-57 are scheduled for publication. Subscription prices are available upon request from the Publisher or from the Customer Service Department nearest you or from this journal’s website (http://www.elsevier.com/locate/parkreldis). Further information is available on this journal and other Elsevier products through Elsevier’s website: (http//www.elsevier.com). Subscriptions are accepted on a prepaid basis only and are entered on a calendar year basis. Issues are sent by standard mail (surface within Europe, air delivery outside Europe). Priority rates are available upon request. Claims for missing issues should be made within six months of the date of dispatch.

For a full and complete Guide to Authors please refer to http://www.elsevier.com/journal/parkreldis Printed by Henry Ling, The Dorset Press, Dorchester, UK

Parkinsonism and Related Disorders 46 (2018) 1

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Editorial
So long, and thanks for all the fish

Just as the Dolphins left a message as they departed Planet Earth in Douglas Adams’ fourth book of his Hitchhikers Guide to the Gal- axy “trilogy,” it seems appropriate that we should also leave a departing message as we make our exit as Co-Editors of Parkin- sonism and Related Disorders after 10 years at the helm. So, here we go.

Ika Natassa has written that “time flies, but not memories” and, to echo Martin Luther, this is most certainly true. It seems like a long time ago that two inexperienced and untested neurologists agreed to assume the role as Co-Editors of Parkinsonism and Related Disorders, not really knowing or understanding just what type of excellent adventure (thank you, Bill and Ted) we were in for. How- ever, we had the very good fortune of inheriting our amazing Man- aging Editor, Susan Calne, and our experienced and capable Publisher, Peter Bakker, to guide us on our journey. We then made what almost certainly was our most important and vital de- cision and accomplishment when we succeeded in assembling a group of Associate Editors e Vincenzo Bonifati, Jonathan Carr, Rob- ert Rodnitzky, and E.K. Tan e who have remained with us during our entire 10-year tenure and have been the most skilled and dedi- cated editorial colleagues that we could have ever envisioned. Memories of our time and work with these talented individuals will remain with us forever and our thanks to them is endless.

During the past 10 years Parkinsonism and Related Disorders has been blessed with significant growth, both in size and in stature. It has received e and continues to receive e strong support from both its publisher, Elsevier, and its sponsoring organization, the Interna- tional Association for Parkinsonism and Related Disorders (IAPRD). We thank both organizations for such wonderful support.

Those readers who are familiar with Hitchhiker’s Guide to the Galaxy will have recognized that the title we have given to this

Farewell Editorial, is wildly inappropriate for what we see as the future for Parkinsonism and Related Disorders. The Dolphins left their message as they were departing Planet Earth just prior to its destruction to make way for a hyperspace bypass. The future for Parkinsonism and Related Disorders is infinitely brighter. With the assumption of the editorial reins by two uniquely qualified co- editors, Vincenzo Bonifati and Hubert Fernandez, we are leaving the Journal in the hands of two highly skilled and accomplished physicians, scientists, and communicators who, we have no doubt, will lead Parkinsonism and Related Disorders to even greater heights and successes, with new ideas, new approaches, and new life.

Our thanks also must go out to the absolutely wonderful group of reviewers who have given selflessly of their time to provide the excellent reviews that have been so instrumental for the growth in prestige of Parkinsonism and Related Disorders during these past 10 years. Of course, we would be completely remiss in not also thank- ing you, our authors and readers, for your support of the Journal during these 10 years.

We are most grateful for the times and the memories that these past 10 years have afforded us as Co-Editors. It has been an immensely enjoyable and educational journey for us both. And we cannot resist leaving you with one (and perhaps one too many!) final “time” quote, this from Groucho Marx: “Time flies like an arrow; fruit flies like a banana.”

Zbigniew K. Wszolek, M.D.*, Ronald F. Pfeiffer, M.D.

* Corresponding author. E-mail address: wszolek.zbigniew@mayo.edu (Z.K. Wszolek).

https://doi.org/10.1016/j.parkreldis.2017.12.027

Parkinsonism and Related Disorders 46 (2018) 2

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Editorial
The beginning of our new journey

Dear Colleagues,

It is with great pride and humility that we jointly announce our acceptance for the role of co-Editors-in-Chief of Parkinsonism & Related Disorders, starting with this edition!

During the past decade, our journal has grown tremendously, under the masterful and dedicated shared leadership of Zbigniew Wszolek and Ronald Pfeiffer. The quantity and quality of the manuscripts submitted to Parkinsonism & Related Disorders have exponentially increased, and so has its Impact Factor, from 1.90 in 2008 to 4.48 in 2016. We congratulate our past Editors-in-Chief and their entire team of Associate Editors, and our beloved Manag- ing Editor, Susan Calne, for their unparalleled achievements. We draw inspiration from your legacy, and we remain indebted to all of you for handing over to us a journal in such a fine standing.

We do realize, as we take over the leadership of the journal, that we, indeed, have big shoes to fill. Nonetheless, we intend to leverage all the gains made by our predecessors to try to further consolidate the position of Parkinsonism & Related Disorders as one of the best journals in our field, and to elevate it to even greater heights. We will certainly give it our very best, and we hope that our comple- mentary expertise, with one Co-Editor entrenched in the commu- nity of basic and translational sciences, and the other at the epicenter of the clinical trials and clinical sciences network, will bring added value in this endeavor.

More importantly, we also realize that the success of a scientific journal is the result of an amazing team work, where the steering role and vision of the Editors-in-Chief are intricately intertwined with the excellence of the Associate Editors, the invaluable dedica- tion of Managing Editors, and, last but not least, of our entire panel of thoughtful ad-hoc Reviewers. We therefore thank all of you: your dedication and support will be the key to the continued success of our journal.

We also wish to thank the International Association for Parkin- sonism and Related Disorders (IAPRD), and the Publisher, Peter Bak- ker and his entire team at Elsevier, for their trust and unwavering support. We look forward to a pleasant and fruitful collaboration and are excited to showcase the continued technological advances in publishing, that Elsevier is known for, within our very own jour- nal. As an example, this year our journal will move to article-based publishing. Article-based publishing will make paginated, final and citable articles available online much faster, without having to wait for an issue to be compiled.

https://doi.org/10.1016/j.parkreldis.2017.12.028

These are truly exciting times for those who work in the field of parkinsonism and related movement disorders. The spectacular recent advances in the basic sciences, such as genetics, gene editing, molecular neurobiology, stem cells biology and cell reprogram- ming, as well as those in the functional and molecular neuroimag- ing, are driving a true revolution in our understanding of the etiology, pathogenesis and pathophysiology of several movement disorders. As a result, disease definitions and boundaries are being redefined. Non-motor features are increasingly appreciated as relevant part of several movement disorders. Biomarkers of disease state and progression are being actively sought. The recognition of pre-motor and prodromal stages of diseases will allow an earlier diagnosis, so that future disease-modifying measures can be deliv- ered when they are crucially needed.

In summary, we genuinely believe that we are at the dawn of a new era of translation of novel molecular insights into therapeutic strategies. At the same time, the development of better therapies for our patients, particularly those in advanced disease states, either pharmacological or surgical, represent relevant areas of unmet needs that warrant our continued growth and understand- ing. We envision Parkinsonism & Related Disorders to be a reflection of this scientific revolution, and to continue to serve the scientific community as a central hub for scientific debate and a major vehicle in the dissemination of new, relevant research on all aspects of parkinsonism and related movement disorders. We also intend to continue the tradition of soliciting comprehensive educational reviews, points of view articles, and editorials on all relevant areas of parkinsonism and movement disorders.

We hope that Parkinsonism & Related Disorders will continue to meet the high standards that our large and expanding international readership expects and deserves.

We are therefore looking forward to receiving your manuscripts!

Vincenzo Bonifati* Department of Clinical Genetics, Erasmus University Medical Center, Rotterdam, The Netherlands

Hubert H. Fernandez

Cleveland Clinic Neurological Institute, Cleveland, OH, USA

* Corresponding author. E-mail address: v.bonifati@erasmusmc.nl (V. Bonifati).

Parkinsonism and Related Disorders 46 (2018) 3e8

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Review article
More than ataxia e Movement disorders in ataxia-telangiectasia

He lio Afonso Ghizoni Teive a, *, Carlos Henrique Ferreira Camargo b, Renato Puppi Munhoz c

a Movement Disorders Unit, Neurology Service, Internal Medicine Department, Hospital de Clínicas, Federal University of Parana , Curitiba, Brazil b Neurology Service, Hospital Universita rio, State University of Ponta Grossa, Ponta Grossa, Brazil
c Movement Disorders Centre, Toronto Western Hospital, Toronto University, Toronto, ON, Canada

articleinfo

Article history:

Received 8 September 2017 Received in revised form
5 December 2017
Accepted 11 December 2017

Keywords:

Movement disorders Ataxia Ataxia-telangiectasia

1. Introduction

Ataxia-telangiectasia (AT), or ATM syndrome, previously known as Louis-Bar or Boder-Sedgwick syndrome, is a rare neurodegen- erative disease that represents the second most common auto- somal recessive ataxia after Friedreich’s ataxia [1e8]. With the exception of consanguineous populations, individuals of all races and ethnicities are affected equally by AT. The prevalence is esti- mated to be < 1e9/100,000, although incidences as high as 1 in 40,000 and as low as approximately 1 in 300,000 have been re- ported [9,10].

AT is characterized by progressive neurological dysfunction with multisystem involvement and cancer predisposition [1e5]. Symp- toms typically start in early childhood and classical neurological signs include progressive cerebellar ataxia, oculomotor apraxia, chorea and cognitive dysfunction [1e5]. Multisystem involvement includes immunodeficiency that commonly manifests as recurrent

* Corresponding author. Rua General Carneiro 1103/102, Centro, Curitiba, PR 80060-150, Brazil.

E-mail addresses: hagteive@mps.com.br (H.A.G. Teive), chcamargo@uol.com.br (C.H.F. Camargo), renatopuppi@yahoo.com (R.P. Munhoz).

https://doi.org/10.1016/j.parkreldis.2017.12.009

abstract

Ataxia-telangiectasia (AT) is a rare autosomal recessive neurodegenerative disease caused by mutations in the ATM gene with progressive neurological dysfunction, multisystem abnormalities and cancer predisposition. Classically, AT is associated with cerebellar ataxia, oculocutaneous telangiectasia and oculomotor apraxia. The aim of this review is to describe the movement disorders observed in patients with AT. Movement disorders in AT patients described in the literature are reviewed. The selected articles were analyzed with a focus on clinical presentation, presence of movement disorders, and atypical cases or variants of the syndrome. In AT patients, particularly adults, chorea and dystonia are the most common movement disorders, besides cerebellar ataxia. Myoclonus, tremor and parkinsonism have been described less frequently in patients with AT. Archetypal findings, such as oculocutaneous abnormalities may not be uniformly present. AT can present with different movement disorders, in isolation or com- bined, with or without cerebellar ataxia or oculocutaneous telangiectasias. Neurologists with expertise in movement disorders should be aware of AT when investigating patients with movement disorders of unknown etiology.

sinopulmonary infections. endocrinopathy, radiosensitivity, chro- mosomal instability, and oculocutaneous telangiectasia (“spider veins”, small dilated blood vessels near the surface of the skin or mucous membranes) [1e5,11e16]. AT is also associated with a high incidence of cancer, predominantly leukemia and lymphoma [1e5,15e18]. Laboratory findings include elevated serum alpha- fetoprotein (AFP) levels and immunoglobulin deficiency (IgA and IgG) [1e3,18e20]. AT is caused by biallelic mutations in the ATM (ataxia telangiectasia mutated) gene on chromosome 11q22.3e23.1, which encodes the ATM protein, a 350-kDa serine- threonine kinase belonging to the phosphoinositide 3-kinase- related protein kinase family. ATM is involved in several cellular functions, including cell cycle checkpoint control, apoptosis, DNA damage response, oxidative stress and mitochondrial metabolism [1e3,21e24]. After the ATM gene was identified, several studies described atypical clinical pictures of the condition, or AT variants, with or without cerebellar ataxia, ocular or cutaneous telangiec- tasias in genetically proven cases of AT, suggesting that the classical designation of the disorder may be inappropriate [1e3,25e31]. Moreover, cases of AT variants include unique and heterogeneous clinical phenotypes, such as the absence of mild neurological manifestations, predisposition to breast cancer (monoallelic ATM

4 H.A.G. Teive et al. / Parkinsonism and Related Disorders 46 (2018) 3e8

mutation carriers), peripheral neuropathy, spinal cord atrophy and ocular abnormalities other than oculomotor apraxia (saccade ab- normalities, absence of smooth pursuit, optokinetic nystagmus and periodic alternating nystagmus) with no cerebellar ataxia or ocu- locutaneous telangiectasia [1e3,6,12,14,17,24e36]. Finally, in addi- tion to cerebellar ataxia, different movement disorders have been described in patients with AT, including chorea, dystonia, myoc- lonus, postural, rest and kinetic tremor, and parkinsonism [1,2,5,26,37e51].

2. Methods

A literature search was performed on Pubmed in December 2016, for English-language articles in the time period from 1992 to 2016, using the terms “ataxia telangiectasia”, “ATM” and “move- ment disorders”, “parkinsonism”, “chorea”, “dystonia”, “tremor”, “myoclonus”. A total of 176 references were initially found and selected using the following main exclusion criteria: (a) duplicate articles; (b) articles non-related to the purpose of the study; (c) articles non-related to ataxia telangiectasia; (d) articles not including sufficient relevant data. A total of 68 references were found to be appropriate for the purpose of this review. To these references, we added important historical articles published in the period between 1926 and 1991 (10 references), and two book chapters that emphasized former clinical descriptions of AT and relevant cases series. These selected articles were reviewed with a focus on clinical presentation, presence of movement disorders, and atypical cases or variants of the syndrome.

2.1. Movement disorders in AT patients – pathophysiology

AT is one of several DNA repair disorders which results in neurodegeneration. In the brain, the ATM kinase is involved not only with DNA damage repair, but in neurons, where a substantial fraction of the ATM protein is cytoplasmic (rather than only nu- clear) and involved with vesicle trafficking, implicating it in neuronal activity. A more objective definition of this role could be provided by assessing possible neuronal activity changes in asymptomatic heterozygous carriers of ATM mutations [52].

A series of cellular mechanisms have been proposed to explain neurodeneration in AT: (a) defective DNA damage response or repair; (b) defective response to oxidative stress; (c) mitochondrial dysfunction; (d) defects in neuronal function (e.g., failed cell cycle regulation, synaptic/vesicular dysregulation and altered epige- netics); (d) defects in brain vasculature; (e) altered brain turnover [1].

Overall, the most disabling symptoms of AT result from pro- gressive cerebellar degeneration, characterized by the gradual loss and/or aberrant location of Purkinje cells and, to a lesser extent, functional damage of granule cells. However, the dilemma of basal ganglia dysfunction occurring along with consistent cerebellar- focused pathology remains unexplained [52]. Eilam et al. [53] compared dopaminergic areas and pathways of ATM-deficient or wild-type control mice using tyrosine hydroxylase (TH) immuno- stained brain sections finding a marked decrease (up to 75%) in the number of TH-positive neurons in the SNc. Significant reductions were also seen in the dopaminergic parts of the ventral tegmental area and efferents of the SNc in the striatum. Interestingly, ATM- related findings were selective for dopaminergic innervation. On the other hand, another study that performed direct analysis of ventral mesencephalon, striatum and cerebellum in a cohort of ATM-knockout mice did not find a reduction in neither the major neurotransmitter monoamines (including dopamine) nor NADþ, which is a cofactor in dopamine biosynthesis via TH [54]. Despite this conflicting body of evidence and the existence of anecdotal AT

patients with levodopa responsive dystonia and parkinsonism, the role of dopaminergic pathways on the pathophysiology of AT is questionable [39,44,45]. For instance, none of the five AT patients studiedbyMe neretetal.[44]with123I-FP-CITSPECTimaging showed signs of dopaminergic denervation, including one with levodopa responsive parkinsonism. Differently, Koepp et al. [45] reported a 6-year-old girl with AT, severe progressive dystonia, and decreased tracer uptake in the striatum bilaterally on brain 123I-FP-CIT SPECT imaging.

Finally, except for single case reports, functional brain imaging studies that might provide insight into how cerebellar degenera- tion influences activity in projection regions and associated cere- bellar networks have not been performed [52]. Recently, a fluorodeoxyglucose labeled positron emission tomography (FDG- PET) study in AT cases showed uniformly reduced glucose meta- bolism in the cerebellum and increased metabolism in the globus pallidus. These findings correlated with the level motor disability, suggesting a physiopathologic relationship [52]. More robust confirmatory studies may consolidate the role and direct relation- ship between the basal ganglia, the cerebellum and the genesis of movement disorders in AT patients.

2.2. Movement disorders in AT patients e clinic overview

The presence of movement disorders other than ataxia are well recognized in cases of AT [1e8]. For instance, Boder and Sedgwick [55] assessed the clinical features of 101 cases of AT finding cere- bellar ataxia in all cases, but also choreoathetosis (91%) and ocu- lomotor apraxia (84%). In fact, among the most common pediatric presentation of AT, the presence of these two features seems ubiquitous [1e5]. This phenotype was indeed noted in the first description of AT from 1926 by Syllaba and Henner [56] who re- ported three teenage siblings with progressive choreoathetosis and ocular telangiectasia. Similar cases were published later by Wells and Shy [57] in two sisters. Among adults, a wider range of atypical cases of AT has been described, including different movement disorders, such as dystonia, myoclonus, chorea, parkinsonism and tremor [1,37e51]. In 2014, Me neret et al. [44] described pleiotropic movement disorders in a series of 14 adult AT patients in a wide spectrum of movement disorders, the most common being dysto- nia (86%) and subcortical myoclonus (86%) followed by tremor (43%). At least three different movement disorders were present in 86% of them. These authors emphasize that adults with unex- plained movement disorders should be investigated for AT and elevated AFP can be a useful clue for the diagnosis [44]. Recently, Person [58] reviewed hyperkinetic movement disorders in child- hood autosomal recessive ataxia syndromes confirming that they may occur in the majority of patients with AT, including dystonia, or dystonia with myoclonus, with predominant upper limb and cer- vical involvement, in addition to generalized chorea (video with three cases).

Supplementary video related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.12.009.

From a molecular standpoint, no area of the ATM gene is espe- cially susceptible to mutations, which have been identified in its proximal, central and distal regions [1] (Fig. 1). Therefore, not sur- prisingly, genotype/phenotype correlation in these cases is com- plex and differences in ATM genotype partially but not fully account for clinical variability. Verhagen et al. [32] studied clinical and the respective molecular findings of 51 genetically proven AT patients, concluding that the presence of ATM protein and residual kinase activity correlated with the phenotype. Disability correlated posi- tively with truncated mutations that resulted in total absence of the ATM kinase activity, whereas patients with milder phenotypes had a missense or splice site mutation resulting in expression of ATM

withresidualkinaseactivity.Me neretetal.[44]resultsarein accordance with these data, as 79% of their patientsdcharacterized overall by slower progression or delayed onset as well as later confinement to wheelchair and longer survivaldhad at least one missense mutation, compared to 36% of the typical patients with AT.

Overall, most of the literature points out that differences in ATM genotyping do not fully explain the phenotypic variability, but evaluating both protein function and the level of residual ATM protein can be helpful in variant forms of AT [44]. Table 2 sum- marizes the most common movement disorders published in pa- tients with AT (case reports and series) before the ATM gene mapping. Table 3 shows movement disorders published in patients with AT with genetically proven AT.

2.3. Cerebellar ataxia in AT

Cerebellar ataxia is found in association with dysarthria, mild chorea, cognitive dysfunction and oculomotor abnormalities, such as oculomotor apraxia and nystagmus [1e8]. Brain magnetic reso- nance imaging shows diffuse cerebellar atrophy, predominantly in the vermis and also in cerebellar hemispheres, and there is evi- dence of loss of Purkinje and granular cells on neuropathological examination [1e3,51,59]. In general, patients are wheelchair-bound by the age of 10 years and die around the age of 20 years [1e6,59]. Historically, Louis-Bar [60] published in 1941 the first description of cerebellar ataxia, associated with cutaneous telangiectasia, in a 9- year-old Belgian boy with AT. In 1958, Centerwall and Miller [61]

Table 1

Movement disorders in patients with AT e Historical cases published.

described two cases of cerebellar ataxia in AT. In 1963, Boder and Sedgwick [55] published a large case series of 101 patients with AT and 100% of them had cerebellar ataxia. In the case reports pub- lished by Syballa and Henner [56] in 1926 (3 patients) and Wells and Shy [57] in 1957 (2 patients) with AT, none had cerebellar ataxia, only choreoathetosis (Table 1).

To date, there are no epidemiological studies investigating the real prevalence of cerebellar ataxia in atypical or variant adult cases of genetically proven AT.

2.4. Chorea in AT patients

In the first case series of AT patients published in the last cen- tury, which included predominantly cases with childhood onset, choreadthen referred to as choreoathetosisdwas high prevalent [1e5,56,58,62e65]. In the study by Boder and Sedgwick, for example, 91% of the 101 AT cases had this condition [55]. In the literature on AT, chorea has traditionally been defined as a feature almost typical as cerebellar ataxia and oculocutaneous telangiec- tasias [1e5,60e66]. Recently, however, the frequency of chorea in AT patients, particularly adult patients, has been debated. Verhagen et al. [67] studied retrospectively the clinical spectrum of AT in adulthood (nine families, thirteen patients) and found that gener- alized chorea was present in 70% of cases. In contrast, Me neret et al. [44] failed to find chorea in their study with fourteen consecutive adult patients. A peculiar case report was published by Klein et al. [41] in 1996 of an AT case with ataxia and no telangiectasia masquerading as benign hereditary chorea.

61 (91%) (choreoathetosis) Ataxia 1⁄4 Cerebellar ataxia; Choreoathetosis 1⁄4 a mixture of chorea and athetosis, nowadays dystonia).

H.A.G. Teive et al. / Parkinsonism and Related Disorders 46 (2018) 3e8 5

Fig. 1. Schematic illustration of the ATM gene and ATM Protein.
A e An illustration of the gene and respective main domains correspondent exons (The size of the exons and the distance between them are not indicative of the sizes/distances in the gene/protein). B – The protein with important areas such as substrate binding domains.

Authors

Syballa & Henner 1926 [56] Louis-Bar 1941 [60]
Wells & Shy 1957 [57] Centerwall & Miller 1958 [61] Boder & Sedwick 1963 [55]

Total of cases

3
1
2
2 101

Movement Disorders (cases)

Ataxia

Chorea

3 (choreoathetosis)

2 (choreoathetosis)

Dystonia

Myoclonus Tremor Parkinsonism

2
101 (100%)

3 (choreoathetosis) – e – 2 (choreoathetosis) – e – 61 (91%) – (choreoathetosis)

– – – – – – – – – –

6 H.A.G. Teive et al. / Parkinsonism and Related Disorders 46 (2018) 3e8

Table 2

Movement Disorders in patients with AT e Case reports and case series before ATM gene mapping.

Authors

Bodensteiner et al., 1980 [49] Taylor et al., 1987 [24] Huang et al., 1991 [42] Churchyard et al., 1991 [26] Woods & Taylor 1992 [69] Koepp et al., 1994 [45]

De Graff et al., 1995 [72] Klein et al., 1996 [41] Yanofsky et al., 2009 Carrilo et al., 2009 [68]

Age in years.

Table 3

Age of onset of symptoms

before 10 childhood 1
e

2
7e10
9 and 12 months 2.5
15

Total of cases

Movement Disorders (cases)

1 2 1 1 70 1 4 2 3 1

Ataxia

1 (mild) 2
e
1 (mild) 70

e
e
2 (in the follow up) 1
e

Chorea

e e 68 e e 2 e e

Dystonia

1 (severe)
2
1 (torticollis) 1
55
1
e
e
2
1

Myoclonus

1 18

e e 1 e 11 2 4 e 1

Myoclonus

e e e e e e 2 e e e

Tremor

rest 1⁄4 9 e

Tremor Parkinsonism

– – – – – – – – – – – – – – – – – – – –

Parkinsonism

– –

Movement Disorders in patients with genetically proved AT.

Authors

Goyal & Behari 2002 [48] Verhagen et al., 2012 [32]

Saunders-Pulmann et al., 2012 [38] Shaikh et al.*, 2013 [40] Charlesworth et al., 2013 [39] Nissenkorn et al., 2013

Cummins et al., 2013 [50] Meissner et al., 2013 [47] Me neret et al., 2014 [44] Termsarasab et al., 2015 [43] van Egmond et al., 2015 [37] Kuhm et al., 2015 [46] Nakayama et al., 2015 [51]

Age of onset of symptoms

12
5 yo 1⁄4 24

!6 yo 1⁄4 10 12 (1-20) 12.1 (5-34) 13 (11-15) 11.2 ± 3.9 2

6 (1-14) 2.5 and 5 8e21 childhood 5

Total of cases

Movement Disorders (cases) Ataxia Chorea Dystonia

1 e 34 34

13 0 80 80 3 e 17 þ 1 1 4 e 14 10 2 2 4 e 1 e 1 1

e 1 24 21

e 13
e 1
e DRD 1⁄4 3 þ e
e 1
e 4
0 12
e e
e e
e 1
e 1

– action 1⁄4 79 rest 1⁄4 48 – e –

e þ e – dystonic 1⁄4 4 – 6 1 e – e – e – e –

* 1⁄4 Disorders of upper limb movements in AT; DRD 1⁄4 Dopamine Responsive Dystonia; variant/atypical forms of AT. Age in years (mean ± standard deviation).

2.5. Dystonia in AT patients

Dystonia has been described in several case reports or case se- ries of patients with AT. The types of dystonia described include focal (cervical, oromandibular), segmental (frequently upper-limb and craniocervical), and generalized dystonia [2,38e40,42,44e49,68]. Saunders-Pullman et al. [38] studied twenty Canadian Mennonite adult patients with primary dystonia and screened for mutations in the DTY1, DTY6 and ATM genes. Their results are particularly interesting as the genetic analysis revealed the presence of a mutation in the ATM gene in 13 members of three of the families that did not have DYT1 or DYT6 mutations. These patients had no cerebellar ataxia or ocular telangiectasias, no cerebellar atrophy on brain MRI, and the phenotype was similar to that of DYT6 patients, with dystonia more prominent in the cervical, cranial and brachial areas. The presence of myoclonus-dystonia in two patients was also remarkable [38]. Woods and Taylor [69] evaluated 70 AT cases (2e42 years of age) in the British Isles in 1992 and showed that dystonia was present in 55 patients. Shaikh et al. [40] published a study in 2013 evaluating upper-limb move- ment disorders in 80 patients with genetically proven AT. They found dystonia in isolation and in association with myoclonus and/ or tremor. In the case series of adult patients described by Me neret et al. [44], dystonia was observed in 12 out of 14 patients. It was predominantly focal or segmental (cervical and upper-limb) in eight and multifocal (mild) in six. Some patients presented with both myoclonus and dystonia [44]. Another remarkable case report was published in 2013 by Charlesworth et al. [39] describing a patient with genetically proven AT manifesting as dopa-responsive

þ/- 1⁄4 Presence or absence, without number of cases; Classic/Variant 1⁄4 Classic and

cervical dystonia. This atypical presentation of AT was later emphasized by Wijemanne and Jankovic [70] in 2015. Interestingly, the initial publications on AT, including the seminal papers by Syllaba and Henner [56] (1926, three patients), Wells and Shy [57] (1957, two patients) and Boder and Sedgwick [55,64] (1958 – eight patients, and 1963 – 101 patients), described patients with pro- gressive choreoathetosis (ranging from 85 to 91% in these series). These authors suggested that AT was in fact a “basal ganglia dis- order” and choreoathetosis was particularly more common in older children, often masking features of cerebellar ataxia [9,55,57]. Taylor et al. [24] also described variant forms of AT in 1987, including two patients with choreoathetosis and dystonia in addition to ataxia. All of these authors used the term chor- eoathetosis. Nowadays, however, athetosis and dystonia are considered to refer to the same condition, and dystonia was prob- ably therefore underestimated in case series of AT patients pub- lished in the 20th century. Dystonia currently represents the most common movement disorder in different case series of AT patients, with or without confirmatory genetic molecular tests [2,38,40,44e49,66,68].

2.6. Myoclonus in AT patients

Myoclonus can be observed in AT patients either in isolation or in combination with dystonia, which can be segmental, multifocal or generalized. In general, myoclonus is often of subcortical origin and not stimulus-sensitive [2,37,38,40,43,44,50,51,58,67,71,72]. In 2012, Saunders-Pullman et al. [38] described eight patients with myoclonus and jerky dystonia, predominantly in the neck and

upper limbs. Myoclonic head jerks, associated with axial dystonia, were described in 2013, by Cummins et al. [50] in the variant form of AT. In the Me neret’s case series, myoclonus was found in 12 out of 14 patients, and in 10 of these patients dystonia was also present, while two of them had isolated myoclonus [44]. Termsarasab et al. [43] reported two AT patients in whom myoclonus was the pre- dominant presenting feature, both with prominent axial and appendicular myoclonic jerks. Van Egmond et al. [37] described a case of AT among a series of four patients with molecularly defined childhood-onset neurogenetic disorders (aged 8e21 years) associ- ated with myoclonus. In this case the initial signs were the invol- untary jerky movements physiologically classified as cortical myoclonus at age 3, added by ataxia on follow up. In 2015, Nakayama et al. [51] published a case report of another atypical AT case presenting myoclonic axial jerks associated with dystonia and cerebellar ataxia.

2.7. Rare movement disorders in AT patients

Parkinsonism and different forms of tremor have also been rarely described in AT patients [2,44]. In the Me neret’s et al. [44] case series one out of 14 patients presented with levodopa responsive parkinsonism. Cases of tremor reported in the literature are also infrequent and include the full spectrum of rest, postural or kinetic predominantly upper extremities tremors [2e5] In the above mentioned Me neret’s et al. [44] study, tremor was observed in 43% of patients (six patients). Three patients had slow cerebellar action tremor, one had mixed rest and action tremor and two pa- tients had dystonic tremor, associated with segmental dystonia. On the other hand, Shaikh et al. [40] studied upper extremity oscilla- tions in patients with AT quantitatively using a three-axis acceler- ometer. Their recordings showed features that the authors classified as tremors in 79 out of a total of 80 cases studied. The most common subtype, resting tremor, has been found in 61% of the patients.

2.8. Movement disorders in AT patients – treatment

Several lines of rehabilitation therapy as well as exercise may be helpful reducing the burden of disability and loss of functionality but have no interference in the process of neurodegeneration [1].

In terms of pharmacological management, Nissenkorn et al. [71] reported a trial of amantadine in 17 AT children finding effective- ness for chorea in 32.5%, and ataxia in 25.3% of patients. There are no other specific studies for the treatment of movement disorders in AT. Conventional treatments, such as botulinum toxin for focal dystonia, should be used until more evidence-based and specific studies for AT can be performed. All drugs should be prescribed by a neurologist familiar with the assessment and treatment of in- dividuals with movement disorders [1].

Currently, no specific treatment is able to slow or stop the progression of the neurologic deficits in AT. However, a number of clinical trials are underway including novel approaches such as a phase III study of intra-erythrocyte dexamethasone sodium phos- phate IV infusion therapy, two studies using cell-based approached with pluripotent stem cells to modify the natural history of AT, in addition to studies to prevent systemic problems such as cancer and immunological changes [73].

3. Conclusion

AT or ATM syndrome represents a multisystem entity with pleomorphic neurological and systemic manifestations. Along with the classical clinical picture of cerebellar ataxia, oculocutaneous telangiectasia and oculomotor apraxia, this syndrome can present

with different movement disorders such as dystonia, myoclonus, chorea, parkinsonism and tremor in isolation or combined, with or without cerebellar ataxia or telangiectasias [2,36,38,40,44,66]. Movement disorders experts should be aware of AT when investi- gating movement disorders of unknown etiology [2,44,66]. Elevated AFP serum levels can be a useful diagnostic clue [1e3,5,19,20,44]. Further studies with large numbers of genetically proven AT patients are needed to elucidate the frequency and phenomenology of movement disorders in children and particu- larly adults.

Funding

None.

Financial disclosure

None.

Conflicts of interest
The authors report no conflict of interest.

Ethics statement

Not applicable for this review.

References

[1] C. Rothblum-Oviatt, J. Wright, M.A. Lefton-Greif, S.A. McGrath-Morrow, T.O. Crawford, H.M. Lederman, Ataxia telangiectasia: a review, Orphanet J Rare Dis 11 (2016) 159.

[2] H.A.G. Teive, A. Moro, M. Moscovich, W.O. Arruda, R.P. Munhoz, S. Raskin, T. Ashizawa, Ataxia-telangiectasia e a historical review and a proposal for a new designation: ATM syndrome, J. Neurol. Sci. 355 (2015) 3e6.

[3] L. Bott, C. Thumerelle, J.C. Cuvellier, A. Deschildre, L. Valle e, A. Sardet, Ataxia- telangiectasia: a review, Arch. Pediatr. 13 (2006) 293e298.

[4] M. Moin, A. Aghamohammadi, A. Kouhi, et al., Ataxia-telangiectasia in Iran: clinical and laboratory features of 104 patients, Pediatr. Neurol. 37 (2007) 21e28.

[5] A. Nissenkorn, B. Bem-Zeev, Ataxia telangiectasia, Handb Clin Neurol 132 (2015) 199e214.

[6] H.A.G. Teive, T. Ashizawa, Primary and secondary ataxias, Curr. Opin. Neurol. 28 (2015) 413e422.

[7] S. Vermeer, B.P.C. van de Warrenburg, M.A.A.P. Willemsen, et al., Autosomal recessive cerebellar ataxias: the current state of affairs, J. Med. Genet. 48 (2011) 651e659.

[8] M. Anheim, C. Tranchant, M. Koenig, The autosomal recessive cerebellar ataxias, N. Engl. J. Med. 366 (2012) 636e646.

[9] R.P. Sedgwick, E. Boder, Ataxia-telangiectasia, in: P.J. Vinken, G.W. Bruyn, H.L. Klawans (Eds.), Handbook of Clinical Neurology. 60. Hereditary Neurop- athies and Spinocerebellar Atrophies. Revised Series 16. De Jong JMBV, Elsevier Science Publishers, Amsterdam, 1991, pp. 347e423.

[10] M. Swift, et al., The incidence and gene frequency of ataxia-telangiectasia in the United States, Am. J. Hum. Genet. 39 (1986) 573e583.

[11] F. Suarez, N. Mahlaoui, D. Canioni, et al., Incidence, presentation, and prog- nosis of malignancies in ataxia-telangiectasia: report from the French Na- tional registry of primary immune deficiencies, J. Clin. Oncol. 33 (2015) 202e208.

[12] E. Byrne, J.F. Hallpike, J.I. Manson, G.R. Sutherland, Y.H. Thong, Ataxia- without-telangiectasia. Progressive multisystem degeneration with IgE defi- ciency and chromosomal instability, J. Neurol. Sci. 66 (1984) 307e317.

[13] E. Boder, R.P. Sedgwick, Ataxia-telangiectasia. (Clinical and immunological aspects), Psychiatr Neurol Med Psychol Beih 13-14 (1970) 8e16.

[14] K.L. Ying, W.E. Decoteau, Cytogenetic anomalies in a patient with ataxia, immune deficiency, and high alpha-fetoprotein in the presence of telangiec- tasia, Canc. Genet. Cytogenet. 4 (1981) 311e317.

[15] G. Lanzi, U. Balottin, D. Franciotta, E. Maserati, A. Ottolini, F. Pasquali, P. Veggiotti, Clinical, cytogenetic and immunological aspects in 4 cases resembling ataxia telangiectasia, Eur. Neurol. 32 (1992) 121e125.

[16] H.H. Chun, R.A. Gatti, Ataxia-telangiectasia, an evolving phenotype, DNA Repair 3 (2004) 1187e1196.

[17] L.L. Paglia, A. Lauge , J. Weber, et al., ATM germline mutations in women with familial breast cancer and a relative with haematological malignancy, Breast Canc. Res. Treat. 119 (2010) 443e452.

[18] A. Guleria, S. Chandna, ATM kinase: much more than a DNA damage

H.A.G. Teive et al. / Parkinsonism and Related Disorders 46 (2018) 3e8 7

8 H.A.G. Teive et al. / Parkinsonism and Related Disorders 46 (2018) 3e8

responsive protein, DNA Repair (15) (2015) 30122e30131, https://doi.org/

10.1016/jdnarep.2015.12.009. Dec 29. Pii: S1568-7864.

. [19] T.A. Waldmann, K.R. McIntire, Serum-alpha-fetoprotein levels in patients with
ataxia-telangiectasia, Lancet 2 (1972) 1112e1115.

. [20] J.H. Schieving, M. de Vries, J.M. van Vugt, et al., Alpha-fetoprotein, a fasci-
nating protein and biomarker in neurology, Eur. J. Paediatr. Neurol. 18 (2014)
243e248.

. [21] R.A. Gatti, I. Berkel, E. Boder, G. Braedt, P. Charmley, P. Concannon, et al.,
Localization of an ataxia-telangiectasia gene to chromosome 11q22-23, Na-
ture 336 (1988) 577e580.

. [22] K. Savitsky, A. Bar-Shira, S. Gilad, G. Rotman, Y. Ziv, L. Vanagaite, et al., A single
ataxia telangiectasia gene with a product similar to PI-3 kinase, Science 268
(1995) 1749e1753.

. [23] G.S. Stewart, J.I. Last, T. Stankovic, et al., Residual ataxia telangiectasia mutated
protein function in cells from ataxia telangiectasia patients, with 576ins137 and 7271T-G mutations, showing a less severe phenotype, J. Biol. Chem. 276 (2001) 30133e30141.

. [24] A.M. Taylor, E. Flude, B. Laher, M. Stacey, E. McKay, J. Watt, et al., Variant forms of ataxia-telangiectasia, J. Med. Genet. 24 (1987) 669e677.

. [25] E. Maserati, A. Ottolini, P. Veggiotti, G. Lanzi, F. Pasquali, Ataxia-without-tel- angiectasia in two sisters with rearrangements of chromosomes 7 and 14, Clin. Genet. 34 (1988) 283e287.

. [26] A. Churchyard, R. Stell, F.L. Mastaglia, Ataxia telangiectasia presenting as an extrapyramidal movement disorder and ocular motor apraxia without overt( telangiectasia, Clin. Exp. Neurol. 28 (1991) 90e96.

. [27] J.H. Friedman, A. Weitberg, Ataxia without telangiectasia, Mov. Disord. 8 (1993) 223e226.

. [28] P.J. Willems, B.C. Van Roy, W.J. Kleijer, M. Van der Kraan, J.J. Martin, Atypical clinical presentation of ataxia telangiectasia, Am. J. Med. Genet. 45 (1993) 777e782.

. [29] C.M. McConville, T. Stankovic, P.J. Byrd, G.M. McGuire, Q.Y. Yao, G.G. Lennox, A.M. Taylor, Mutations associated with variant phenotypes in ataxia-telangi- ectasia, Am. J. Hum. Genet. 59 (1996) 320e330.

. [30] G.G. Trimis, C.K. Athanassaki, M.M. Kanariou, A.A. Giannoulia-Karantana, Unusual absence of neurologic symptoms in a six-year old girl with ataxia- telangiectasia, J. Postgrad. Med. 50 (2004) 270e271.

. [31] N. Alterman, A. Fattal-Valevski, L. Moyal, et al., Ataxia-telangiectasia: mild neurological presentation despite null ATM mutation and severe cellular phenotype, Am J Med Genet A 143A (2007) 1827e1834.

. [32] M.M. Verhagen, J.I. Last, F.B. Hogervorst, et al., Presence of ATM protein and residual kinase activity correlates with the phenotype in ataxia-telangiectasia: a genotype-phenotype study, Hum. Mutat. 33 (2012) 561e571.

. [33] S. Gilad, L. Chessa, R. Khosravi, P. Russell, Y. Galanty, M. Piane, et al., Genotype- phenotype relationships in ataxia-telangiectasia and variants, Am. J. Hum. Genet. 62 (1998) 551e561.

. [34] R. Stell, A.M. Bronstein, G.T. Plant, A.E. Harding, Ataxia telangiectasia: a reappraisal of the ocular motor features and their value in the diagnosis of atypical cases, Mov. Disord. 4 (1989) 320e329.

. [35] F. Hoche, K. Seidel, M. Theis, et al., Neurodegeneration in ataxia telangiectasia: what is new? What is evident? Neuropediatrics 43 (2012) 119e129.

. [36] A.M. Taylor, Z. Lam, J.I. Last, P.J. Byrd, Ataxia telangiectasia: more variation at clinical and cellular levels, Clin. Genet. 87 (2015) 199e208.

. [37] M.E. van Egmond, J.W. Elting, A. Kuiper, et al., Myoclonus in childhood-onset neurogenetic disorders: the importance of early identification and treatment, Eur. J. Paediatr. Neurol. 19 (2015) 726e729.

. [38] R. Saunders-Pullman, D. Raymond, A.J. Stoessl, et al., Variant ataxia- telangiectasia presenting as primary-appearing dystonia in Canadian Men- nonites, Neurology 78 (2012) 649e657.

. [39] G. Charlesworth, M.D. Mohire, S.A. Schneider, M. Stamelou, N.W. Wood, K.P. Bhatia, Ataxia telangiectasia presenting as dopa-responsive cervical dystonia, Neurology 81 (2013) 1148e1151.

. [40] A.G. Shaikh, D.S. Zee, A.S. Mandir, H.M. Lederman, T.O. Crawford, Disorders of upper limb movements in ataxia-telangiectasia, PLos One 8 (2013), e67042.

. [41] C. Klein, G.K. Wenning, N.P. Quinn, C.D. Marsden, Ataxia without telangiec- tasia masquerading as benign hereditary chorea, Mov. Disord. 11 (1996) 217e220.

. [42] S.J. Huang, J.M. Chu, Y.T. Tsai, P.J. Wang, K.H. Hsieh, Ataxia-telangiectasia associated with torticollis: report of a case, Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi 32 (1991) 191e195.

. [43] P. Termsarasab, A.C. Yang, S. Frucht, Myoclonus in ataxia-telangiectasia, Tremor Other Hyperkinet Mov 5 (2015), https://doi.org/10.7916/D88P5Z9X.

. [44] A. Me neret, Y. Ahmar-Beaugendre, G. Rieunier, et al., The pleotropic move- ment disorders phenotype of adult ataxia-telangiectasia, Neurology 83 (2014) 1087e1095.

. [45] M. Koepp, L. Schelosky, I. Cordes, M. Cordes, W. Poewe, Dystonia in ataxia telangiectasia: report of a case with putaminal lesions and decreased striatal [123I]iodobenzamine binding, Mov. Disord. 9 (1994) 455e459.

. [46] C. Kuhm, C. Gallenmüller, T. Do€rk, M. Menzel, S. Biskup, T. Klopstock, Novel ATM mutation in a German patient presenting as generalized dystonia

without classical signs of ataxia-telangiectasia, J. Neurol. 262 (2015) 768e770. [47] W.G. Meissner, M. Fernet, J. Couturier, et al., Isolated generalized dystonia in biallelic missense mutations of the ATM gene, Mov. Disord. 28 (2013)

1897e1899.
[48] V. Goyal, M. Behari, Dystonia as presenting manifestation of ataxia telangi-

ectasia: a case report, Neurol. India 50 (2002) 187e189.
[49] J.B. Bodensteiner, R.M. Goldblum, A.S. Goldman, Progressive dystonia masking

ataxia in ataxia telangiectasia, Arch. Neurol. 37 (1980) 464e465.
[50] G. Cummins, T. Jawad, M. Taylor, T. Lynch, Myoclonic head jerks and extensor axial dystonia in the variant form of ataxia telangiectasia, Park. Relat. Disord.

19 (2013) 1173e1174.
[51] T. Nakayama, Y. Sato, M. Uematsu, et al., Myoclonic axial jerks for diagnosing

atypical evolution of ataxia telangiectasia, Brain Dev. 37 (2015) 362e365. [52] N.D. Volkow, D. Tomasi, G.J. Wang, Y. Studentsova, B. Margus, T.O. Crawford, Brain glucose metabolism in adults with ataxia-telangiectasia and their asymptomatic relatives, Brain 137 (6) (2014) 1753e1761. http://doi.org/10.

1093/brain/awu092.
[53] R. Eilam, Y. Peter, A. Elson, et al., Selective loss of dopaminergic nigro-striatal

neurons in brains of Atm-deficient mice, Proceedings of the National Academy

of Sciences of the United States of America 95 (1998) 12653e12656.
[54] H.T. Mount, J.C. Martel, P. Fluit, Y. Wu, E. Gallo-Hendrikx, C. Cosi, M.R. Marien, Progressive sensorimotor impairment is not associated with reduced dopa- mine and high energy phosphate donors in a model of ataxia-telangiectasia,

J. Neurochem. 88 (2004) 1449e1454.
[55] E. Boder, R.P. Sedgwick, Ataxia-telangiectasia. A review of 101 cases, in:

G. Walsh (Ed.), Little Club Clinics in Develop, Heinemann Medical Books,

London, 1963, pp. 110e118.
[56] L.Syllaba,K.Henner,Contributiona l’inde pendancedel’athe tosedouble

idiopatique et conge nitale. Atteinte familiale, syndrome dystrophique, signe du re seau vasculaire conjonctival, inte grite psychique, Rev. Neurol. 1 (1926) 541e562.

[57] C.E. Wells, G.M. Shy, Progressive familial choreoathetosis with cutaneous telangiectasia, J. Neurol. Neurosurg. Psychiatr. 20 (1957) 98e104.

[58] T.S. Pearson, More than ataxia: hyperkinetic movement disorders in child- hood autosomal recessive ataxia syndromes, Tremor Other Hyperkinet Mov (NY) 6 (2016), https://doi.org/10.7916/D8H70FSS.

[59] I. Sahama, K. Sinclair, K. Pannek, M. Lavin, S. Rose, Radiological imaging in ataxia telangiectasia: a review, Cerebellum 13 (2014) 521e530.

[60] D.Louis-Bar,Surunsyndromeprogressifcomprenantdeste langiectasies capilares cutane es et conjonctivalles syme triques, a disposition naevode et de trobles ce rebelleux, vol. 4, Confin Neurol (Basel), 1941, pp. 32e42.

[61] W.R. Centerwall, M.M. Miller, Ataxia, telangiectasia, and sinopulmonary in- fections. A syndrome of slowly progressive deterioration in childhood, Am. J. Dis. Child. 95 (1958) 385e396.

[62] L. Martin, Aspect chore oathe tosique du syndrome d’ataxie-te langiectasie, Acta Neurol. 64 (1964) 802e819.

[63] A. Biemond, Paleocerebellar atrophy with extra-pyramidal manifestations in association with bronchiectasis and telangiectasis of the conjunctiva bulbi as a familial syndrome, in: L. van Bogaert, J. Radermecker (Eds.), Proc. 1st Intern. Congr. Of Neurological Sciences – Brussels, Pergamonon Press, London, 1957, p. 206.

[64] E. Boder, R.P. Sedgwick, Ataxia-telangiectasia a familial syndrome of pro- gressive cerebellar ataxia, oculocutaneous telangiectasia and frequent pul- monary infection, Pediatrics 21 (1958) 526e554.

[65] R.P. Sedgwick, E. Boder, Progressive ataxia in childhood with particular reference to ataxia telangiectasia, Neurology 10 (1960) 705e715.

[66] H.G. Dunn, H. Meuwissen, C.S. Livingstone, K.K. Pump, Ataxia-telangiectasia, Can. Med. Assoc. J. 91 (1964) 1106e1118.

[67] M.M. Verhagen, W.F. Abdo, M.A. Willemsen, et al., Clinical spectrum of ataxia- telangiectasia in adulthood, Neurology 73 (2009) 430e437.

[68] F. Carrillo, S.A. Schneider, A.M. Taylor, V. Srinivasan, R. Kapoor, K.P. Bhatia, Prominent oromandibular dystonia and pharyngeal telangiectasia in atypical ataxia telangiectasia, Cerebellum 8 (2009) 22e27.

[69] C.G. Woods, A.M. Taylor, Ataxia telangiectasia in the British Isles: the clinical and laboratory features of 70 affected individuals, Q J Med 82 (1992) 169e179.

[70] S. Wijemanne, J. Jankovic, Dopa-responsive dystonia e clinical and genetic heterogeneity, Nat. Rev. Neurol. 11 (2015) 414e424.

[71] A. Nissenkorn, S. Hassin-Baer, S.F. Lerman, Y.B. Levi, M. Tzadok, B. Ben-Zeev, Movement disorder in ataxia-telangiectasia: treatment with amantadine sulfate, J. Child Neurol. 28 (2013) 155e160.

[72] A.S. de Graaf, G. de Jong, W.J. Kleijer, An early-onset recessive cerebellar disorder with distal amyotrophy and, in two patients, gross myoclonia: a probable ataxia telangiectasia variant, Clin. Neurol. Neurosurg. 97 (1995) 1e7.

[73] Ataxia Telangiectasia, Recruiting, not yet recruiting studies, in: Clinical- Trials.gov. National Institute of Health e US National Library of Medicine, 2017. https://clinicaltrials.gov/ct2/results?cond1⁄4Ataxia- telangiectasia&term1⁄4&cntry11⁄4&state11⁄4&Search1⁄4Search&recrs1⁄4a&recrs1⁄4b. (Accessed 5 December 2017).

Parkinsonism and Related Disorders 46 (2018) 9e15

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Incobotulinum toxin A in Parkinson’s disease with foot dystonia: A double blind randomized trial

Isabelle Rieu a, Bertrand Degos b, Giovanni Castelnovo c, Christophe Vial d,
Elodie Durand a, Bruno Pereira e, Marion Simonetta-Moreau f, Sophie Sangla b, Fre de rique Fluche re g, Dominique Guehl h, Pierre Burbaud h, Christian Geny i, Dominique Gayraud j, Fabienne Ory-Magne f, Françoise Bouhour d, Elisabeth Llinares c, Philippe Derost a, Ana Marques a, Franck Durif a, *

a Service de neurologie, CHU Clermont-Ferrand, Universit e Clermont Auvergne, Clermont-Ferrand, France
b D epartement de neurologie, Centre Expert et Inter-R egional de Coordination de la Maladie de Parkinson, Groupe Hospitalier Universitaire Piti e- Salp^etri ere, APHP, Paris, France
c Service de neurologie, Centre Hospitalo-Universitaire Caremeau, Nimes, France
d Service ENMG et pathologies neuromusculaires, Ho^pital Neurologique Pierre Wertheimer, Hospices Civils de Lyon, Lyon, France
e CHU Clermont-Ferrand, Biostatistics Unit, DRCI, France
f Neurosciences Department, University Hospital of Toulouse, University of Toulouse 3, UMR 1214, INSERM, UPS ToNIC Toulouse, France
g Department of Neurology and Movement Disorders, Po^le de Neurosciences Cliniques, Timone university hospital, Aix-Marseille University, Marseille, France
h Service de neurophysiologie clinique, CHU de Bordeaux, IMN UMR CNRS 5293, Universit e de Bordeaux, Bordeaux, France
i Service de m edecine Interne et Gerontologie, Centre Antonin Balmes, Montpellier, France
j Service de Neurologie, Centre Hospitalier Aix-Pertuis, Aix-en Provence, France

articleinfo abstract

Article history:

Received 8 June 2017 Received in revised form 7 September 2017 Accepted 16 October 2017

Keywords:

Foot dystonia
Parkinson’s disease
Botulinum toxin
Clinical trials randomized controlled

Introduction: Plantar flexion of toe dystonia is very painful and leads to difficulties in walking. The objective of this study was to investigate the effect of incobotulinum toxin A (Xeomin) in the treatment of this type of dystonia in parkinsonian patients, using a randomized, double blind, placebo-controlled trial.

Methods: 45 parkinsonian patients with painful dystonic plantar flexion of toes were injected either with incobotulinum toxin A (Btx group), or with placebo in two muscle targets: the Flexor digitorum longus and the Flexor digitorum brevis. Three groups were compared: the first group received placebo in the Flexor digitorum longus and 100UI of Btx in the Flexor digitorum brevis (n 1⁄4 16); the second group received 100 UI of Btx in the Flexor digitorum longus and placebo in the Flexor digitorum brevis (n 1⁄4 13); and the third group, 2 injections of placebo (n 1⁄4 16). The patients were injected in the same way twice with an interval of 3 months. The primary endpoint was measured six weeks after injections with the Clinical Global Impression (CGI) of change. Dystonia severity and associated pain were also assessed.

Results: Mean CGI was improved in the Btx group compared to the placebo group (P 1⁄4 0.039). A sig- nificant reduction of pain and dystonia severity were observed in patients treated with Btx compared to baseline but no improvement was noted when compared to placebo group. No difference of efficacy was highlighted between the two injection sites.

Conclusions: Btx injections are effective for improving clinical state of parkinsonian patients with plantar flexion of toe dystonia.

1. Introduction

* Corresponding author. Service de Neurologie, Centre Hospitalier Universitaire Gabriel Montpied, 58 rue Montalembert, 63003, Clermont-Ferrand cedex 1, France.

E-mail address: fdurif@chu-clermontferrand.fr (F. Durif).

https://doi.org/10.1016/j.parkreldis.2017.10.009

Various clinical features of foot dystonia (FD) are observed in idiopathic Parkinson’s disease (IPD), ranging from simple forms such as inversion or hallux extension, to complex forms combining

10 I. Rieu et al. / Parkinsonism and Related Disorders 46 (2018) 9e15

inversion, plantar and toe flexion, intrarotation, and dorsal and hallux extension [1]. FD can occur in the morning before the first dose of levodopa has been taken and/or during daily off-periods, but also during on-periods.

Several medical options have been used to treat FD such as adaptation of dose frequency, dose size and galenic formulation of levodopa, dopamine agonist, apomorphine injection, anticholiner- gics and lithium [2]. Botulinum toxin A could be an interesting alternative as it improves FD locally without modifying the pa- tient’s antiparkinsonian treatment [3,4]. An open-label pilot study showed that onabotulinum toxin A reduces pain with a concomi- tant improvement of dystonic spasm in 30 IPD patients with FD [1]. This was recently supported by case studies in which onabotulinum toxin A significantly improved FD and lower limb functional out- comes in 6 IPD patients [5].

In our medical practice, prolonged plantar flexion of toe dys- tonia (PFTD), a specific form of FD, is very painful and leads to difficulties in walking. Although botulinum toxin A can improve FD, no specific study has shown the efficacy of incobotulinum toxin A for this form of FD.

The main objective of this double blind, placebo controlled, randomized study was to demonstrate that incobotulinum toxin A is beneficial in reducing PFTD and associated pain in IPD patients. Furthermore, as no consensus exists on the methodologies used, two injection sites in lower limb muscles were compared: Flexor digitorum longus vs Flexor digitorum brevis.

2. Methods

2.1. Participants

Forty-six IPD patients were recruited in 10 French centers from 2011 to 2014. The sample characteristics included patients aged 30e80 years, with an idiopathic IPD diagnosed according to the UK PD Brain Bank criteria by a neurologist skilled in movement dis- orders [6,7]. Patients had to report unilateral or bilateral PFTD (>1 year) for at least 1 h per day, inducing walking difficulties. Patients may have received a Btx injection up to three months before in- clusion. Antiparkinsonian treatment had to be stable for at least 3 months before inclusion and during the study. Overall, disease severity was graded according to the modified Hoehn and Yahr scale [8]. The main exclusion criteria were atypical Parkinsonian syndrome or contraindication to Btx injection. Patients with extension of the big toe dystonia were not included.

2.2. Standard protocol approvals, registrations, and patient consents

The study protocol was approved by the local Medical School Ethical Committee (#2008-32) and performed according to the principles set out in the Declaration of Helsinki and to French legislation. The study was also registered on the specific clinical trial website (NCT 00909883). The nature and potential risks of the study were fully explained and written informed consent was ob- tained from each participant.

2.3. Protocol design

In this double-blind parallel study, patients were injected twice: the first time at baseline and the second one after 12 weeks. The injections were performed in two different sites: the lower limb muscles Flexor digitorum longus and Flexor digitorum brevis. Patients with bilateral FD were injected only in the side where FD was more severe.

Patients were randomized in one of the 2 following groups: the

“Btx group” which received 100UI of incobotulinum toxin A (n 1⁄4 29) or the “Placebo group” (Pl) (n 1⁄4 16). The Btx group was divided into (1) the “Extrinsic muscle group” (E) (n 1⁄4 14), i.e. pa- tients receiving one injection of incobotulinum toxin A (100UI) in the Flexor digitorum longus and one injection of Placebo in the Flexor digitorum brevis; and (2) the “Intrinsic muscle group” (I) (n 1⁄4 16), i.e. patients receiving one injection of incobotulinum toxin A (100UI) in the Flexor digitorum brevis and one injection of Placebo in the Flexor digitorum longus. The “Placebo group” (Pl) included pa- tients receiving only placebo injections, whatever the muscle considered. (Fig. 1).

Randomization was done by the MERZ Pharmaceuticals using RANCODE (Version 3.6, IDV Datenanalyse und Versuchsplanung, Gauting, Germany). Randomization was done in the ratio 1:1:1. This allocation was concealed to both the physicians and the patients involved.

Vials containing either incobotulinum toxin A (XEOMIN®) or placebo were indistinguishable and provided by MERZ Pharma- ceuticals. Incobotulinum toxin A or placebo was diluted with 1 ml saline 0.9% to obtain a concentration of 100U/ml. An injection of 100UI or placebo was performed in one site in each muscle under electromyogram detection guidance using a 26 gauge injectrode. The location of the injection was not specified in the protocol; the investigators were allowed to inject at their convenience according to the results of the electromyogram detection.

2.4. Assessments

The primary outcome measure of this study was the improve- ment of the patient’s clinical state evaluated by the Clinical Global Impression of change (CGI) [9,10] measured 6 weeks and 18 weeks after the injections.

Secondary outcome measures were applied at baseline and þ6, þ12 and þ 18 weeks after injection as follows: dystonia was assessed using the Burke-Fahn-Marsden (BFM) Scale [11]. This scale evaluates dystonia in different body areas but in this study, only the lower limb injected was evaluated. Dystonia severity was estimated by a subjective scale (1 light, 2: moderate, 3: severe, 4: very severe). Dystonia associated pain was assessed using the Visual-Analogue Scale (VAS) [12] and quality of life by the 39-item Parkinson’s Disease PDQ39 [13] The pain linked to Btx/Placebo injections was also evaluated after each injection, by using the Visual-Analogue Scale (VAS) [12]. Assessments were made by the same physicians who injected the patients.

2.5. Statistical analyses

According to the data available from the literature, in which only number of patients improved regarding dystonia severity and pain after incobotulinum toxin A injection were given [1], it did not appear possible to calculate an optimal sample size to demonstrate that intramuscular injections of incobotulinum toxin A are bene- ficial to reduce FD. Thus sample size was estimated according to Cohen’s recommendations which defined effect-size (ES) bounds as: small (ES: 0.2), medium (ES: 0.5) and large (ES: 0.8, “grossly perceptible and therefore large”) [14]. In addition, we decided to include half as many patients in the placebo group. Finally, we calculated that 24 patients in the Btx group and 12 patients in the Pl group would enable highlighting an effect size equal to 1 for a two- tailed type I error a 1⁄4 0.05 and statistical power at 80%, corre- sponding to a minimal difference at 1 point (with the standard- deviation equal to 1) for the primary endpoint. Considering possible lost to follow-up, we decided to include 30 patients in the Btx group (15 patients injected in the extrinsic muscle group and 15 patients injected in the intrinsic muscle group) and 15 patients in

I. Rieu et al. / Parkinsonism and Related Disorders 46 (2018) 9e15 11

Fig. 1. Sequence of study events.
Patients who fulfilled the inclusion criteria (n 1⁄4 46) were randomized in one of the 3 following groups: the “Extrinsic muscle group” (E) (n 1⁄4 14), i.e. patients receiving one injection of incobotulinum toxin A (100U) in the Flexor digitorum longus and one injection of Placebo in the Flexor digitorum brevis and (2) the “Intrinsic muscle group” (I) (n 1⁄4 16), i.e. patients receiving one injection of incobotulinum toxin A (100U) in the Flexor digitorum brevis and one injection of Placebo in the Flexor digitorum longus, or the “Placebo group” (Pl) (n 1⁄4 16), i.e. patients receiving only placebo injections whatever the muscle considered. Patients were injected twice with similar doses, one at baseline and one at þ12 weeks.

the Pl group, thereby studying an effect-size around 0.9. Statistical analysis was performed using Stata 13 software (StataCorp LP, College Station, US). The tests were two-sided, with a type I error set at a 1⁄4 0.05. Subjects’ characteristics were presented as mean (±SEM) or median [interquartile range] according to sta- tistical distribution data (normality assessed using the Shapir- oeWilk test) for continuous parameters. Comparisons between groups were performed using chi-squared or Fisher’s exact tests for categorical variables, and the Student t-test or Mann-Whitney test for quantitative parameters (homoscedasticity was studied using the Fisher-Snedecor test). These analyses were completed by random-effects models to study fixed effects groups, time-points and their interaction and take into account within and between patient variability. Multivariate analyses were performed to take into account adjustment on possible confounding factors,

determined on the basis of univariate results and clinical relevance such as age, gender, and treatments, at the inclusion and during the follow-up.

3. Results

3.1. Sample characteristics

Forty-six patients were enrolled in the study and 45 were finally analysed (Fig. 1). One patient of the E group was excluded because of protocol deviation; this patient has received the same day, in- jections into the 2 lower limbs and for a blepharospasm. The Btx and Pl groups were comparable for demographic data, disease duration and severity and antiparkinsonian treatment at inclusion (Table 1). There were no difference regarding dystonia

I. Rieu et al. / Parkinsonism and Related Disorders 46 (2018) 9e15

Table 1

Sample characteristics at inclusion.

Age (y)
Sex, Male, n (%)
Weight (kg)
Time since diagnosis (y)
Age at the beginning of disease(y)
Hoehn and Yahr score
Type of dystonia (%)
OFF dystonia alone
ON dystonia alone
ON þ OFF dystonia
Biphasic dystonia alone
Biphasic þ OFF dystonia
Dystonia Severity
BFM -lower limb item-
Dystonia pain VAS (mm)
Dopamine agonist dose (expressed as levodopa equivalent dose (mg/day))
Levodopa dose (mg/day)
Levodopa and Dopamine agonist doses (expressed as levodopa equivalent dose (mg/day)) PDQ39 SI

Group Btx Group Pl P (n1⁄429) (n1⁄416) BtxvsP

67.1 ± 1.6 64.9 ± 2.2 0.81 14 (48.3) 10 (62.5) 0.36 68.5 ± 2.6 76.3 ± 3.8 0.09 8.7 ± 0.9 8.9 ± 1.2 0.96 58.4 ± 1.9 56.0 ± 2.3 0.73 2.2 ± 0.1 2.5 ± 0.6 0.34

61 60 0.57 8 7
23 26
8 0

0 7
3.2 ± 0.1 3.3 ± 0.1 0.40 8.2 ± 0.5 8.3 ± 0.7 0.20 55.9 ± 4.7 50.6 ± 5.4 0.30 217 ± 242 248 ± 379 0.94 627 ± 394 547 ± 433 0.31 823 ± 469 794 ± 571 0.26 8.0 ± 0.5 8.8 ± 0.9 0.88

Group E (n 1⁄4 13)

67.7 ± 1.8 7 (53.9) 72.3 ± 3.5 8.5 ± 1.3 59.1 ± 2.3 2.2 ± 0.2

60
0
40
0
0
3.1 ± 0.1 7.5 ± 0.8 49.9 ± 7.7 193±41 767 ± 133 961 ± 124 7.8 ± 0.6

Group I P (n1⁄416) IvsE

66.6 ± 2.5 0.72 7 (43.8) 0.59 65.3 ± 3.7 0.18 8.8 ± 1.2 0.91 57.8 ± 3.0 0.73 2.3 ± 0.2 0.69

62 0.21 12
12
12

0
3.3 ± 0.2 0.26 8.8 ± 0.7 0.25 60.8 ± 5.6 0.27 236±75 0.76 507±69 0.14 711 ± 117 0.08 8.2 ± 0.8 0.71

characteristics (dystonia severity, BFM score), pain associated with dystonia, and quality of life between the Btx and Pl groups. Eigh- teen percent of the patients enrolled in the study had already had a toxin injection for foot dystonia. Among them, the mean duration between the last injection and the inclusion in the study was 7 ± 1 months.

3.2. Effect of incobotulinum toxin A injections

Patients treated by incobotulinum toxin A showed a moderate improvement after injections with a CGI score of 3.14 ± 0.22 at þ6 weeks and 2.90 ± 0.20 at þ 18 weeks (Fig. 2). Incobotulinum toxin A injections significantly decreased pain associated with dystonia (þ6 weeks vs. baseline, P<0.001; þ18 weeks vs. baseline, P1⁄40.002), dystonia severity (þ6 weeks vs. baseline, P < 0.001; and þ18 weeks vs. baseline, P < 0.001) and the BFM score (lower limb item) (þ6 weeks vs. baseline, P 1⁄4 0.026; þ18 weeks vs. baseline, P 1⁄4 0.001) (Fig. 3).

In the placebo group, a very slight improvement of the CGI score was observed (3.67 ± 0.16 at 6 weeks and 3.40 ± 0.25 at 18 weeks) (Fig. 2). Dystonia associated pain, dystonia severity and the BFM score (lower limb item) were not improved in the placebo group at 6 weeks (P1⁄4NS) and slightly improved for severity dystonia at 18

weeks (Fig. 3).
Comparisons between Btx and Pl groups showed that CGI, the

primary outcome, was improved in the Btx group: mean CGI (average between 6 weeks and 18 weeks) was significantly higher in the Btx group compared to the Pl group (3.02 ± 0.21 vs 3.53 ± 0.21 respectively, ES 1⁄4 [0.8], 95% confidence interval [0.35; 1.25], P 1⁄4 0.039) (Fig. 2). Dystonia severity and associated pain levels and BFM score (lower limb item) were not different between the Btx and the Pl groups, independently of the time considered.

Pain related to injection was not significantly different between the Btx and Pl groups (43.8 ± 4.9 mm for Btx group vs 33.6 ± 7.3 mm for Pl group at baseline; 52.6 ± 4.3 mm for the Btx group vs 42.9 ± 7.1 mm for the Pl group at þ12 weeks).

Quality of life (either PDQ 39 SI or the others sub-scores) was similar between the 2 groups (P1⁄4 NS) and was not modified by injections (data not shown).

3.3. Comparison of the two injection sites

Patients showed a moderate improvement of their clinical state after injections with a CGI score of 3.00 ± 0.30 at þ6 weeks and 2.77 ± 0.33 at þ18 weeks for group E, and at 3.27 ± 0.33 at þ6 weeks and 3.00 ± 0.27 at þ18 weeks for group I. The percentage of patients improved by injections was 61.5% and 69.2% at þ6 weeks and þ18 weeks respectively in group E and 46.7% and 56.3% at þ6 weeks and þ18 weeks respectively in group I.

Incobotulinum toxin A injections significantly decreased dys- tonia severity of group E at 6 weeks and 18 weeks, respectively, compared to baseline (Supplemental Fig. 1). The BFM score (lower limb item) was significantly improved after the second injection only at 18 weeks, whereas pain associated dystonia was not significantly modified. Incobotulinum toxin A injections signifi- cantly decreased pain associated dystonia and dystonia severity of group I at 6 weeks and 18 weeks; the BFM score (lower limb item) was only significantly reduced at 18 weeks (Supplemental Fig. 1).

In both groups quality of life was not modified by incobotulinum toxin A injections whatever the period (P1⁄4NS, data not shown).

There were no significant differences for CGI, pain, severity, BFM score (lower limb item) of dystonia or quality of life between the 2 injection sites evaluated (P1⁄4NS). Only CGI tended to be improved when injections were made in extrinsic muscles (P 1⁄4 0.08, all times considered). Furthermore, pain experienced at the time of injection

7 6 5 4 3 2 1

Btx group Pl group

ES = 0.80, CI95% [0.35; 1.25]

average

Fig. 2. Clinical improvement induced by incobotulinum toxin A/Pl injections evaluated by the CGI.

P= 0.039

Improvement

Aggrava on CGI score

I. Rieu et al. / Parkinsonism and Related Disorders 46 (2018) 9e15

4 3.5 3 2.5 2 1.5 1

B- 70 60

50 40 30 20 10

* ***

A-

***

Btx group Pl group

12 10 8 6 4 2 0

C-

Baseline

+6 weeks

+12 weeks

+18 weeks

+6 weeks

***

+18 weeks

*** **

Baseline

+12 weeks

Baseline

+6 weeks

+12 weeks

+18 weeks

* P<0.05 vs Baseline; ** P<0.01 vs Baseline; *** P<0.001 vs Baseline for Btx group # P<0.01vs Baseline for Pl group

Fig. 3. Effect of incobotulinum toxin A/Pl injections on pain associated dystonia (VAS scale) and dystonia characteristics (severity and BFM score).

was similar independently of the group considered (46.1 ± 7.5 mm for group E vs 41.8 ± 6.6 mm for group I at baseline; 51.3 ± 8.0 mm for group E vs 53.6 ± 4.8 mm for group I at þ12 weeks).

3.4. Adverse events

Nine falls were reported: three in group E, four in group I and two in group Pl. Two falls in group I were recorded a few days after the toxin injections but the investigators did not impute these side effects to Btx treatment. With regard to the other side effects, a patient of group E described a 3-day loss of sensation in his leg, a few days after the toxin injections. This was linked to the Btx treatment by the investigator. Moreover, localized foot pain was reported a few days after the injections in a patient of group I but was not attributed to the treatment. Finally, a patient of group Pl developed gastroparesis, independently of the treatment, which led to hospitalization for 3 days.

4. Discussion

This randomized placebo controlled double-blind study demonstrated that patients’ clinical state evaluated by the CGI was significantly improved when patients were treated by incobotuli- num toxin A vs placebo. Intragroup analysis showed that incobo- tulinum toxin A injections decreased the severity of PFTD and associated pain in IPD patients. However, no significant difference regarding efficacy was observed between incobotulinum toxin A injections performed in extrinsic vs intrinsic muscles.

PFTD is often a very painful dystonia when walking or at rest, causing significant functional impairment. The pharmacological

treatment of FD requires careful analysis of its relationship with L- Dopa treatment, both in terms of timing and dose sizes. The advantage of incobotulinum toxin A treatment is its local effect on FD without the need to modify antiparkinsonian treatment, espe- cially when adapted.

Botulinum toxin A has frequently been used successfully to treat a variety of symptoms related to Parkinson’s disease, such as cer- vical dystonia and blepharospasm [4,15e19]. This toxin is already used in current practice for the treatment of foot and toe dystonia but this use has not been supported by robust published studies. Onabotulinum toxin A was used to treat “off” painful FD in 30 IPD patients in an open label pilot study and improved pain within 10 days with no side effects [1]. In a recent study, onabotulinum toxin A injections improved FD and pain in 6 IPD patients treated by deep brain stimulation [5]. The patients included in these two studies were treated for different types of foot dystonia (simple or complex forms of dystonia). In this study, we focused on foot dystonia with plantar flexion of the toes. To our knowledge, there are no studies confirming the efficacy and safety of Btx for the treatment of FD and PFTD in IPD. In our controlled, randomized, double-blind study we showed that incobotulinum toxin A injections significantly improved the clinical state of our patients when compared with Pl injections. The CGI score was significantly improved in Btx group; nevertheless the other clinical points (severity, pain associated dystonia) remained similar in Btx and Pl groups. This absence of significant differences between the 2 groups could be explained by a lack of statistical power and should be confirmed by others studies with higher number of subjects. Intragroup analysis underlined that all the clinical features of FD (BFM score-lower limb item, severity and pain associated dystonia) were improved in the

BFM score (lower limb item)

VAS score (mm) Dystonia severity score

14 I. Rieu et al. / Parkinsonism and Related Disorders 46 (2018) 9e15

Btx group.
Although onabotulinum toxinA (Botox) and incobotulinum

toxinA (Xeomin) are two different drugs, a 1/1 conversion ratio is generally admitted [20]. In this study, we used a dose of 100 IU whatever the muscle injected. This dose has been used in several studies, especially in spasticity or dystonia treatment [1,5,21]. Regarding published data using Botox injections in the treatment of foot dystonia, doses are ranging depending on the studies from 40 to 150 IU by muscle [1,5]. We have chosen to inject a relatively higher dose (100 IU by muscle) in an attempt to obtain a significant result with a reduced risk of side effects.

Incobotulinum toxin A injections were performed under elec- tromyography detection to ensure accurate injection into the muscles targeted. We avoided electrical muscle stimulation in or- der to reduce pain, especially in the Flexor digitorum brevis muscle. Injections were performed only in one site for each muscle, in order to decrease the pain induced by the needle, especially for the intrinsic muscles. Thus we cannot exclude higher efficacy of the multiple injection sites, in particular in the Flexor digitorum longus muscle. Moreover, we chose to inject only one muscle with inco- botulinum toxin A. We cannot rule out that injected both muscles at the same time, for a more complete coverage of the dystonia, could be more effective in reducing toe dystonia. Further experiments were needed to address this question.

Two sets of identical injections (one at baseline and one at 12 weeks) were planned to seek a synergic effect between the two injections. However, we failed to demonstrate such an effect, although the duration and severity dystonia tended to improve after the 2nd Btx injection.

No difference of clinical state, severity dystonia and associated pain were noted between the two injection sites. CGI tend to improve when the injections were performed in the Flexor dig- itorum longus compared to the Flexor digitorum brevis (P 1⁄4 0.08, all times considered). This was probably due to a lack of power related to the low numbers of patients (n 1⁄4 16 in group I vs n 1⁄4 13 in group E). Higher numbers of subjects are needed to address this question.

Identical dose of incobotulinum toxin A (100IU) were used for both muscles in order to reduce bias. If there was a difference in effectiveness between the two muscles, it would not have been attributed to a difference in dose but to a difference in the char- acteristics of the 2 muscles. Moreover, we chose to inject the same dose of incobotulinum toxin A into the two target muscles despite the fact that the Flexor digitorum brevis is smaller than the Flexor digitorum longus; However, it is important to note that comparison between two muscles depends on providing an equivalent dose (which could directly be impacted by the size or the mass of the muscle) and we are not sure that the dosing for both were equiv- alent. Further experiments are needed to address this question.

Studies investigating the effects of botulinum toxin A injection on FD showed that this treatment is generally well tolerated with few side effects [1,5]. In our study, the number of falls were similar between the 3 groups studied (3 in group E, 4 in group I and 2 in group Pl). Two falls in group I were reported to be related to toxin injections from the patients’ viewpoint but not from that of the investigators. A patient of group E recorded a loss of sensation in the lower limb for 3 days, a few days after the toxin injections. This adverse event related to incobotulinum toxin A treatment could be explained more by a technical problem during the injection than to the product itself.

5. Conclusion

Incobotulinum toxin A injections were found to be effective to improve PFTD in Parkinson’s disease. This treatment could be considered as a first line therapy for the treatment of FD. Further

investigations including higher numbers of subjects are needed to assess whether these injections are more effective when made in extrinsic or intrinsic muscles.

Funding

This work was funded by MERZ Pharma.

Author’s contribution

I Rieu: study coordination, data acquisition, contribution to data analysis and interpretation, and writing the manuscript. E Durand: data acquisition, contribution to data analysis and interpretation. B Pereira: statistical analysis. C Vial, S Sangla, P Burbaud, F Fluche re, C Geny, D Gayraud, F Ory-Magne, F Bouhour: neurological assess- ment. E Llinares: data acquisition. P Derost: study design and neurological assessment. G Castelnovo G, B Degos, M Simonetta- Moreau, D Guehl, A Marques: neurological assessment and revising manuscript. F Durif: study design, study supervision, obtaining erpretation of data and revising manuscript.

Conflict of interest

The authors report no disclosures relevant to the manuscript, except S. Sangla who is a consultant for MERZ Pharma. However, she did not receive financial support for this research.

Role of funding source

MERZ Pharma provided financial support for the conduct of the research but was not involved in the study design, in the collection, analysis and interpretation of data, in the writing of the report and the decision to submit the article for publication.

Acknowledgments

The authors would like to acknowledge MERZ pharma that provided financial support for this research.

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.009.

References

[1] C.Pacchetti,G.Albani,E.Martignoni,L.Godi,E.Alfonsi,G.Nappi,“Off”painful dystonia in Parkinson’s disease treated with botulinum toxin, Mov. Disord. 10 (1995) 333e336.

[2] E. Tolosa, Y. Compta, Dystonia in Parkinson’s disease, J. Neurol. 253 (Suppl 7) (2006) 7e13.

[3] P. Limousin, B. Memin, P. Pollak, Treatment of dystonia occurring in parkin- sonian syndromes by botulinum toxin, Eur. Neurol. 37 (1997) 66e67.

[4] R. Mills, L. Bahroo, F. Pagan, An update on the use of botulinum toxin therapy in Parkinson’s disease, Curr. Neurol. Neurosci. Rep. 15 (2015) 511.

[5] Gupta AD1, R. Visvanathan, Botulinum toxin for foot dystonia in patients with Parkinson’s disease having deep brain stimulation: a case series and a pilot study, J. Rehabil. Med. 48 (2016) 559e562.

[6] W.R.G. Gibb, A.J. Lees, The relevance of the Lewy body to the pathogenesis of idiopathic Parkinson’s disease, J. Neurol. Neurosurg. Psychiatry 51 (1988) 745e752.

[7] A.J. Hughes, Y. Ben-Shlomo, S.E. Daniel, A.J. Lees, What features improve the accuracy of clinical diagnosis in Parkinson’s disease: a clinicopathologic study, Neurology 42 (1992) 1142e1146.

[8] M.M. Hoehn, M.D. Yahr, Parkinsonism: onset, progression and mortality, Neurology 17 (1967) 427e442.

[9] W. Guy, Clinical Global Impression. ECDEU Assessment Manual for Psycho- pharmacology, revised National Institute of Mental Health, Rockville, MD, 1976.

[10] J. Busner, S.D. Targum, The clinical global impressions scale: applying a

research tool in clinical practice, Psychiatry 4 (2007) 28e37.

. [11] R.E. Burke, S. Fahn, C.D. Marsden, S.B. Bressman, C. Moskowitz, J. Friedman, Validity and reliability of a rating scale for the primary torsion dystonias,
Neurology 35 (1985) 73e97.

. [12] J. Scott, E.C. Huskisson, Graphic representation of pain (EVA), Pain 2 (1976)
175e184.

. [13] V. Peto, C. Jenkinson, Fitzpatrick R.PDQ-39: a review of the development,
validation and application of a Parkinson’s disease quality of life questionnaire
and its associated measures, J. Neurol. 245 (1998) 10e14.

. [14] J. Cohen, Statistical Power Analysis for the Behavioral Sciences, second ed.,
Lawrence Earlbaum Associates, Hillsdale, NJ, 1988.

. [15] D. Truong, Botulinum toxins in the treatment of primary focal dystonias,
J. Neurol. Sci. 316 (2012) 9e14.

. [16] M. Hallett, A. Albanese, D. Dressler, K.R. Segal, D.M. Simpson, D. Truong,
J. Jankovic, Evidence-based review and assessment of botulinum neurotoxin for the treatment of movement disorders, Toxicon 67 (2013) 94e114.

[17] R. Benecke, D. Dressler, Botulinum toxin treatment of axial and cervical dys- tonia, Disabil. Rehabil. 29 (2007) 1769e1777.

[18] D. Truong, D.D. Duane, J. Jankovic, C. Singer, L.C. Seeberger, C.L. Comella, M.F. Lew, R.L. Rodnitzky, F.O. Danisi, J.P. Sutton, P.D. Charles, R.A. Hauser, G.L. Sheean, Efficacy and safety of botulinum type A toxin (Dysport) in cervical dystonia: results of the first US randomized, double-blind, placebo-controlled study, Mov. Disord. 20 (2005) 783e791.

[19] Comella CL1, J. Jankovic, D.D. Truong, A. Hanschmann, S. Grafe, U.S. XEOMIN Cervical Dystonia Study Group. Efficacy and safety of incobotulinumtoxinA (NT 201, XEOMIN®, botulinum neurotoxin type A, without accessory proteins) in patients with cervical dystonia, J. Neurol. Sci. 308 (2011) 103e109.

[20] F. Scaglione, Conversion ratio between Botox®, Dysport®, and Xeomin® in clinical practice, Toxins 8 (2016) 65.

[21] L. Nalysnyk, S. Papapetropoulos, P. Rotella, J.C. Simeone, K.E. Alter, A. Esquenazi, OnabotulinumtoxinA muscle injection patterns in adult spas- ticity: a systematic literature review, BMC Neurol. 8 (13) (2013) 118.

I. Rieu et al. / Parkinsonism and Related Disorders 46 (2018) 9e15 15

Parkinsonism and Related Disorders 46 (2018) 16e23

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Cerebrospinal fluid levels of coenzyme Q10 are reduced in multiple system atrophy

Yaroslau Compta a, b, Darly M. Giraldo a, Esteban Mun~oz a, Francesca Antonelli a, Manel Ferna ndez a, Paloma Bravo a, Marta Soto a, Ana Ca mara a, Ferran Torres c, María Jose Martí a, *on behalf of theCatalan MSA Registry (CMSAR)1

a Parkinson’s Disease and Movement Disorders Unit, Neurology Service, ICN, Hospital Clínic, IDIBAPS, CIBERNED, University of Barcelona, Barcelona, Catalonia, Spain
b Department of Biomedicine, University of Barcelona, Barcelona, Catalonia, Spain
c Medical Statistics Core Facility, IDIBAPS & Biostatistics Unit, Faculty of Medicine, Universitat Auto noma de Barcelona, Catalonia, Spain

articleinfo abstract

Article history:

Received 7 April 2017 Received in revised form 20 September 2017 Accepted 17 October 2017

Keywords:

Multiple system atrophy Parkinson’s disease Progressive supranuclear palsy Atypical parkinsonisms Cerebrospinal fluid

Biomarker Coenzyme Q10

Introduction: The finding of mutations of the COQ2 gene and reduced coenzyme Q10 levels in the cer- ebellum in multiple system atrophy (MSA) suggest that coenzyme Q10 is relevant to MSA pathophysi- ology. Two recent studies have reported reduced coenzyme Q10 levels in plasma and serum (respectively) of MSA patients compared to Parkinson’s disease and/or control subjects, but with largely overlapping values, limited comparison with other parkinsonisms, or dependence on cholesterol levels. We hypothesized that cerebrospinal fluid (CSF) is reliable to assess reductions in coenzyme Q10 as a candidate biomarker of MSA.

Methods: In this preliminary cross-sectional study we assessed CSF coenzyme Q10 levels in 20 patients with MSA from the multicenter Catalan MSA Registry and of 15 PD patients, 10 patients with progressive supranuclear palsy (PSP), and 15 control subjects from the Movement Disorders Unit Biosample Collection of Hospital Clinic de Barcelona. A specific ELISA kit was used to determine CSF coenzyme Q10 levels. CSF coenzyme Q10 levels were compared in MSA vs. the other groups globally, pair-wise, and by binary logistic regression models adjusted for age, sex, disease severity, disease duration, and dopami- nergic treatment.

Results: CSF coenzyme Q10 levels were significantly lower in MSA than in other groups in global and pair-wise comparisons, as well as in multivariate regression models. Receiver operating characteristic curve analyses yielded significant areas under the curve for MSA vs. PD, PSP and controls.
Conclusions: These findings support coenzyme Q10 relevance in MSA. Low CSF coenzyme Q10 levels deserve further consideration as a biomarker of MSA.

1. Introduction

Multiple system atrophy (MSA) is an incurable and rapidly progressive neurodegenerative condition, clinically grouped with atypical parkinsonisms and neuropathologically classified as a synucleinopathy [1]. From a clinical standpoint, MSA has

* Corresponding author. Parkinson’s Disease & Movement Disorders Unit, Neurology Service, Clinical Neuroscience Institute (ICN), Hospital Clínic & Univer- sity of Barcelona, IDIBAPS, CIBERNED, 170 Villarroel, 08036 Barcelona, Catalonia, Spain.

E-mail address: mjmarti@clinic.cat (M.J. Martí). 1 See Appendix.

https://doi.org/10.1016/j.parkreldis.2017.10.010

parkinsonism (parkinsonian variant: MSAp) or ataxia (cerebellar variant: MSAc), plus dysautonomia (which is mandatory for the clinical diagnosis) as its core features [1]. Currently MSA can only be confirmed neuropathologically, with the clinico-pathological mismatch being particularly due to misdiagnose as Parkinson’s disease (PD) or progressive supranuclear palsy (PSP), another atypical parkinsonism, but with underlying tauopathy instead of synucleinopathy [2,3]. In view of this and the foreseeable emer- gence of disease-specific treatments, many efforts are underway to develop reliable biomarkers for these conditions.

The finding of mutations of the COQ2 gene [4] and reduced coenzyme Q10 in the cerebellum in MSA [5,6], have led to suggest that coenzyme Q10 is relevant to the pathophysiology of MSA

(perhaps through mitochondrial dysfunction and oxidation of alpha-synuclein) [7] and that it might be both a biomarker and a potential therapeutic target in this condition. Two recent studies have respectively shown reduced coenzyme Q10 levels in serum [8] and plasma [9] of MSA patients. However, in both studies the overlap among groups was remarkable, in the plasma study MSA was compared only to controls, and in the serum study differences were significant only after controlling for cholesterol levels, a known confounder of blood coenzyme Q10 levels.

One of the most widely explored sources of biomarkers of neurodegeneration is cerebrospinal fluid (CSF), after its close rela- tionship with brain tissue and the fact that, if properly collected, it is less likely of being influenced by systemic metabolites, as opposed to blood. Moreover, albeit its collection is invasive, it can be routinely, easily and safely performed [10].

For all these reasons we hypothesized that CSF is reliable to assess coenzyme Q10, and that CSF coenzyme Q10 levels are low- ered in MSA vs. other neurodegenerative parkinsonisms and controls.

2. Methods

2.1. Design

This is a pilot hypothesis-driven cross-sectional convenience study based on samples from the multicenter Catalan MSA Registry (CMSAR) and the Movement Disorders Biosample Collection of Hospital Clinic de Barcelona (MDBC-HCB). In short, the CMSAR is a multicenter initiative in which movement disorders specialists across Catalonia identify possible or probable MSA patients and offer them to participate in the registry and to donate biosamples (blood, urine, fibroblasts, CSF) at Hospital Clinic de Barcelona. The MDBC-HCB is a single centre biorepository consisting of several biosamples such as CSF of patients with degenerative parkinson- isms, including PD and PSP patients. Both the CMSAR and the MDBC-HCB along with the present CSF biomarker study have received approval from the competing Institutional Review Board, and all participants provided their written informed consent.

2.2. Participants

Sixty participants were studied: 20 sporadic MSA cases from the CMSAR clinically diagnosed according to currently accepted criteria [11] (10 classified as MSAp and 10 as MSAc; 13 as probable and 7 as possible MSA), and 40 additional subjects having contributed to the MDBC-HCB, and consisting of: 15 non-demented PD patients (all with a clinically definite diagnosis according to the United Kingdom Parkinson’s Disease Society Brain Bank criteria [12]); 10 patients with clinically probable PSP according to the NINDS-SPSP criteria [13]; and 15 controls who after comprehensive clinical history and examination as well as brain MRI without remarkable changes were judged not to have any known neurodegenerative disease. These control subjects donated CSF during their admission for knee replacement surgery with intradural anaesthesia. Part of the PD, PSP and control subjects have been reported in previous CSF studies [14,15].

Hoehn & Yahr stage at the time of inclusion was recorded [16] for MSA, PD and PSP participants. Additionally, the scores of the unified MSA rating scale (UMSARS) [17] and the motor section of the unified PD rating scale (UPDRS-III) [18] were available for MSA and PD patients, respectively. Mini mental state examination [19] and Mattis dementia rating scale (MDRS-2) [20] were used for cognitive assessment. Levodopa equivalent daily dose (LEDD) was calculated [21].

2.3. CSF collection, storage and analysis

All patients underwent lumbar puncture (LP) in L3-L4 space, using a 22G needle, between 8 and 10 a.m., after overnight fasting and under off-medication. The first 2 mL of CSF were used for routine studies. The following 10 mL of CSF were collected in polypropylene tubes and immediately centrifuged for 10 min at 4,000g and 4 C, and stored at 80 C in 300 mL polypropylene aliquots until analyses. After LP, patients stayed in bed for at least 2 h and were advised to increase water intake. To determine CSF coenzyme Q10 levels we used a commercially available specific ELISA technique (Human CoQ10 ELISA Kit; MBS701260; MyBio- Source; San Diego, CA, USA). All CSF samples were analyzed in duplicate using two ELISA kits with the same batch number, with 11 samples being tested in both experiments to control for inter-assay variability.

2.4. Statistical analyses

All data were analyzed using SPSS 20.0 (IBM, New York, USA). No formal statistical power calculations were carried out due to: (a) the pilot and exploratory nature of the study, (b) the rarity of the disease and the consequent unfeasibility to recruit large samples, and (c) the paucity of previous data on coenzyme Q10 in MSA, in general, and using CSF, in particular. Yet, considering both previous studies of coenzyme Q10 using peripheral blood with sample sizes in the range of ours [8,9] and prior CSF studies of other biomarkers with significant findings using similar sample sizes [14,15,22], the available cohort was deemed likely to allow for detecting relevant associations. Qualitative variables are presented as relative and absolute frequencies and compared with Fisher’s exact test. Quantitative variables are reported as medians and interquartile ranges, and compared across all groups by means of Kruskal-Wallis followed by pairwise Mann-Whitney’s U tests between MSA and each of the other groups (conditioned to only when the overall group test was significant), with post-hoc Bonferroni’s correction for multiplicity (all reported pairwise p-values are Bonferroni- corrected). Linear correlations were explored with Spearman cor- relation. Binary logistic regression models allowed for further testing of CSF coenzyme Q10 levels as a continuous quantitative predictor (expressed in ng/mL) with adjust for potential qualitative or quantitative modifiers (age, sex, disease duration, disease severity as per motor scales, LEDD). In the statistical analyses plan, at first several models were run separately, each one adjusted for part of the aforementioned potential modifiers, since the cohort size made it less reliable introducing all the potential modifiers at once in a single model, and as some of these variables were not applicable to the control group (disease duration, motor scales, LEDD). However, one such model pooling all the variables together was also run in the end, for exploratory purposes and as additional data. The results of these binary logistic regression models are presented as odds ratios (OR) and the respective 95% confidence intervals (95%CIs), with OR>1 indicating an increase in the outcome risk per each unit increase in the tested variable, with the opposite applying for OR<1. Finally, CSF coenzyme Q10 levels were addi- tionally tested as a potential MSA biomarker by means of receiver operating characteristic (ROC) analyses, where areas under the curve (AUCs) and 95%CIs greater than 0.5 indicate significant discriminant ability. All analyses were two-tailed, with p-threshold set at 0.05.

Y. Compta et al. / Parkinsonism and Related Disorders 46 (2018) 16e23 17

18 Y. Compta et al. / Parkinsonism and Related Disorders 46 (2018) 16e23

3. Results

3.1. Demographic and clinical features in MSA vs. PD vs. PSP vs. controls

The data of four study groups (MSA, PD, PSP and controls) are summarized in Table 1. There were no differences in sex distribu- tion. Regarding age at time of inclusion, PD and controls were significantly older than MSA and PSP participants. Age at disease onset was comparable across MSA, PD and PSP, whereas disease duration was significantly longer in PD. Conversely, Hoehn & Yahr stages were significantly worse in both MSA and PSP vs. PD, while LEDD was significantly higher in PD vs. both MSA and PSP. There were no differences in MMSE and MDRS-2.

3.2. DemographicandclinicalfeaturesinMSAsubgroups(MSApvs. MSAc; possible vs. probable MSA)

There were no differences in demographic and clinical variables between MSAp and MSAc cases except for trend towards greater UMSARS scores and a significantly greater LEDD in MSAp vs. MSAc (Suppl. Table 1). Similarly, possible and probable MSA cases were comparable in all demographic and clinical variables (Suppl. Table 2).

3.3. UnadjustedcomparisonsandcorrelationsofCSFcoenzymeQ10 levels in MSA vs. PD vs. PSP vs. controls

Intra- and inter-assay coefficients of variation were 13% and 17%, respectively. CSF coenzyme Q10 levels were significantly different in the global comparison of the four groups (p 1⁄4 0.003) due to significantly lower levels in MSA vs. PD, PSP and controls in pair- wise comparisons (MSA vs. PD, p 1⁄4 0.033; MSA vs. PSP, p 1⁄4 0.003; MSA vs. controls, p 1⁄4 0.036; all pairwise p-values are Bonferroni-corrected; Table 1; Fig. 1). There were two MSA outliers with high CSF coenzyme Q10 levels (Fig. 1). Both cases had been classified as probable MSAp. Comparative analyses among study group removing both these cases resulted in even larger differences (global comparison, p < 0.001; pairwise Bonferroni-corrected comparisons: MSA vs. PD, p 1⁄4 0.003; MSA vs. controls, p 1⁄4 0.009; MSA vs. PSP, p < 0.001).

Since, as mentioned before, no a priori power calculations had been carried out, we determined the 95%CIs of the medians of CSF coenzyme Q10 levels in each group as a measure of our study un- certainty, with those being consistent with the observed differ- ences in CSF coenzyme Q10 levels (Suppl. Table 3).

CSF coenzyme Q10 levels did not show any significant correla- tions with age at onset, age at inclusion, disease duration, motor and cognitive scales (where available and applicable), or LEDD, either in the entire cohort nor in the disease groups separately (data not shown).

3.4. Adjusted regression models and ROC analyses of CSF coenzyme Q10 as a predictor of MSA

In univariate binary logistic regression model with the MSA diagnosis as outcome and CSF coenzyme Q10 as predictor, consid- ering the entire cohort, CSF coenzyme Q10 levels had an OR of 0.916 (95%CI 1⁄4 0.869e0.965), indicating a significant increase in the risk of MSA per each ng/mL reduction of CSF coenzyme Q10 levels (p 1⁄4 0.002). Of the other variables (age at inclusion, sex, disease dura- tion, Hoehn & Yahr, and LEDD), only age at inclusion also yielded significant association in univariate binary logistic regression models (OR 1⁄4 0.854; 95%CI 1⁄4 0.775e0.941) (see the left columns of Table 2). In a covaried binary logistic regression model pooling

variables applicable to the entire cohort (CSF coenzyme Q10 levels, age at inclusion, sex; see middle columns of Table 2), again both CSF coenzyme Q10 levels (OR 1⁄4 0.891; 95%CI 1⁄4 0.827e0.960) and age at inclusion (OR 1⁄4 0.784; 95%CI 1⁄4 0.666e0.922) were significant predictors. Finally, in a covaried binary logistic regression model including variables applicable to disease-groups (CSF coenzyme Q10 levels, age at inclusion, sex, disease duration, Hoehn & Yahr, and LEDD; see the right columns of Table 2), CSF coenzyme Q10 levels were the only significant predictor of MSA (OR 1⁄4 0.851; 95% CI 1⁄4 0.760e0.954).

ROC curve analyses (Fig. 2) yielded significant AUCs and 95%CIs for lowering CSF coenzyme Q10 levels as predictor of MSA diagnosis when comparing MSA vs. all other groups together (AUC 1⁄4 0.744, 95%CI 1⁄4 0.641; 0.906; p 1⁄4 0.001; Fig. 2A), vs. PD (AUC 1⁄4 0.822; 95% CI 1⁄4 0.659; 0.986; p 1⁄4 0.002; Fig. 2B), vs. PSP (AUC 1⁄4 0.959; 95% CI 1⁄4 0.875; 1.000; p < 0.001; Fig. 2C), and vs. controls (AUC 1⁄4 0.800; 95%CI 1⁄4 0.644; 0.956; p 1⁄4 0.003; Fig. 2D).

3.5. CSF coenzyme Q10 in MSA subgroups (MSAp/MSAc; possible/ probable MSA)

CSF coenzyme Q10 levels did not differ between MSAp and MSAc (Suppl. Table 1; Suppl. Fig. 1A), nor did between possible and probable MSA cases (Suppl. Table 2; Suppl. Fig. 1B). Exclusion of possible MSA cases, due to the greater diagnostic uncertainty of this category, still resulted in lowered CSF coenzyme Q10 in probable MSA both in global (p 1⁄4 0.023) and pair-wise comparisons (MSA vs. PD, p 1⁄4 0.033; MSA vs. PSP, p 1⁄4 0.006; MSA vs. controls, p 1⁄4 0.052). As additional data, through follow-up since inclusion in the CMSAR all cases initially classified as possible MSA still have MSA as their most likely clinical diagnosis as per expert clinical opinion (n 1⁄4 5) or have been reclassified as probable MSA (n 1⁄4 2).

4. Discussion

In this pilot study we report on CSF levels of coenzyme Q10 as a candidate biomarker of MSA. The overlap in plasma and serum coenzyme Q10 levels in two previous studies [8,9], along with the dependence on other peripheral modifiers or confounders, as cholesterol [8] led us to hypothesize that CSF might be an alter- native and reliable source to assess this candidate MSA biomarker. Accordingly, we have found significant lowering of CSF coenzyme Q10 levels in MSA vs. other parkinsonisms and controls, with this association being independent of potential confounders present in the study cohort due to the convenience design and the intrinsic differences across the studied conditions.

Although the search for specific biomarkers of degenerative parkinsonisms is also being directed to atypical parkinsonisms, including MSA [23], most of the existing evidence comes from studies focused in PD, occasionally including subgroups with MSA and PSP as comparators. Still, several studies have been published on possible CSF biomarkers of MSA (for a review see Ref. [23]). Most of those have focused on a-synuclein and markers of axonal loss (neurofilament), with the limitations of inconsistent findings and overlapping values (similar CSF a-synuclein profiles in PD and MSA as synucleinopathies; similar CSF neurofilament findings in MSA and PSP as atypical and more agressive parkisonisms). The currently available knowledge of a link between coenzyme Q10 and MSA with potential pathophysiological implications has opened the possibility of exploring coenzyme Q10 as a candidate MSA-specific biomarker. While the genetic link between COQ2 and MSA in East Asian populations has not been replicated in American and Euro- pean cohorts [24,25], and familial cases are extraordinary in MSA, previously another protein (a-synuclein) has proven relevant to sporadic PD in spite of mutations or multiplications of its gene

Table 1

Demographic, clinical and CSF data across the study groups.
MSA PD

PSP
(n 1⁄4 10) (n1⁄415)

Y. Compta et al. / Parkinsonism and Related Disorders 46 (2018) 16e23 19

Controls p-value

Sex
(women)
Age at onset
(y)
Age at inclusion (y)

Disease duration (mo) Hoehn &Yahr stage > III UMSARS
UPDRS-III

MMSE MDRS-2

LEDD
(in mg/d) CSF coQ10 (in ng/mL)

(n 1⁄4 20) (n1⁄415)

10 6
(50%) (40%)
60.00 59.00 [51.25e63.00] [50.00e61.00] 64.00 71.00 [60.25e66.25] [65.00e76.00] 54.00 120.00 [36.00e81.00] [84.00e180.00] 13 4
(65%) (27%)
66 NA
[42e76]
NA 27

[22e36] 28 28

[28e30] [26e30]
134 135
[132e143] [132e138] 425.00 1250.00 [262.50e838.75] [610.00e1600.00] 22.99 46.96 [17.95e30.73] [27.64e52.40]

8 0.896

5
(50%) (53%)

60.00 [57.50e63.00] 64.00

NA 0.767

77.00 0.001a [61.75e67.00] [72.00e78.00]

NA 0.003b NA 0.039c NA NA NA NA

30 0.139 [24e29] [28e30]

48.00 [36.00e75.00] 7
(70%)
NA

NA 525.00

NA 26

NA 0.675

NA 0.021d

31.10 0.003e [35.81e52.00] [25.63e49.79]

[0.00e900.00] 47.67

NA 1⁄4 not applicable or not available.
The p-values refer to global comparisons as per Kruskal-Wallis test.
The p-values of pairwise comparisons as per Mann-Whitney’s U test with Bonferroni correction (only where global comparisons resulted in significant differences) between MSA and each of the other groups were as follows.

. a MSA vs. PD, p 1⁄4 0.003; MSA vs. PSP, p 1⁄4 1.000; MSA vs. controls, p 1⁄4 0.003.

. b MSA vs. PD, p 1⁄4 0.002; MSA vs. PSP, p 1⁄4 1.000.

. c MSA vs. PD, p 1⁄4 0.082; MSA vs. PSP, p 1⁄4 1.000.

. d MSA vs. PD, p 1⁄4 0.016; MSA vs. PSP, p 1⁄4 1.000.

. e MSA vs. PD, p 1⁄4 0.033; MSA vs. PSP, p 1⁄4 0.003; MSA vs. controls, p 1⁄4 0.036.

being extremely rare.
The results are in keeping with the a priori hypothesis that CSF

might be an alternative and reliable source to detect coenzyme Q10, less likely to carry the limitations of peripheral biofluids of being influenced by systemic confounders. Thus, we have detected co- enzyme Q10 in all cases (both diseased and controls), with their reduction in MSA being statistically significant, without need for adjust for any other analyte or metabolite, and proving indepen- dent of several potential modifiers. However, caution is needed when interpreting our findings, since the studied cohort is rela- tively small, the study lacks a validation cohort, and the intra- and inter-site variabilities of the determination of CSF coenzyme Q10 levels remain to be further characterized.

It is to be acknowledged that our groups significantly differed in age at inclusion (among other variables), with PD and particularly controls being older, due to the convenience design of the study, where MSA cases included in the CMSAR project have been compared to PD, PSP and control cases previously collected as part of the MDBC-HCB, which has not allowed for perfect matching of all groups for age. Nevertheless, the adjusted regression models have partly compensated for the limitation of different age and it might be further speculated that such a difference in age, rather than accounting for the results, might even have resulted in an under- estimation of CSF coenzyme Q10 lowering in MSA, since reduction of coenzyme Q10 with aging has been consistently reported in previous studies [26,27]. Future studies with cases strictly matched for age will be needed to fully rule out (or not) an effect of age on the observed differences in CSF coenzyme Q10 levels in MSA vs. other parkinsonisms and controls.

Remarkably, CSF coenzyme Q10 levels were not significantly correlated to any demographic or clinical variable, including motor and cognitive scores. By contrast, CSF coenzyme Q10 levels were equally lowered in possible and probable MSA cases, suggesting

that the reduction of CSF coenzyme Q10 already exists in cases not severe enough to meet the probable MSA definition. However, the present preliminary study has not focused on early MSA cases and thus it remains to be seen whether CSF coenzyme Q10 levels might be an early diagnostic biomarker, rather than a marker of pro- gression, and if this biomarker might even assist the differential diagnosis of clinically unclassifiable parkinsonisms. The prospec- tive longitudinal phase of the CMSAR project as well as other ongoing multicenter atypical parkinsonisms cohorts will provide valuable information on this.

Similarly, CSF levels of coenzyme Q10 did not differ between MSAp and MSAc. The available neuropathological evidence of reduced coenzyme Q10 in MSA is restricted to the cerebellum, albeit only a reduced number of brain areas have been assessed, namely frontal cortex and the cerebellum in one study, and the cerebellum, the striatum and the occipital cortex in the other one [5,6]. The fact that CSF coenzyme Q10 levels are similar in both MSAc and MSAp could be due to pathological involvement of the cerebellum in MSAp. Alternatively, coenzyme Q10 reduction might be more widespread within the brain than previously shown. Further neuropathological studies of coenzyme Q10 levels in brain areas not assessed to date shall address this.

Our findings and those of the recent genetic, neuropathological and peripheral blood studies [4e6,8,9], raise the question of whether coenzyme Q10 might be a therapeutic target for MSA. This is a complex issue, since this protein has been implicated in, and proposed as a potential treatment for, several different neurode- generative conditions, but to date coenzyme Q10 supplementation proved unsuccessful in PD [28] and PSP [29]. However, given the available evidence [4e9] it is conceivable that coenzyme Q10 deficiency might be more specifically related to MSA than to other neurodegenerative entities. In such setting, subjects carrying COQ2 mutations, or those without mutations but with biochemical

20 Y. Compta et al. / Parkinsonism and Related Disorders 46 (2018) 16e23

Fig. 1. Box plots of CSF coenzyme Q10 levels in MSA, PD, PSP and controls, showing significant differences in the global comparison (Kruskal-Wallis test) due to significant decrease in MSA relative to the other groups in pair-wise comparisons (Mann-Whitney’s U test with Bonferroni’s correction).

Table 2

Binary logistic regression models with MSA diagnosis as outcome (dependent dichotomic variable) and CSF coenzyme Q10 levels in ng/mL and different demographic and clinical variables as potential predictors (independent variables) in univariate and multivariate models, both in the entire cohort (n 1⁄4 60) and in disease-groups (n 1⁄4 45) (please, note that variables only present in disease groups were only assessable in this subset of the cohort).

CSF coenzyme Q10 levels (in ng/mL)
Age at inclusion
(in years)

Sex
(women) Disease duration (in months) Hoehn &Yahr stage > III
LEDD
(in mg/d)

a Statistically significant.

0.891
(n 1⁄4 60) 0.784
(n 1⁄4 60) 1.526
(n 1⁄4 60)

Univariate (entire cohort or disease- groups, depending on the variable)

Covaried (entire cohort)

Covaried (disease groups)

OR 95% CI

0.916 0.869e0.965 (n1⁄460)
0.854 0.775e0.941 (n1⁄460)

1.105 0.378e3.235 (n1⁄460)
0.989 0.974e1.004 (n1⁄445)

0.423 0.126e1.421 (n1⁄445)
0.999 0.997e1.000 (n1⁄445)

p-value

95% CI

0.827e0.960 0.666e0.922 0.318e7.321 NA

NA NA

p-value OR

0.002a 0.851 (n1⁄445)

0.003a 0.804 (n1⁄445)

0.598 1.885 (n1⁄445)

NA 0.980 (n1⁄445)

NA 0.176 (n1⁄445)

NA 1.000 (n1⁄445)

95% CI

0.760e0.954 0.608e1.064 0.182e19.504 0.944e1.017 0.016e1.986 0.998e1.003

p-value

0.006a 0.127 0.595 0.279 0.160 0.762

0.001a
0.001a
0.855
0.148 NA 0.164 NA 0.096 NA

evidence of defective coenzyme Q10, might be candidates for co- enzyme Q10 supplementation. In this vein, recently one such case with MSA carrying compound heterozygous COQ2 mutations has been reported to remain clinically stable and to show an increase in CSF and plasma coenzyme Q10 levels after a three-year course of ubiquinol (a reduced form of coenzyme Q10) [30].

Besides the already discussed limitations, most importantly the fact that indeed this is a pilot study, it is also likely to be statistically underpowered for the smaller subgroups analyses (MSAp vs. MSAc;

probable vs. possible MSA) and lacks neuropathological confirma- tion and longitudinal data, particularly regarding the cases with lesser diagnostic certainty (possible MSA). We have not genotyped our cases for COQ2 mutations either. However, all the included MSA cases are sporadic and COQ2 mutations have not been previously found in our area (compared to East Asian regions) [4,24,25], and are being prospectively followed, diagnostically reassessed, and offered to be brain donors along with additional cases as part of the prospective Catalan MSA Registry, which envisages recruiting and

Y. Compta et al. / Parkinsonism and Related Disorders 46 (2018) 16e23 21

Fig. 2. ROC curves with the respective AUCs and their 95% CIs for the lowering of coenzyme Q10 levels as a diagnostic marker of MSA: A) Comparison of MSA vs. the other groups altogether; B) Comparison of MSA vs. PD; C) Comparison of MSA vs. PSP; D) Comparison of MSA vs. controls.

studying up to 100 MSA cases. Also, we have not determined CSF cholesterol levels and therefore not controlled CSF coenzyme Q10 levels for this parameter. However, it is noteworthy that in a pre- vious study using serum it was necessary to control for cholesterol levels for significant differences to emerge [8], rather than to ensure that the coenzyme Q10 findings were independent of this variable. Thus, it can be speculated that cholesterol appears to mask coenzyme Q10 differences rather than to drive them, which in our case might have hypothetically resulted in even clearer differences should we have controlled for it. Still, this is a conjecture and will certainly need to be addressed in future studies measuring both CSF

coenzyme Q10 and cholesterol levels, including our cohort once it has been enriched with more cases.

The strengths of this report are being the first one on CSF levels of coenzyme Q10 in MSA comparing it not only to controls and PD, but also to PSP, another challenging differential diagnosis of PD and MSA, along with the fact that collection and sample processing protocol was identical for all the cases and the analyses were made in a centralized way at the same lab, using two ELISA kits with the same batch number. Moreover, the adjusted regression models have compensated in part for some baseline inter-group differ- ences, most importantly elder age in PD and controls.

22 Y. Compta et al. / Parkinsonism and Related Disorders 46 (2018) 16e23

In summary, in our cohort of MSA, PD, PSP and control subjects, CSF coenzyme Q10 levels have shown a significant reduction in MSA patients vs. all the other participants, independently of po- tential modifiers and variables differing among study groups at baseline, and with significant diagnostic potential. Therefore CSF coenzyme Q10 levels warrant further assessment in larger and prospective studies as a specific MSA biomarker.

Author contributions

Study concept and design and/or acquisition, analysis, or inter- pretation of data: Compta, Giraldo, Mun~oz, Antonelli, Martí.

Drafting of the manuscript or revising it critically for important intellectual content: All authors.

Final approval of the version to be submitted: All authors.

Conflict of interest
The authors report no conflicts of interest relevant to this study.

Role of the funding source

The funding source had no role in the design and conduct of the study, the collection, management, analysis, or interpretation of the data and the preparation, review, nor the approval of the manu- script; and decision to submit it for publication.

Acknowledgements

The authors are most grateful to all the study participants for their generosity and goodwill. The CMSAR is funded by “Fundacio la Marato de TV3” (PI043296; PI: Dr. M.J. Marti).

Appendix. Members of the CMSAR

4 5 Asuncio n Avila (MD; PhD) , Angels Baye s (MD; PhD) , Teresa

Botta-Orfila (PhD)6, Núria Caballol (MD)7, Matilde Calopa (MD)8, Jaume Campdelacreu (MD; PhD)8, Mario Ezquerra (PhD)1, Oriol de

91 Fa bregues (MD; PhD) , Rube n Ferna ndez-Santiago (PhD) , Jorge

Herna ndez-Vara (MD; PhD)9, Serge Jauma (MD)8, Domenica Marchese (PhD)6, Javier Pagonabarraga (MD; PhD)10, Pau Pastor (MD; PhD)11, Lluís Planellas (MD)1, Claustre Pont-Sunyer (MD, PhD)12, Víctor Puente (MD)13, Montserrat Pujol (MD)14, Josep Saura (PhD)15, Gian Gaetano Tartaglia (PhD)6, Eduard Tolosa (MD; PhD)1, Francesc Valldeoriola (MD; PhD)1.

1. Parkinson’s disease and Movement Disorders Unit, Neurology Service, ICN, Hospital Clínic, IDIBAPS, CIBERNED, University of Barcelona, Barcelona, Catalonia, Spain.

4. Neurology Service Hospital General de l’Hospitalet, Consorci Sanitari Integral, L’Hospitalet de Llobregat, Barcelona, Catalonia, Spain.

5. Parkinson’s disease and Movement Disorders Unit, Clínica Teknon, Barcelona, Catalonia, Spain.

6. Gene function and evolution group; Centre for genomic regulation (CRG), Institucio Catalana de Recerca i Estudis Avançats (ICREA) Universitat Pompeu Fabra (UPF), Barcelona, Catalonia, Spain.

7. Department of Neurology, Hospital Sant Joan Despí Moise s Broggi, Consorci Sanitari Integral, Barcelona, Catalonia, Spain.

8. Parkinson’s disease and Movement Disorders Unit, Neurology Service, Hospital de Bellvitge, Hospitalet de Llobregat, Catalonia, Spain.

9. Parkinson’s disease and Movement Disorders Unit, Neurology Service, Hospital del Vall d’Hebron, Barcelona, Catalonia, Spain.

Neurology Service. Hospital de la Creu i Sant Pau, Barcelona, Cata- lonia, Spain.

11. Neurology Service, Hospital Mútua de Terrassa, Terrassa, Catalonia, Spain.

12. Neurology Service, Hospital General de Granollers, Catalonia, Spain.

13. Neurology Service, Hospital del Mar, Barcelona, Catalonia, Spain.

14. Neurologia Service, Hospital de Santa Maria, Lleida, Cata- lonia, Spain.

15. Biochemistry and Molecular Biology Unit, School of Medi- cine, Neuroradiology Section, Magnetic Resonance Unit, Centre de IDIBAPS, University of Barcelona, Barcelona, Catalonia, Spain.

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.010.

References

[1] A. Fanciulli, G.K. Wenning, Multiple-system atrophy, N. Engl. J. Med. 372 (2015) 249e263.

[2] A.J. Hughes, S.E. Daniel, Y. Ben-Shlomo, A.J. Lees, The accuracy of diagnosis of parkinsonian syndromes in a specialist movement disorder service, Brain 125 (2002) 861e870.

[3] S. Koga, N. Aoki, R.J. Uitti, J.A. van Gerpen, W.P. Cheshire, K.A. Josephs, Z.K. Wszolek, J.W. Langston, D.W. Dickson, D.L.B. When, PD, and PSP masquerade as MSA: an autopsy study of 134 patients, Neurology 85 (2015) 404e412.

[4] Multiple-System Atrophy Research Collaboration, Mutations in COQ2 in fa- milial and sporadic multiple-system atrophy, N. Engl. J. Med. 369 (2013) 233e244.

[5] L.V. Schottlaender, C. Bettencourt, A.P. Kiely, A. Chalasani, V. Neergheen, J.L. Holton, I. Hargreaves, H. Houlden, Coenzyme Q10 levels are decreased in the cerebellum of multiple-system atrophy patients, PLoS One 11 (2016) e0149557.

[6] E.Barca,G.Kleiner,G.Tang,M.Ziosi,S.Tadesse,E.Masliah,E.D.Louis,P.Faust, U.J. Kang, J. Torres, E.P. Cortes, J.P. Vonsattel, S.H. Kuo, C.M. Quinzii, Decreased coenzyme Q10 levels in multiple system atrophy cerebellum, J. Neuropathol. Exp. Neurol. 75 (2016) 663e672.

[7] K. Ubhi, P.H. Lee, A. Adame, C. Inglis, M. Mante, E. Rockenstein, N. Stefanova, G.K. Wenning, E. Masliah, Mitochondrial inhibitor 3-nitroproprionic acid en- hances oxidative modification of alpha-synuclein in a transgenic mouse model of multiple system atrophy, J. Neurosci. Res. 87 (2009) 2728e2739.

[8] T. Kasai, T. Tokuda, T. Ohmichi, R. Ishii, H. Tatebe, M. Nakagawa, T. Mizuno, Serum levels of coenzyme Q10 in patients with multiple system atrophy, PLoS One 11 (2016) e0147574.

[9] J. Mitsui, T. Matsukawa, T. Yasuda, H. Ishiura, S. Tsuji, Plasma coenzyme Q10 levels in patients with multiple system atrophy, JAMA Neurol. 73 (2016) 977e980.

[10] E.R. Peskind, R. Riekse, J.F. Quinn, J. Kaye, C.M. Clark, M.R. Farlow, C. Decarli, C. Chabal, D. Vavrek, M.A. Raskind, D. Galasko, Safety and acceptability of the research lumbar puncture, Alzheimer. Dis. Assoc. Disord. 19 (2005) 220e225.

[11] S. Gilman, G.K. Wenning, P.A. Low, D.J. Brooks, C.J. Mathias, J.Q. Trojanowski, N.W. Wood, C. Colosimo, A. Dürr, C.J. Fowler, H. Kaufmann, T. Klockgether, A. Lees, W. Poewe, N. Quinn, T. Revesz, D. Robertson, P. Sandroni, K. Seppi, M. Vidailhet, Second consensus statement on the diagnosis of multiple system atrophy, Neurology 71 (2008) 670e676.

[12] A.J. Hughes, S.E. Daniel, L. Kilford, A.J. Lees, Accuracy of clinical diagnosis of idiopathic Parkinson’s disease: a clinico-pathological study of 100 cases, J. Neurol. Neurosurg. Psychiatry 55 (1992) 181e184.

[13] I. Litvan, Y. Agid, D. Calne, G. Campbell, B. Dubois, R.C. Duvoisin, C.G. Goetz, L.I. Golbe, J. Grafman, J.H. Growdon, M. Hallett, J. Jankovic, N.P. Quinn, E. Tolosa, D.S. Zee, Clinical research criteria for the diagnosis of progressive supranuclear palsy (Steele-Richardson-Olszewski syndrome): report of the NINDS-SPSP international workshop, Neurology 47 (1996) 1e9.

[14] Y.Compta,M.J.Martí,N.Ibarretxe-Bilbao,C.Junque ,F.Valldeoriola,E.Mun~oz, M. Ezquerra, J. Ríos, E. Tolosa, Cerebrospinal tau, phospho-tau, and beta- amyloid and neuropsychological functions in Parkinson’s disease, Mov. Dis- ord. 24 (2009) 2203e2210.

[15] C. Luk, Y. Compta, N. Magdalinou, M.J. Martí, G. Hondhamuni, H. Zetterberg, K. Blennow, R. Constantinescu, Y. Pijnenburg, B. Mollenhauer, C. Trenkwalder, J. Van Swieten, W.Z. Chiu, B. Borroni, A. Ca mara, P. Cheshire, D.R. Williams, A.J. Lees, R. de Silva, Development and assessment of sensitive immuno-PCR assays for the quantification of cerebrospinal fluid three- and four-repeat tau isoforms in tauopathies, J. Neurochem. 123 (2012) 396e405.

[16] M.M. Hoehn, M.D. Yahr, Parkinsonism: onset, progression, and mortality,

10. Parkinson’s disease and movement disorders unit.

Neurology 17 (1967), 427e427.

. [17] G.K. Wenning, F. Tison, K. Seppi, C. Sampaio, A. Diem, F. Yekhlef, I. Ghorayeb,
F. Ory, M. Galitzky, T. Scaravilli, M. Bozi, C. Colosimo, S. Gilman, C.W. Shults, N.P. Quinn, O. Rascol, W. Poewe, Multiple system atrophy study group, development and validation of the unified multiple system atrophy rating scale (UMSARS), Mov. Disord. 19 (2004) 1391e1402.

. [18] S. Fahn, R.L. Elton, Members of the UPDRS development committee, unified Parkinson’s disease rating scale, in: S. Fahn, C.D. Marsden, D.B. Calne, A. Lieberman (Eds.), Recent Developments in Parkinson’s Disease. Florham Park, NJ, McMillan Health Care Information, 1987, pp. 153e163.

. [19] M.F. Folstein, S.E. Folstein, P.R. McHugh, “Mini-mental state”. A practical method for grading the cognitive state of 735 patients for the clinician, J. Psychiatr. Res. 12 (1975) 189e198.

. [20] S. Mattis, Dementia Rating Scale: Professional Manual, Psychological Assess- ment Resources, Odessa, FL, 1988.

. [21] R. Wenzelburger, B.R. Zhang, S. Pohle, S. Klebe, D. Lorenz, J. Herzog, H. Wilms, G. Deuschl, P. Krack, Force overflow and levodopa-induced dyskinesias in Parkinson’s disease, Brain 125 (2002) 871e879.

. [22] T.M. Marques, H.B. Kuiperij, I.B. Bruinsma, A. van Rumund, M.B. Aerts, R.A. Esselink, B.R. Bloem, M.M. Verbeek, MicroRNAs in cerebrospinal fluid as potential biomarkers for Parkinson’s disease and multiple system atrophy, Mol. Neurobiol. (2016 Nov 14), https://doi.org/10.1007/s12035-016-0253- 0 [Epub ahead of print] PubMed PMID: 27844283.

. [23] B. Laurens, R. Constantinescu, R. Freeman, A. Gerhard, K. Jellinger, A. Jeromin, F. Krismer, B. Mollenhauer, M.G. Schlossmacher, L.M. Shaw, M.M. Verbeek, G.K. Wenning, K. Winge, J. Zhang, W.G. Meissner, Fluid biomarkers in multiple system atrophy: a review of the MSA Biomarker Initiative, Neurobiol. Dis. 80 (2015) 29e41.

[24] L.V. Schottlaender, H. Houlden, Multiple-system atrophy (MSA) Brain Bank collaboration, mutant COQ2 in multiple-system atrophy, N. Engl. J. Med. 71 (2014) 81.

[25] M. Sharma, G. Wenning, R. Krüger, European multiple-system atrophy study group (EMSA-SG), mutant COQ2 in multiple-system atrophy, N. Engl. J. Med. 371 (2014) 80e81.

[26] G. Ravaglia, P. Forti, F. Maioli, R.C. Scali, F. Boschi, A. Cicognani, P. Morini, A. Bargossi, G. Gasbarrini, Coenzyme Q10 plasma levels and body composition in elderly males, Arch. Gerontol. Geriatr. 22 (S1) (1996) 539e543.

[27] P. Niklowitz, S. Onur, A. Fischer, M. Laudes, M. Palussen, T. Menke, F. Do€ring, Coenzyme Q10 serum concentration and redox status in European adults: influence of age, sex, and lipoprotein concentration, J. Clin. Biochem. Nutr. 58 (2016) 240e245.

[28] A. Negida, A. Menshawy, G. El Ashal, Y. Elfouly, Y. Hani, Y. Hegazy, S. El Ghonimy, S. Fouda, Y. Rashad, Coenzyme Q10 for patients with Parkinson’s disease: a systematic review and meta-analysis, CNS. Neurol. Disord. Drug. Targets 15 (2016) 45e53.

[29] D. Apetauerova, S.A. Scala, R.W. Hamill, D.K. Simon, S. Pathak, R. Ruthazer, D.G. Standaert, T.A. Yacoubian, CoQ10 in progressive supranuclear palsy: a randomized, placebo-controlled, double-blind trial, Neurol. Neuroimmunol. Neuroinflamm 3 (2016) e266.

[30] J. Mitsui, K. Koguchi, T. Momose, M. Takahashi, T. Matsukawa, T. Yasuda, S.I. Tokushige, H. Ishiura, J. Goto, S. Nakazaki, T. Kondo, H. Ito, Y. Yamamoto, S. Tsuji, Three-year follow-up of high-dose ubiquinol supplementation in a case of familial multiple system atrophy with compound heterozygous COQ2 mutations, Cerebellum (2017 Feb 1), https://doi.org/10.1007/s12311-017- 0846-9.

Y. Compta et al. / Parkinsonism and Related Disorders 46 (2018) 16e23 23

Parkinsonism and Related Disorders 46 (2018) 24e29

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Association between abnormal nocturnal blood pressure profile and dementia in Parkinson’s disease*

Ryota Tanaka 1, Yasushi Shimo 1, Kazuo Yamashiro, Takashi Ogawa, Kenya Nishioka, Genko Oyama, Atsushi Umemura, Nobutaka Hattori*

Department of Neurology, Juntendo University, 2-1-1, Hongo, Bunkyo Ward, Tokyo, 113-0033, Japan

articleinfo abstract

Article history:

Received 19 December 2016 Received in revised form
12 September 2017 Accepted 18 October 2017

Keywords:

Parkinson’s disease
Risk of dementia
Ambulatory blood pressure monitoring Riser pattern
Orthostatic hypotension

Background: Circadian blood pressure alterations are frequently observed in Parkinson’s disease, but the association between these changes and dementia in the condition remains unclear. Here, we assess the relationship between abnormal nocturnal blood pressure profiles and dementia in Parkinson’s disease. Methods: We enrolled 137 patients with Parkinson’s disease, who underwent 24 h ambulatory blood pressure monitoring, following cognitive and clinical assessment.

Results: Twenty-seven patients (19.7%) were diagnosed with dementia in this cohort. We observed significant associations of dementia with age, male gender, Hoehn-Yahr (H-Y) stage, diabetes mellitus, history of stroke, presence of cerebrovascular lesions on MRI, and orthostatic hypotension. Univariate logistic regression analysis showed that among the patterns of nocturnal blood pressure profiles, the riser pattern was significantly associated with dementia (OR 11.6, 95%CI: 2.14e215.0, P < 0.01), and this trend was observed after adjusting for all confounding factors except orthostatic hypotension (OR 19.2, 95%CI: 1.12e1960.3, P 1⁄4 0.04). However, coexistence of a riser pattern and orthostatic hypotension was related to a higher prevalence of dementia (45.2%) than was a riser pattern alone (9.5%). Furthermore, coexistence of a riser pattern and orthostatic hypotension was significantly more associated with de- mentia than was a riser pattern alone, even after adjusting for confounders (OR 1625.1, 95%CI: 21.9 e1343909.5, P < 0.01).

Conclusions: Our results suggest a relationship between a riser pattern coexisting with orthostatic hy- potension and dementia in Parkinson’s disease. Further prospective studies are warranted to investigate whether abnormal nocturnal blood pressure profiles predict dementia in Parkinson’s disease.

1. Introduction

Parkinson’s disease (PD) is a neurodegenerative disorder with onset in middle age. It manifests as progressive motor symptoms, including bradykinesia, muscular rigidity, tremor at rest, and postural or gait disturbance [1,2]. These motor symptoms are important therapeutic targets for PD treatment. In contrast, non- motor symptoms such as cognitive decline and autonomic dysfunction are also further important factors affecting the prog- nosis of PD [3,4]. Epidemiological studies have shown that about

* Funding sources for study: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

* Corresponding author.
E-mail address: n_hattori@juntendo.ac.jp (N. Hattori).

1 R.T. and Y.S. equally contribute to this study.

https://doi.org/10.1016/j.parkreldis.2017.10.014

30% of patients with PD develop dementia [4]. Furthermore, de- mentia is associated with worsened quality of life, increased care- giver burden, and a higher risk of mortality [5,6]. Age, older age at onset, akinetic-rigid Parkinson’s subtypes, and non-motor symp- toms, such as visual hallucinations, rapid eye movement sleep behavior disorders, and orthostatic hypotension have been shown to be independent risk factors for cognitive decline in PD [7,8].

Abnormal nocturnal blood pressure (BP) profiles, such as the non-dipper and riser patterns, are frequently observed in PD [9,10]. These are associated with cardiovascular events such as coronary heart disease, stroke, and mortality [11]. Recent evidence demon- strates that abnormal nocturnal BP profiles are associated with cognitive decline in elderly populations [12,13]. Furthermore, nocturnal blood pressure is correlated with burden of cerebrovas- cular lesions on MRI, which are also a risk factor for cognitive decline in PD [14]. However, there have been no studies to assess

the association between abnormal nocturnal BP profiles and de- mentia in PD. In this study, we assessed abnormal blood pressure alteration (ABPM) over 24 h in 137 people with PD. We also analyzed correlations between BP alterations and dementia, as well as associations with cerebrovascular lesions on MRI. Furthermore, we assessed whether the presence of orthostatic hypotension (OH) is associated with certain patterns of nocturnal BP profile and whether it affects cognitive decline.

2. Methods

We conducted a retrospective review of 141 patients with PD admitted to Juntendo University Hospital for diagnostic assess- ment, drug adjustment, or evaluation for deep brain stimulation between January 2014 and July 2016. We excluded patients with PD admitted for the treatment of acute illnesses, such as acute infec- tion, ileus, and heart failure. The diagnosis of PD was made ac- cording to the UK Brain Bank criteria [1]. Following informed consent, patients were assessed for BP patterns using ABPM, and orthostatic hypotension. Of 141 participants, three were excluded for unreliable ambulatory BP data, and one participant was also excluded for absence of OH evaluation. We also calculated the levodopa equivalent daily dose (LEDD) for each participant [15].

The study protocol was approved by the ethics committee of Juntendo University Hospital.

2.1. Ambulatory blood pressure monitoring (ABPM)

All patients underwent ABPM for 24 h using an automated system (FB-270; Fukuda Denshi, Tokyo, Japan). Exclusion criteria included an inability to cooperate with ABPM, and complications resulting from acute illnesses, such as infection, ileus, and heart failure. Measurements were recorded every 30 min during the day and every 60 min at night (10:00PM to 7:00AM) using the oscil- lometric method. BP during sleep (sleep BP) was defined as the average of BP measurements during the time the patient was in bed. BP while awake (awake BP) was defined as the average of BP recordings during the rest of the day. Nocturnal falls in BP were classified as follows: a) riser: a nocturnal BP fall of <0%; b) non- dipper: a fall of !0% and <10%; c) dippers: a fall of between 10% and 20%; and d) extreme-dipper: a fall of >20%.

2.2. Orthostatic hypotension (OH) and supine hypertension (SH)

After at least 15 min resting in the supine position, BP was measured, using an electronic sphygmomanometer (ES-H55; Ter- umo). The first measurement was taken while the patient remained supine, followed by BP in a standing position. OH is defined as a 20 mmHg drop in systolic BP and/or a 10 mmHg drop in diastolic BP within the first 3 min after standing. Among the participants, eight had previously been diagnosed with OH and were already treated with midodrine hydrochloride, droxidopa or fludrocortisone ace- tate. SH was defined as a systolic BP of 140 mmHg or higher, or a diastolic BP of 90 mmHg or higher, when in a supine position.

2.3. Cognitive assessment and diagnosis of dementia

Cognitive function was assessed using the Mini-Mental State Examination (MMSE), and the Hasegawa dementia scale-revised (HDS-R) [16]. We enrolled patients with Parkinson’s disease with dementia (PDD) but not dementia with Lewy bodies (DLB) based on the “1-year rule” [17], and the diagnosis of PDD was based on the diagnostic criteria from the Movement Disorder Society Task Force [18]. Cerebrovascular lesions, such as periventricular hyper- intensity (PVH) and deep white matter hyperintensity (DWMH)

were assessed with magnetic resonance imaging (MRI) using semiquantitative visual scales [19].

2.4. Statistical analysis

Continuous variables were compared with either Student’s t- test or one-way ANOVA with Dunnett’s multiple comparison post hoc test. The frequency of categorical variables was compared with the c2 test. We performed multivariate logistic regression analyses to evaluate the association of dementia with OH and ABPM pa- rameters. Clinical variables that were significant following univar- iate analysis were included. The statistical analyses were performed using the JMP Version 12.0 software program (SAS Inc. Cary, NC, USA). A value of P < 0.05 was considered to be statistically significant.

3. Results

3.1. Baseline demographics and risk for dementia in PD

Table 1 presents baseline demographics and the presence of risk factors for dementia in PD for included patients. Of 137 participants, 27 (19.7%) had dementia. Gender, H-Y stage, diabetes mellitus, a history of stroke, and cerebrovascular lesions were significantly associated with dementia in PD. OH and SH were also associated with dementia. Among the pattern of nocturnal blood pressure fall, the riser pattern was significantly associated with dementia (32.7% in the non-dementia group vs. 59.3% in the dementia group, P 1⁄4 0.01), while the dipper pattern was negatively associated with dementia (23.6% in the non-dementia group vs. 3.7% in the de- mentia group, P 1⁄4 0.02). Average systolic BP (including awake, sleep, and 24-h BP) was significantly elevated in the dementia group compared with the non-dementia group.

3.2. Association between abnormal nocturnal BP fall and dementia in PD

Supplemental Table 1 shows the prevalence of dementia among all types of nocturnal BP fall. The highest prevalence of dementia was in the riser pattern (30.8%), followed by the extreme-dipper (28.6%), the non-dipper (15.7%), and the dipper pattern (3.7%) (Supplemental Table 1, P 1⁄4 0.03). Average sleep systolic and dia- stolic BP was significantly higher in the riser type than in the other nocturnal BP types. Semi quantitative evaluation of cerebrovascular lesions showed that these were also more frequent in the riser pattern, while only PVH in the riser pattern was significantly more severe than in the dipper type (Supplemental Table 1, P < 0.05). Anti-hypertensive medication was prescribed to 25 patients (18.2%), and 10 patients (7.3%) also received pharmacological treatment for OH, including droxidopa, midodrine hydrochloride, and fludrocortisone. However there were no significant differences in ABPM profile between patients receiving different kinds of anti- hypertensive or anti-orthostatic hypotension medication (Supplemental Table 1).

The univariate ORs for dementia in PD were significantly higher in the riser or nocturnal hypertension (non-dipper and riser) pat- terns compared with the dipper pattern (Table 2; OR, 11.6; 95%CI: 2.14e215.0; P < 0.01, and OR, 7.9; 95%CI: 1.55e144.5; P < 0.01, respectively). This difference remained significant even after adjusting for possible confounding factors, such as age, gender, H-Y scale, diabetes, a history of stroke, and cerebrovascular lesions (model 1; OR, 19.2; 95%CI: 1.12e1960.3, P 1⁄4 0.04, and OR, 24.9; 95% CI: 1.38e3663.7; P 1⁄4 0.02, respectively). However, these statistical differences disappeared when OH was included in the model (Table 2; model 2).

R. Tanaka et al. / Parkinsonism and Related Disorders 46 (2018) 24e29 25

R. Tanaka et al. / Parkinsonism and Related Disorders 46 (2018) 24e29

Table 1

Baseline clinical demographics and ABPM profiles.

Basal clinical demographics

Number of participants Gender (female, %)
Age
Family history, (%)

BMI
Disease duration, year
Hoehn and Yahr Scale
Education, year
Levodopa Equvalent Dose (LEDD) Smoking status
Non-smoker
Past smoker
Current smoker
Regular drinker, (%)

Cognitive scale

HDS-R

MMSE

Vascular risk and disease

Hypertension, (%) Dyslipidemia, (%) Diabetes mellitus, (%) Stroke

Coronary artery disease

Peripheral artery disease

Autonomic dysfunction

Orthostataic hypotension, (%)

Supine hypertension

Cerebrovascular lesion

PVH

DWMH

ABPM

Dipper (DI)
Extreme dipper (ED) Non-dipper (ND) Riser (RI)
Noctural HT (ND þ RI) Mean blood pressure Awake SBP (mmHg) Awake DBP (mmHg) Awake HR (/min) Sleep SBP (mmHg) Sleep DBP (mmHg) Sleep HR (/min)
24-h SBP (mmHg) 24-h DBP (mmHg) 24-h HR (/min)

Total

137
74 (54.0%) 64.1 ± 10.5 22 (16.1%) 21.4 ± 3.7 10.9 ± 6.2
3.0 ± 0.9
13.5 ± 2.5 957.2 ± 360.1

97 (70.8%) 30 (21.9%) 10 (7.3%) 25 (18.2%)

26.2 ± 5.4 26.6 ± 4.5

29 (21.1%) 25 (18.2%) 14 (10.2%) 5 (3.7%)

4 (3.0%) 0

69 (50.3%) 19 (14.2%)

0.74 ± 0.75 0.78 ± 0.71

27 (19.7%) 7 (5.1%)
51 (37.2%) 52 (38.0%) 103 (75.2%)

116.8 ± 11.9 78.2 ± 8.9 81.2 ± 9.7 113.0 ± 15.1 72.2 ± 10 67.4 ± 8.7 115.1 ± 12.0 75.8 ± 8.4 76.3 ± 8.7

Dementia ( )

110
66 (60.0%) 62.2 ± 10.2 19 (17.3%) 21.5 ± 3.7 11.1 ± 6.1
2.8 ± 0.8
13.7 ± 2.2 960.8 ± 379.1

80 (72.7%) 21 (19.1%) 9 (8.2%) 18 (16.4%)

28.2 ± 1.8 28.3 ± 1.8

22 (20.0%) 19 (17.3%) 7 (6.4%)
2 (1.8%)

2 (1.8%) 0

44 (40.0%) 10 (9.1%)

0.6 ± 0.7 0.71 ± 0.66

26 (23.6%) 5 (4.6%) 43 (39.1%) 36 (32.7%) 79 (71.8%)

115.7 ± 12.3 77.9 ± 8.9 81.9 ± 9.8 111.5 ± 15.0 71.6 ± 9.5 67.1 ± 8.8 113.9 ± 12.3 75.4 ± 8.3 76.7 ± 8.9

Dementia

27
8 (29.6%)
71.6 ± 8.4
3 (11.1%)
20.7 ± 3.7
10.1 ± 6.7
3.7 ± 0.8
12.8 ± 3.3
942.5 ± 274.7 0.81

17 (63.0%) 0.32 9 (33.3%) 0.11 1 (3.7%) 0.42 7 (25.9%) 0.25

18.1 ± 7.3 <0.01 20.0 ± 6.1 <0.01

7 (25.9%) 0.5
6 (22.2%) 0.55 7 (25.9%) <0.01 3 (11.1%) 0.02 2 (7.4%) 0.12 0

25 (92.6%) <0.01 9 (37.5%) <0.01

1.36 ± 0.64 <0.01 1.12 ± 0.83 <0.01

1 (3.7%) 0.02 2 (7.4%) 0.55 8 (29.6%) 0.36 16 (59.3%) 0.01 24 (88.9%) 0.07

121.1 ± 9.3 0.03 79.7 ± 8.9 0.33 78.6 ± 9.2 0.12 119.4 ± 14.3 0.01 75.1 ± 11.8 0.1 68.5 ± 8.1 0.45 119.9 ± 9.7 0.02 77.6 ± 8.8 0.22 74.9 ± 8.0 0.35

(þ) p

<0.01 <0.01 0.43 0.33 0.45 <0.01 0.11

Data are presented as mean ± s.d. or count (proportion). BMI, body mass index; HDS-R, Hasegawa’s dementia rating scale-revised; MMSE, mini mental state examination; PVH, periventricular hyperintensity; DWMH, deep white matter hyperintensity; SBP, systolic blood pressure; DBP, diastolic blood pressure; HR, heart rate.

Table 2

Multivariable logistic regression analysis for risk of dementia based on the types of nocturnal blood pressure fall.

Dipper
Extreme dipper
Non dipper
Riser
NHT (Riser and Non dipper)

Univariate OR (95% CI)

1 (ref)
10.4 (0.85e251.3) 4.84 (0.82e92.3) 11.6 (2.14e215.0) 7.9 (1.55e144.5)

p value

0.07 0.09 <0.01 <0.01

Multivariate (model 1) OR (95% CI)

1 (ref)
5.24 (0.13e625.2) 21.4 (1.21e2461.1) 19.2 (1.12e1960.3) 24.9 (1.38e3663.7)

p value

0.41 0.03 0.04 0.02

Multivariate (model 2) OR (95% CI)

1 (ref)
0.45 (0.00e150.6) 11.0 (0.30e4543.4) 15.7 (0.46e6020.7) 20.5 (0.48e14939.1)

p value

0.79 0.24 0.16 0.15

OR: odds ratio; CI: confidence interval; model 1: adjusting for age, sex, Hoehn and Yahr Scale, diabetes, history of stroke, and cerebrovascular lesions; model 2: adjusting for model 1 and orthostatic hypotension.

3.3. Co-existing OH with abnormal nocturnal BP fall and risk of dementia in PD

OH has been shown to be a non-motor risk factor for dementia in PD, and may also influence nocturnal BP. Therefore, we assessed

the influence of OH with abnormal nocturnal BP profiles on de- mentia in PD (Table 3). Among patients without OH (n 1⁄4 68), there were no cases of dementia in the dipper, extreme dipper, and non- dipper types. Only the riser type without OH showed any cases of dementia (9.5% of patients; Table 3). Among patients with OH

Table 3

Comparison of cognitive score for each type of nocturnal blood pressure fall with or without OH.

OH ( ) dipper

extreme dipper

4
0
28.8 ± 1.5 29±0.8
0.5 ± 0.58 0.5 ± 0.58 120.8 ± 17.6 79.8 ± 8.5 91.8 ± 10.7 59±8.7 111.8 ± 16.4 73.3 ± 9.1

non dipper

24
0
28.2 ± 2.0 28.7 ± 1.3 0.42 ± 0.58 0.5 ± 0.51 112 ± 10.5 74.6 ± 6.8 107.8 ± 11 69.2 ± 6.3 110.3 ± 10.7 72.5 ± 6.6

riser

21
2 (9.5%)
26.9 ± 4.6
27.2 ± 3.4
0.95 ± 0.8
0.86 ± 0.57* 114.1 ± 11.5 75.5 ± 8.4 120.4 ± 12.3*## 75.8 ± 6.7*## 115.7 ± 11.7 75.1±6

p value

0.20 0.28 0.13 0.046 0.06 0.47 0.17 <0.01 <0.01 0.33 0.67

OH (þ) dipper

8
1 (12.5%) 25.1 ± 9.4 24.9 ± 9.0 0.71 ± 0.95 1.00 ± 1.15 121.1 ± 13.3 82.9 ± 7.8 105.1 ± 9.8 68.1 ± 4.8 115.1 ± 11.9 77.3 ± 6.0

extreme dipper

3
2 (66.7%) 22.0 ± 4.0 25.0 ± 2.0 2.0 ± 1.0## 1.33 ± 0.58 129.0 ± 7.9 91.7 ± 7.8 106.7 ± 24.0 73.3 ± 18.9 119.0 ± 13.5 83.7 ± 12.2

non dipper

27
8 (29.6%)
25.2 ± 6.7 25.3 ± 5.5 0.70 ± 0.67 0.89 ± 0.8 119.2 ± 12.7 79.9 ± 9.1 114.0 ± 12.2## 73.6 ± 9.1## 117.0 ± 12.5 77.2 ± 8.9

riser

31
14 (45.2%)
23.9 ± 6.5# 25.0 ± 5.4# 0.97 ± 0.73 1.03 ± 0.73# 117.6 ± 12.0 78.1 ± 10.5 123.6 ± 13.9*## 77.4 ± 11.2## 119.2 ± 12.4 77.7 ± 10.2

p value

0.19 0.81 1.00 0.04 0.79 0.47 0.12 <0.01 0.16 0.84 0.73

R. Tanaka et al. / Parkinsonism and Related Disorders 46 (2018) 24e29

Dementia, (%) HDS-R
MMSE
PVH

DWMH Awake SBP Awake DBP Sleep SBP Sleep DBP 24-h SBP 24-h DBP

19
0
28.5 ±
28.6 ±
0.47 ±
0.42 ±
115.5 ± 10.3 79.2 ± 7.2 99.8 ± 9.1 64.7 ± 7.6 110.1 ± 9.9 74.1 ± 7.1

1.4 1.8 0.61 0.51

Data are presented as mean ± s.d. or count (proportion). OH, orthostatic hypotension; HDS-R, Hasegawa’s dementia rating scale-revised; MMSE, mini mental state exami- nation; SBP, systolic blood pressure; DBP, diastolic blood pressure; 24-h, 24 h *p < 0.05, **p < 0.01 vs. dipper. #p < 0.05, ##p < 0.01 vs. OH( ) dipper.

(n 1⁄4 69), there was a higher prevalence of dementia in the extreme dipper (66.7%), riser (45.2%), and non-dipper (29.6%) types than in the dipper type (12.5%). We used Dunnett’s test to compare cognitive scores for each ABPM profile with or without OH. There were only significant differences between the riser with OH group and the dipper without OH group (Table 3, HDS-R: 23.9 ± 6.5 vs. 28.5 ± 1.4, P 1⁄4 0.0166; MMSE: 25.0 ± 5.4 vs. 28.6 ± 1.8, P 1⁄4 0.025). Cerebrovascular lesions on MRI, especially DWMH, were more se- vere in the riser group than the other groups. Furthermore, there were significant differences between the riser pattern with or without OH, and in the dipper group without OH (Table 3). The univariate OR for dementia in PD was significantly higher in the riser with OH type than in the dipper without OH type (Table 4, OR, 7.82; 95%CI: 1.83e54.6; P < 0.01). This difference remained signif- icant even after adjusting for possible confounding factors such as age, sex, H-Y scale, diabetes, history of stroke, and cerebrovascular lesions (Table 4, OR, 1625.1; 95%CI: 21.9e1343909.5, P < 0.01).

4. Discussion

Circadian BP alterations have frequently been observed in PD and are associated with several non-motor symptoms, including depression and psychosis [20,21]. Previous studies have reported high frequencies of loss of nocturnal BP fall in up to 31e80% of patients with PD [7,9,10,22], which is similar to our findings. In our cohort, we observed that the riser pattern was an important risk for dementia, and the co-existence of the riser pattern with OH was a stronger risk factor for dementia than this pattern without OH. Abnormal nocturnal BP profiles have been associated with mild cognitive impairment (MCI) in elderly populations without de- mentia, and the incidence of MCI was highest in the riser pattern (50%) than in other nocturnal BP patterns [12]. A recent report also demonstrated that the riser pattern was associated with MCI in patients with heart failure [13]. Nagai et al. has reported that ambulatory systolic BP, especially during sleep, is negatively asso- ciated with total brain volume and cognition, while nocturnal

Table 4

Multivariable logistic regression analysis for risk of dementia in riser pattern with or without OH. Univariate

systolic BP dipping was positively associated with these parameters in elderly subjects with one or more cardiovascular risk factors [23]. Other groups have shown that systolic circadian BP variation, especially nighttime BP increases, are associated with the extent of cerebral white matter lesions (WML) measured with MRI [14]. Such WML are one of the risk factors for cognitive decline in PD [24,25]. Because the severity of cerebrovascular lesions visible on MRI was also greater in the riser pattern than in the dipper pattern among our participants (Supplemental Table 1, Table 3), reduced night time blood pressure fall may influence cognition partly as a result of promotion of cerebrovascular damage in PD. Alzheimer’s pathology is also an important substrate of dementia in PD [26]. In patients with Alzheimer’s disease, higher night time blood pressure than in controls has been observed [27]. Reduced night time blood pres- sure falls is also associated with a greater Ab burden of the posterior cingulate in patients with amnestic MCI than in cases with dipper patterns [28]. These studies support the idea that reduced nocturnal BP falls is associated with cognitive decline and the development of dementia in PD. In early PD, cognitive dysfunction has been related to OH and SH, but not in those with a non-dipper pattern [7]. However, this previous study only enrolled early cases of PD, with a mean duration of motor symptoms of 1.8 years, and subjects had never received antiparkinsonian medication. Furthermore, both non-dipper and riser patterns were included in their non-dipper group (<10% nocturnal decrease). These meth- odological differences might explain why our results differ from these. OH has been associated with significantly lower MMSE scores in PD [29]. Orthostatic BP drop has also been shown to be a strong predictor for dementia in PD, with a 7-fold increase in de- mentia risk for every 10 mmHg decrease in systolic BP [30]. Our study also showed a significantly higher prevalence of OH in PD with dementia than in those without dementia (92.3% vs. 40%, P < 0.0001). The pathophysiological association between OH and cognitive decline in PD has not yet been elucidated. However, a recent study demonstrated that OH is related to regional brain hypoperfusion, which may play a role in cognitive decline in PD

OR (95% CI)

riser without OH 1
riser with OH 7.82 (1.83e54.6)

p value

<0.01

Multivariate OR (95% CI)

1
1625.1 (21.9e1343909.5)

p value

<0.01

OH, orthostatic hypotension; OR, odds ratio; CI, confidence interval; this differences were also significant even after adjusting for possible confounding factors such as age, sex, Hoehn and Yahr Scale, diabetes, history of stroke and cerebrovascular lesions.

28 R. Tanaka et al. / Parkinsonism and Related Disorders 46 (2018) 24e29

[31]. As a result of the higher number of patients with OH in this study, the influence of the riser pattern on the risk of dementia might be weakened after adjustment for OH. Further studies will be needed to investigate which factors may be more associated with a greater cognitive burden in PD.

Although our results demonstrate the influence of nocturnal BP rise and co-existing OH on dementia in PD, our study does have some limitations. First, the sample size was small, such that the number of patients with extreme dipper profiles was also small (two subjects out of seven had dementia). Thus, it is difficult to draw any conclusions about the association between extreme dipper profiles and dementia, and our results should be considered preliminary. Second, a recent study has shown that cognition is also transiently impaired during OH in PD [32]. We tested the cognitive scale with participants in a sitting position, but we did not measure blood pressure in this situation for all patients. Third, whether the relationship between circadian BP alterations and dementia is causative or associative remains unclear. Fourth, there may be some uncertainty about the type of dementia our participants had. It is possible that the type of dementia affects the relationship with nocturnal dipper profiles, and this should be considered in future studies with large samples. Furthermore, we investigated patients in a hospital setting rather than at home, and it is possible inpa- tient/outpatient status may also have an effect on the results. Prospective studies that enroll large numbers of patients with early PD without dementia and follow them over time can investigate the nature of the association between abnormal nocturnal BP profiles and the development of dementia.

In conclusion, abnormal circadian BP fall is common in PD and the riser pattern is significantly associated with dementia. Although prospective study of the pathophysiological role of abnormal BP alteration in PD is warranted, our data emphasize the importance of ambulatory BP monitoring and its synergistic effects with autonomic dysfunction on cognitive decline in PD.

Acknowledgment

We thank Dr. Yuichiro Yano for his technical advice regarding ABPM.

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.014.

Author’s contribution

R.T., Y.S. and N.H. planned this study protocol. R.T., Y.S., K.Y., T.O., K.N., G.O., A.U. were involved in data collection. R.T. and K.Y. were involved in statistical analysis. R.T. wrote the manuscript. Y.S. and N.H. reviewed the manuscript and organized the study.

Disclosure of conflicts of interest

R.T. reports grants from (1) Grants from Japan Society for the Promotion of Science; Grant-in-Aid for Scientific Research (C) Grants (JSPS KAKENSHI Grant Number 16K09699) from Japan Agency for Medical Research and Development, and personal fees from honoraria: not related to the current work: (2) Takeda Phar- maceutical Co.,Ltd., (3) Nippon Boehringer Ingelheim,Co.,Ltd, (4) Dai-Nippon Sumitomo Pharma Co.,Ltd., Japan, (5) Bayer Yakuhin, Ltd, (6) Otsuka Pharmaceutical, Co.,Ltd, Japan, (7) FP Pharmaceu- tical Co., (8) Mitsubishi Tanabe Pharma Co., (9) Pfizer Japan., (10) Novartis Pharma K.K, (11) Sanofi K.K., (12) DAIICHI SANKYO Co.,Ltd., (13) Janssen Pharmaceutical K.K., and (14) SHIONOGI & Co., Ltd.

Y.S. reports grants from (1) Grants from Japan Society for the Promotion of Science; Grant-in-Aid for Scientific Research (C) (JSPS KAKENSHI Grant Number 15K09359), speaker honoraria from (2) Medtronic, (3) Boston Scientific, (4) Novartis Pharma, (5) Nihon Medi Physics, (6) Otsuka Pharmaceutical, Co., Ltd, Japan, (7) Sumitomo Dainippon Pharma Co., Ltd., Japan, and (8) Kyowa Hakko Kirin Co., unrelated to the current work.

K.Y. reports grants from Grants from (1) Grant-in-Aid for Sci- entific Research (C) (JSPS KAKENSHI Grant Number 15K07439), and personal fees from honoraria for work unrelated to the current study: (2) Pfizer Inc, (3) Takeda Pharmaceutical Co.,Ltd., (4) Dai- Nippon Sumitomo Pharma Co.,Ltd., Japan, (5) Bayer Yakuhin, Ltd, (6) Otsuka Pharmaceutical, Co.,Ltd. Japan, and (7) Novartis Pharma K.K.

T.O. reports the following disclosures: None.

K.N. reports the following disclosures: (1) Grant from Japan Society for the Promotion of Science KAKENHI Grant (JSPS KAKENSHI Grant Number 16K09678) not granted for the current study.

G.O. reports grants from (1) Grants from Japan Society for the Promotion of Science; Grant-in-Aid for Young Scientists (B) (JSPS KAKENSHI Grant Number 15K19498), honoraria from work unre- lated to the current study: (2) Medtronic, (3) Boston Scientific, (4) Novartis Pharma, MSD, (5) Nihon Medi Phisics, (6) Otsuka Phar- maceutical,Co., Ltd, Japan (7) Kyowa Hakko Kirin, and (8) Abbvie.

A.U. reports grants from (1) Grants from Japan Society for the Promotion of Science; Grant-in-Aid for Scientific Research (C) (JSPS KAKENSHI Grant Number 26462219), Research Grant from (2) Novartis Pharma K.K, consultancy: (3) Boston Scientific Japan, and honoraria: (4) Medtronic Japan Inc., (5) Novartis Pharma K.K, (6) Kyowa Hakko-Kirin Co.,Ltd., (7) Dai-Nippon Sumitomo Pharma Co.,Ltd., Japan, (8) Otsuka Pharmaceutical Co., Ltd., Japan, (9) GlaxoSmithKleine K.K., for work unrelated to the current study.

N.H. reports grants from (1) Grants from Japan Agency for Medical Research and Development, Japan Society for the Promo- tion of Science; Grant-in-Aid for Scientific Research (B) (JSPS KAKENSHI Grant Number 15H04842), (2) Ministry of Education Culture, Sports,Science and Technology Japan (Grant-in-Aid for Scientific Research on Innovative Areas Grant Number 23111003), (3) Health and Labour Sciences Research Grants (Research on Measures for Intractable Diseases), (4) Japan Society for the Pro- motion of Science and (5) Sumitomo Dainippon Pharma Co. Ltd., Japan, personal fees from consultancy: (6) Hisamitsu Pharmaceu- tical, (7) Dai-Nippon Sumitomo Pharma Co.,Ltd. Expert Testimony: (8) Ono Pharmaceutical Co., Ltd; advisory boards participation: (9) Otsuka Pharmaceutical, Co. Ltd., Japan, (10) Novartis Pharma K.K, (11) Takeda Pharmaceutical Co. Ltd.; honoraria: (12) Glax- oSmithKleine K.K, (13) Nippon Boehringer Ingelheim, Co. Ltd., (14) FP Pharmaceutical Corporation, (15) Dai-Nippon Sumitomo Pharma Co. Ltd., Japan, (16) Eisai Co. Ltd., (17) Kissei Pharmaceutical Com- pany, (18) Nihon Medi-physics Co.,Ltd., (19) Kyowa Hakko-Kirin Co. Ltd., (20) Novartis Pharma K.K, (21) Biogen, Acorda Therapeutics, Inc., (22) Otsuka Pharmaceutical, Co. Ltd., Japan, (23) Janssen Pharmaceutical K.K, (24) Medtronic, Inc., (25) Astellas Pharma Inc., other donations: (26) Astellas Pharma., (27) Eisai Co. Ltd., (28) MSD K.K., (29) Daiichi Sankyo Co. Ltd., (30) Novartis Pharma K.K, (31) Takeda Pharmaceutical Co. Ltd., (32) Nihon Medi-physics Co. Ltd., (33) Dainippon Sumitomo Pharma Co., Ltd.., (34) Pfizer Japan., (35) Bayer Yakuhin, Ltd; donations for the endowment of research de- partments: (36) GlaxoSmithKleine K.K, (37) Nippon Boehringer Ingelheim,Co. Ltd., (38) Dainippon Sumitomo Pharma Co., Ltd.., (39) Eisai Co. Ltd., (40) Kissei Pharmaceutical Co., (41) Janssen Phar- maceutical K.K., (42) Nihon Medi-physics Co. Ltd., (43) Kyowa Hakko-Kirin Co. Ltd., (44) Medtronic., (45) Novartis Pharma K.K., (46) Ono Pharmaceutical Co. Ltd., (47) Mitsubishi Tanabe Pharma

Co., (48) Zaiho Co., (49) Hydrogen Health Medical Labo Co., (50) ABIST Co. Ltd., (51) Melodian Co. Ltd., (52) Daiwa Co. Ltd., (53) Biogen Idec Japan Ltd., (54) Bayer Yakuhin, Ltd, (55) Nihon Phar- maceutical Co. Ltd., (56) Asahi Kasei Medical Co. Ltd., (57) MiZ Co. Ltd., unrelated to the current study.

References

. [1] A.J. Hughes, S.E. Daniel, L. Kilford, A.J. Lees, Accuracy of clinical diagnosis of idiopathic Parkinson’s disease: a clinico-pathological study of 100 cases, J. Neurol. Neurosurg. Psychiatry 55 (1992) 181e184.

. [2] C. Marras, A. Lang, Parkinson’s disease subtypes: lost in translation? J. Neurol. Neurosurg. Psychiatry 84 (2013) 409e415.

. [3] C. McDonald, J.L. Newton, D.J. Burn, Orthostatic hypotension and cognitive impairment in Parkinson’s disease: causation or association? Mov. Disord. 31 (2016) 937e946.

. [4] D. Aarsland, J. Zaccai, C. Brayne, A systematic review of prevalence studies of dementia in Parkinson’s disease, Mov. Disord. 20 (2005) 1255e1263.

. [5] P. Martinez-Martin, C. Rodriguez-Blazquez, M.J. Forjaz, B. Frades-Payo, L. Agüera-Ortiz, D. Weintraub, A. Riesco, M.M. Kurtis, K.R. Chaudhuri, Neuropsychiatric symptoms and caregiver’s burden in Parkinson’s disease. Neuropsychiatric symptoms and caregiver’s burden in Parkinson’s disease, Park. Relat. Disord. 21 (2015) 629e634.

. [6] A.D. Macleod, K.S. Taylor, C.E. Counsell, Mortality in Parkinson’s disease: a systematic review and meta-analysis, Mov. Disord. 29 (2014) 1615e1622.

. [7] J.S. Kim, Y.S. Oh, K.S. Lee, Y.I. Kim, D.W. Yang, D.S. Goldstein, Association of
cognitive dysfunction with neurocirculatory abnormalities in early Parkinson
disease, Neurology 79 (2012) 1323e1331.

. [8] P. Svenningsson, E. Westman, C. Ballard, D. Aarsland, Cognitive impairment in
patients with Parkinson’s disease: diagnosis, biomarkers, and treatment,
Lancet Neurol. 11 (2012) 697e707.

. [9] A. Fanciulli, S. Strano, J.P. Ndayisaba, G. Goebel, L. Gioffre , M. Rizzo,
C. Colosimo, C. Caltagirone, W. Poewe, G.K. Wenning, F.E. Pontieri, Detecting nocturnal hypertension in Parkinson’s disease and multiple system atrophy: proposal of a decision-support algorithm, J. Neurol. 261 (2014) 1291e1299.

. [10] C. Schmidt, D. Berg, Prieur S. Herting, S. Junghanns, K. Schweitzer, C. Globas, L. Scho€ls, H. Reichmann, T. Ziemssen, Loss of nocturnal blood pressure fall in various extrapyramidal syndromes, Mov. Disord. 24 (2009) 2136e2142.

. [11] G.F. Salles, G. Reboldi, R.H. Fagard, C.R. Cardoso, S.D. Pierdomenico, P. Verdecchia, K. Eguchi, K. Kario, S. Hoshide, J. Polonia, A. de la Sierra, R.C. Hermida, E. Dolan, E. O’Brien, G.C. Roush, ABC-H investigators, prognostic effect of the nocturnal blood pressure fall in hypertensive patients: the ambulatory blood pressure collaboration in patients with hypertension (ABC- H) meta-analysis, Hypertension 2016 (67) (2016) 693e700.

. [12] H. Guo, Y. Tabara, M. Igase, M. Yamamoto, N. Ochi, T. Kido, E. Uetani, K. Taguchi, T. Miki, K. Kohara, Abnormal nocturnal blood pressure profile is associated with mild cognitive impairment in the elderly: the J-SHIPP study, Hypertens. Res. 33 (2010) 32e36.

. [13] T. Komori, K. Eguchi, T. Saito, Y. Nishimura, S. Hoshide, K. Kario, Riser blood pressure pattern is associated with mild cognitive impairment in heart failure patients, Am. J. Hypertens. 29 (2016) 194e201.

. [14] Y.S. Oh, J.S. Kim, D.W. Yang, J.S. Koo, Y.I. Kim, H.O. Jung, K.S. Lee, Nighttime blood pressure and white matter hyperintensities in patients with Parkinson disease, Chronobiol Int. 30 (2013) 811e817.

. [15] C.L. Tomlinson, R. Stowe, S. Patel, C. Rick, R. Gray, C.E. Clarke, Systematic re- view of levodopa dose equivalency reporting in Parkinson’s disease, Mov. Disord. 25 (2010) 2649e2653.

. [16] K. Hasegawa, Clinical assessment of cognitive functioning in the aged, Pshy- chopharmacol Bull. 19 (1983) 44e51.

. [17] I.G. McKeith, D.W. Dickson, J. Lowe, M. Emre, J.T. O’Brien, H. Feldman, J. Cummings, J.E. Duda, C. Lippa, E.K. Perry, D. Aarsland, H. Arai, C.G. Ballard,

B. Boeve, D.J. Burn, D. Costa, T. Del Ser, B. Dubois, D. Galasko, S. Gauthier, C.G. Goetz, E. Gomez-Tortosa, G. Halliday, L.A. Hansen, J. Hardy, T. Iwatsubo, R.N. Kalaria, D. Kaufer, R.A. Kenny, A. Korczyn, K. Kosaka, V.M. Lee, A. Lees, I. Litvan, E. Londos, O.L. Lopez, S. Minoshima, Y. Mizuno, J.A. Molina, E.B. Mukaetova-Ladinska, F. Pasquier, R.H. Perry, J.B. Schulz, J.Q. Trojanowski, M. Yamada, Consortium on DLB, Diagnosis and management of dementia with Lewy bodies: third report of the DLB Consortium, Neurology 65 (2005) 1863e1872.

[18] M. Emre, D. Aarsland, R. Brown, D.J. Burn, C. Duyckaerts, Y. Mizuno, G.A. Broe, J. Cummings, D.W. Dickson, S. Gauthier, J. Goldman, C. Goetz, A. Korczyn, A. Lees, R. Levy, I. Litvan, I. McKeith, W. Olanow, W. Poewe, N. Quinn, C. Sampaio, E. Tolosa, B. Dubois, Clinical diagnostic criteria for dementia associated with Parkinson’s disease, Mov. Disord. 22 (2007) 1689e1707.

[19] F. Fazekas, J.B. Chawluk, A. Alavi, H.I. Hurtig, R.A. Zimmerman, MR signal abnormalities at 1.5 T in Alzheimer’s dementia and normal aging, Am. J. Roentgenol. 149 (1987) 351e356.

[20] H.E. Park, J.S. Kim, Y.S. Oh, I.S. Park, J.W. Park, I.U. Song, K.S. Lee, Autonomic nervous system dysfunction in patients with Parkinson disease having depression, J. Geriatr. Psychiatry Neurol. 29 (2016) 11e17.

[21] E. Stuebner, E. Vichayanrat, D.A. Low, C.J. Mathias, S. Isenmann, C.A. Haensch, Non-dipping nocturnal blood pressure and psychosis parameters in Parkinson disease, Clin. Auton. Res. 25 (2015) 109e116.

[22] K. Berganzo, B. Díez-Arrola, B. Tijero, J. Somme, E. Lezcano, V. Llorens, I. Ugarriza, R. Ciordia, J.C. Go mez-Esteban, J.J. Zarranz, Nocturnal hypertension and dysautonomia in patients with Parkinson’s disease: are they related? J. Neurol. 260 (2013) 1752e1756.

[23] M. Nagai, S. Hoshide, J. Ishikawa, K. Shimada, K. Kario, Ambulatory blood pressure as an independent determinant of brain atrophy and cognitive function in elderly hypertension, J. Hypertens. 26 (2008) 1636e1641.

[24] N. Kandiah, E. Mak, A. Ng, S. Huang, W.L. Au, Y.Y. Sitoh, L.C. Tan, Cerebral white matter hyperintensity in Parkinson’s disease: a major risk factor for mild cognitive impairment, Park. Relat. Disord. 19 (2013) 680e683.

[25] M.K. Sunwoo, S. Jeon, J.H. Ham, J.Y. Hong, J.E. Lee, J.M. Lee, Y.H. Sohn, P.H. Lee, The burden of white matter hyperintensities is a predictor of progressive mild cognitive impairment in patients with Parkinson’s disease, Eur. J. Neurol. 21 (2014), 922ee50.

[26] M.N. Sabbagh, C.H. Adler, T.J. Lahti, D.J. Connor, L. Vedders, L.K. Peterson, J.N. Caviness, H.A. Shill, L.I. Sue, I. Ziabreva, E. Perry, C.G. Ballard, D. Aarsland, D.G. Walker, T.G. Beach, Parkinson disease with dementia: comparing patients with and without Alzheimer pathology, Alzheimer Dis. Assoc. Disord. 23 (2009) 295e297.

[27] H.F. Chen, H. Chang-Quan, C. You, Z.R. Wang, W. Hui, Q.X. Liu, H. Si-Qing, The circadian rhythm of arterial blood pressure in Alzheimer disease (AD) patients without hypertension, Blood Press 22 (2013) 101e105.

[28] T. Tarumi, T.S. Harris, C. Hill, Z. German, J. Riley, M. Turner, K.B. Womack, D.R. Kerwin, N.L. Monson, A.M. Stowe, D. Mathews, C.M. Cullum, R. Zhang, Amyloid burden and sleep blood pressure in amnestic mild cognitive impairment, Neurology 85 (2015) 1922e1929.

[29] A.D. Hohler, J.R. Zuzua rregui, D.I. Katz, T.J. Depiero, C.L. Hehl, A. Leonard, V. Allen, J. Dentino, M. Gardner, H. Phenix, M. Saint-Hilaire, T. Ellis, Differences in motor and cognitive function in patients with Parkinson’s disease with and without orthostatic hypotension, Int. J. Neurosci. 122 (2012) 233e236.

[30] J.B. Anang, J.F. Gagnon, J.A. Bertrand, S.R. Romenets, V. Latreille, M. Panisset, J. Montplaisir, R.B. Postuma, Predictors of dementia in Parkinson disease: a prospective cohort study, Neurology 83 (2014) 1253e1260.

[31] A.D. Robertson, M.A. Messner, Z. Shirzadi, G. Kleiner-Fisman, J. Lee, J. Hopyan, A.E. Lang, S.E. Black, B.J. MacIntosh, M. Masellis, Orthostatic hypotension, cerebral hypoperfusion, and visuospatial deficits in Lewy body disorders, Park. Relat. Disord. 22 (2016) 80e86.

[32] J. Centi, R. Freeman, C.H. Gibbons, S. Neargarder, A.O. Canova, A. Cronin- Golomb, Effects of orthostatic hypotension on cognition in Parkinson disease, Neurology 88 (2017) 17e24.

R. Tanaka et al. / Parkinsonism and Related Disorders 46 (2018) 24e29 29

Parkinsonism and Related Disorders 46 (2018) 30e35

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

A prospective study of falls in relation to freezing of gait and response fluctuations in Parkinson’s disease

Yasuyuki Okuma a, Ana Lígia Silva de Lima b, c, Jiro Fukae a, d, Bastiaan R. Bloem c, *, Anke H. Snijders c, e

a Department of Neurology, Juntendo University Shizuoka Hospital, Izunokuni, Japan
b Brazilian Federal Agency for Support and Evaluation of Graduate Education – CAPES, Brazil
c Radboud University Medical Centre, Donders Institute for Brain, Cognition and Behavior, Department of Neurology, Nijmegen, The Netherlands d Department of Neurology, Fukuoka University School of Medicine, Fukuoka, Japan
e Department of Neurology, Pantein Ziekenhuis, Boxmeer, The Netherlands

articleinfo abstract

Article history:

Received 9 October 2016 Received in revised form 9 October 2017
Accepted 18 October 2017

Keywords:

Parkinson’s disease Falls
Motor fluctuation Freezing of gait

1. Introduction

Recurrent falls are a disabling feature of Parkinson’s disease (PD) and have a significant negative impact on the patient’s quality of life [1]. Assessing the impact of falls in patients with PD by only asking about prior falls is often unreliable because of recall bias [2]. Therefore, prospective studies are necessary to clarify the actual circumstances surrounding falls.

Few prospective studies specifically aimed to clarify the falling circumstances and the associated impact of falls in PD [3,4]. These studies showed that most falls occur indoors, and they are usually

* Corresponding author. Department of Neurology (935), Radboud University Medical Centre, PO Box 9101, 6500 HB Nijmegen, The Netherlands.

E-mail address: bas.bloem@radboudumc.nl (B.R. Bloem).

https://doi.org/10.1016/j.parkreldis.2017.10.013

Introduction: Falls are a disabling feature of Parkinson’s disease (PD). In this prospective study we investigated: (1) in which motor state patients with PD fallmost often; and (2) whether freezing of gait (FOG) and dyskinesias contribute to falls.
Methods: Patients with PD who had fallen at least once in the previous year and had wearing-off were recruited. During six months, patients complete a standardized fall report. We analyzed data regarding fall circumstances and motor state at the time of each first 10 falls.

Results: We included 36 patients with PD (34 freezers), with mean ± SD age of 67.5 ± 6.3 years and disease duration of 12.4 ± 4.1 years. 50% had Hoehn & Yahr (HY) 2 at ON-state and 56% had a HY 4 at OFF. All 36 patients fell at least once during the follow-up period (total number of falls: 252; mean ± SD: 19.03 ± 33.9). Falls at ON were 50% of the total falls, followed by Transition (30%) and OFF (20%). Overall, 69% of falls were related to FOG, 28% were unrelated to FOG and 3% were related to dyskinesia. There was a significant relationship between motor state and circumstances (c2(2) 1⁄4 31.496,p < 0.001), showing that FOG-related falls happened mostly at OFF-state.

Conclusion: This study showed that patients with PD fall mostly at ON. Additionally, FOG is an important contributor to falls in patients with PD. This information may assist clinicians in optimizing medication to prevent further falls.

“intrinsic” in nature (i.e. mainly due to patient-related factors). The attributed causes mainly included tripping, postural instability and freezing of gait (FOG), although the latter was not ascertained in a great amount of detail [3,4].

Later work identified FOG as a truly important cause of falls [5,6]. However, the relationship between falls and actual medica- tion state (ON, OFF, Transition) at the time of falling has not been established yet. Moreover, it remains largely unknown whether troublesome dyskinesias can also contribute to falls, by generating excessive trunk movements that cause the patient to sway beyond the limits of stability [7,8]. Clarifying those relationships could improve clinical management of falls in PD [9]. Therefore, the aim of this prospective study was to investigate in which medication state patients with PD fall most often, and whether and how FOG and dyskinesias contribute to falls.

2. Methods

2.1. Subject and study procedure

We studied 36 patients with PD (defined according to the UK Brain Bank Criteria) who all showed the wearing-off phenomenon. Patients were recruited from the outpatient clinic of Juntendo University Shizuoka Hospital and were only eligible if they were ambulant, either with or without walking aids, and had fallen at least once in the previous year. Patients with other causes of postural instability, including orthopedic problems, were excluded. Informed consent was obtained from all participants and the study was approved by the ethical committee in Juntendo University Shizuoka Hospital.

After inclusion, participants underwent a baseline assessment performed by the physician, consisting of Hoehn & Yahr stage and the gait ability item from the Unified Parkinson’s Disease Rating Scale. The presence of FOG was assessed using the question: “Do you sometimes feel that your feet get glued to the floor while walking, making a turn, or when trying to initiate walking?” Motor states were defined according to previous work [10] as follows: ON- state was being in a better condition (i.e. easy to perform move- ments); OFF-state was defined as being in a clearly worse condition (i.e. difficult to in performing movements); and Transition-state was defined as a condition between ON and OFF states. Before starting the follow-up, patients were educated about FOG, dyski- nesias and motor sates. The physician demonstrated a typical FOG episode and typical dyskinesias himself. Furthermore, video clips of various FOG episodes were shown, including episodes of turning hesitation leading to balance loss. ON-, OFF-, and Transition-states were explained as previously described.

Following the baseline assessment, patients were instructed to complete a standardized fall report for each fall, during six months. They were asked to note the time and circumstances (i.e. presence of dual tasking and activities developed) at the time of the fall in their own words. Additionally, subjects ticked pre-specified options regarding the environment where the fall had occurred (indoors or outdoors), and motor state as defined earlier. Patients also recorded the severity of dyskinesias at the time of a fall (0, none; 1, mild; 2, moderate; 3,severe) and the usual daytime duration of dyskinesias on the day the fall occurred (0, none; 1,1e25%; 2, 26e50%; 3, 51e75%; 4, 76e100%). Finally, patients had to indicate specifically whether they considered either excessive dyskinesias or FOG to be the major contributor for that particular fall.

During the prospective follow-up period, patients visited our outpatient clinic once a month, and physicians counted the number of falls and checked the fall reports for completeness. If necessary, patients and caregivers were interviewed and re-educated about describing their fall events at each visit.

2.2. Outcomes and statistical analysis

Reports of the first 10 falls were collected and analyzed to reduce the numerical impact of extraordinary fallers (i.e. patients with a number of falls higher than 100, n 1⁄4 4). Patients were classified either as ON-state fallers, OFF-state fallers, or Transition- state fallers according to the timing at which the majority (i.e. more than 50%) of the first 10 falls had occurred. In cases where this cut- off point was not achieved, participants were classified as “not- clear”. We also classified all falls separately, as either occurring in the ON, OFF, or in the Transition states.

After classification, main outcomes calculated were: total and mean ± SD number of falls (excluding outliers); and distribution of falls in relation to: (1) motor state, (2) FOG, (3) dual-tasking, and (4) indoors or outdoors. To calculate these outcomes, motor state was

considered as whether the patients rated themselves as ON, OFF or Transition state at the moment of the fall. The influence of FOG was determined by evaluating the reasons of fall rated by patients, (i.e. reason for falling: FOG episode, loss of balance without connection to FOG or dyskinesia episode). Indoor falls were considered as all falls that happened inside the patient’s house and outdoor falls were considered any fall that occurred outside the patient’s house (including gardens).

Descriptive statistics was used to express the frequencies of categorical variables and means ± SD of continuous variables. Dif- ferences in frequencies of categorical variables were investigated by Chi-square test.

3. Results

3.1. Sample characteristics

We included 36 patients with PD, with mean age of 67.5 ± 6.3 years, mean disease duration of 12.4 ± 4.1 years and not cognitively impaired (Mini-Mental State Examination higher than 24). Among them, 34 reported FOG episodes. Most patients (n 1⁄4 33) presented FOG in the OFF-state, one patient in the ON-state, and 20 in both states. Dyskinesias were present in 26 patients (all at ON-state) with 3 patients considering it disabling. Hoehn and Yahr (HY) distribution varied between ON and OFF-state. At ON-state, 50% (n1⁄418)presentedHY2,39%(n1⁄414)HY3,8%(n1⁄43)HY4and3% (n1⁄41)HY1.AtOFF,thesenumberswere56%(n1⁄420)HY4,29% (n1⁄410)HY5and18%(n1⁄46)HY3(Table1).All36patientstook levodopa (mean daily dose 551 mg), 30 patients used dopamine agonists (mean daily levodopa equivalent dose 189 mg), 19 patients took selegiline (mean daily dose 5.9 mg) and eleven patients took amantadine (mean daily dose 131.8 mg).

3.2. Classification of fallers

All 36 patients fell at least once during the 6 months follow-up, with the number of falls ranging from 1 to more than 600 per pa- tient. After excluding outliers (n 1⁄4 4), the mean number of falls were 19.03 ± 33.9, and the total number of falls analyzed were then 252 (when analyzing only the first 10 falls per patient).

Based on the first 10 falls, 53% of patients (n 1⁄4 19) fell pre- dominantly at the ON-state, whereas 22% (n 1⁄4 8) fell in between ON- and OFF-state (transition) and 18% (n 1⁄4 6) fell mainly during the OFF-state. For 3 patients, it was not possible to determine in which status they predominantly fell. Regarding type of fallers, two thirds (12 of 19 patients) of ON-state fallers, showed HY stage 3 or 4 in the ON-state. In contrast, only one of six OFF-state fallers showed a HY 3 in the ON-state (Table 1).

3.3. Classification of falls

Overall, 69% (n 1⁄4 173) of the falls were related to a FOG episode (FOG-related), 28% (n 1⁄4 72) were related to a loss of balance not caused by a FOG episode (FOG-unrelated) and 3% (n 1⁄4 7) were caused by dyskinesias. The high incidence of fall episodes related to FOG remained unaltered when dual-tasking was taking into consideration. In this case, while patients were performing dual- tasks, 71 falls were FOG-related, 50 were FOG-unrelated and 4 dyskinesia related. Without dual-tasking, the number of FOG- related falls raises to 102, mainly due to increase of FOG related OFF-state falls, and the number of FOG-unrelated falls decreases to 22 (Table 2).

When the motor status is investigated, it is possible to see that 50% (n 1⁄4 124) of all falls were in the ON-state. Among ON-state falls, 52% (n 1⁄4 65) were FOG-related, while 42% (n 1⁄4 52) were FOG-

Y. Okuma et al. / Parkinsonism and Related Disorders 46 (2018) 30e35 31

32 Y. Okuma et al. / Parkinsonism and Related Disorders 46 (2018) 30e35 Table 1

Distribution of falls per patient.
Patient HY ON (n) HY OFF (n) Total falls (n) ON falls (n)a Transition falls (n)a OFF falls (n)a Number of injuries (n)a

On fallers (n 1⁄4 19)

2 3 5 11 10 0 0 2 3 3 5 44 9 1 0 3 4 3 5 35 8 1 1 0 53464200 6 2 4 17 8 2 0 1 82474210 92533000 10 3 5 175 8 2 0 0 14 2 4 17 6 2 2 0 16 4 5 1 1 0 0 1 17 4 5 3 3 0 0 0 19 2 4 3 2 1 0 0 20 4 5 4 4 0 0 0 22 3 4 3 3 0 0 0 23 3 4 8 8 0 0 1 27 3 5 >600 7 1 2 0 29 3 4 1 1 0 0 0 30 2 3 5 5 0 0 0 32 2 4 6 6 0 0 0

Transition fallers (n 1⁄4 8)

12431200 7 2 4 32 1 8 1 0 12 2 4 7 2 5 0 0 15 2 4 3 0 2 1 0 25 2 4 >600 0 9 1 2 28 3 5 94 3 6 1 3 33 2 4 11 0 8 2 0 36 3 4 8 3 5 0 0

OFF fallers (n 1⁄4 6)

11 2 3 7 0 2 5 1 13 3 4 56 2 1 7 3 21 2 3 13 3 2 5 0 26 1 3 >600 2 0 8 2 34 2 4 12 0 2 8 1 35 2 3 4 0 0 4 1

Not clear (n 1⁄4 3)

18 3 4 4 2 0 2 1 24 3 4 19 5 5 0 1 31 2 3 6 0 3 3 2 a Data collected for the first10 falls if the number of falls is >10.

unrelated and 7 (6%) due to a balance loss related to dyskinesias. ON- falls that were FOG-unrelated happened because patients lost bal- ance when changing posture, reaching something, or opening/closing doors. ON-falls related to FOG, either with or without dual tasking, occurred during gait when initiating walking or during turning.

A total of 75 (30%) falls happened at Transition-state (between ON and OFF), as an end-of-dose phenomenon. In that case, 80% (n 1⁄4 60) of the falls were related to a FOG event, while other 20% (n 1⁄4 15) of the falls were due to a loss of balance unrelated to a FOG episode. Finally, 20% (n 1⁄4 53) of the fall episodes happened at the OFF-state, with 91% (n 1⁄4 48) being related to a FOG episode and 9% (n 1⁄4 5) not related to FOG. FOG-related OFF-state falls frequently occurred when going to the bathroom, particularly at night or early in the morning, mostly without dual tasking (39 of 48 falls, Table 2). In either status (transitional or OFF) no falls related to dyskinesia were registered.

Chi-square analysis showed that there was a significant rela- tionship between motor status and circumstances of the fall (c2(2) 1⁄4 31.496,p < 0.001). The analysis revealed that a higher number of FOG-related falls occurred in the OFF-state, in compar- ison to transition and ON. In contrast, a higher number of FOG-unrelated falls happened in the ON-state in comparison to transition or OFF-state.

Concurrent dyskinesias at the time of a fall were reported by 56% of patients, mostly with mild to moderate severity. Two pa- tients reported that they experienced dyskinesias at the time of the fall that were severe enough to undermine their balance (scores for dyskinesia severity and duration were grade 3).

Nineteen patients (53%) sustained an injury following a fall. Two patients had fractures that required hospital admission. One of these patients sustained a hip fracture, necessitating surgery after a fall that was preceded by a sudden loss of consciousness (syncope). Another patient had a fracture in her shoulder due to a fall caused by a sudden FOG episode while walking in the hallway at the OFF- state. Both of these injurious falls occurred at home (overview of injuries per patient is presented in Table 1 and per group is pre- sented in Table 3).

Indoors falls represented 80% (n 1⁄4 201) of the falls. The landing positions associated with FOG-related falls were predominantly forward (n 1⁄4 122), while FOG-unrelated falls (n 1⁄4 36) and dyski- nesia (n 1⁄4 5) were predominantly backward. Outdoor falls (20%, n 1⁄4 51) showed the same pattern (Table 3).

4. Discussion

This study showed the important contribution of FOG to falls.

Y. Okuma et al. / Parkinsonism and Related Disorders 46 (2018) 30e35 33

Table 2

Distribution of falls

in relation to fall circumstances. Dual tasking present

Standing Walking Turning Transfer
Start hesitation Othera

76 15 14 23 00 78 00

215236 33221 0501 0002 41900 00121

23140 0300 0100 0200 0000 01300 25040

04
00 0013 0031

041 276023 4820 0470 62500 2500 39102154

0500 2710 0210 1810 0000 0000 32230

00 01 01 01 00 00 03

FOG-related

FOG-unrelated OFF Total ON Transition

Dyskinesia-related

Dual tasking absent FOG-related
ON Transition 000

FOG-unrelated OFF Total ON Transition

Dyskinesia-related

ON Transition

OFF Total ON Transition

OFF Total

OFF Total ON Transition

OFF Total

Total 31 31 9 71 37 11
a This category includes: stairs for FOG related falls and squatting for FOG unrelated falls.

00 00 04

109 21 34 29

Table 3

Distribution of falls in relation to environment, landing positions and injuries. Indoors

Outdoors

Backward Forward Sideways Injury

32 43 38 3 5 5 4

38 41 122 1 9
5 14

276 5 6 6 2 2 0

3 36 1 12 1 9 2 4

50 0 0 0 0 0 0

05 0 0 0 0 0 0

10 15 13 0 2 4 1

01 3 31 0 2 1 6

50 5 1 4 0 0 1

05 0 6 0 4 0 1

20 0 0 0 0 0 0

02 0 0 0 0 0 0

FOG-related

FOG-unrelated

Dyskinesia-related

FOG-related

FOG-unrelated

Dyskinesia-related

ON Transition

OFF Total

ON Transition

OFF Total

ON Transition

OFF Total

ON Transition

OFF Total

ON Transition

OFF Total

ON Transition

OFF Total

18 15

34 Y. Okuma et al. / Parkinsonism and Related Disorders 46 (2018) 30e35

Here, FOG was the most frequent cause of falls: accounting for 91% of falls during the OFF-state, 80% of falls during the Transition-state, and 52% of falls during the ON-state. Among the 252 falls, 3% (n 1⁄4 7) were due to dyskinesias at ON-state. Falls at ON were 50% (n 1⁄4 124) of the total falls, followed by falls in Transition (30%, n 1⁄4 75) and OFF (20%, n 1⁄4 53). Furthermore, 53% (n 1⁄4 19) of participants were ON-fallers, 22% (n 1⁄4 8) Transition-fallers and 18% (n 1⁄4 6) OFF- fallers. These findings contribute to elucidate the relationship be- tween falls, FOG, motor state and dyskinesias in patients with PD.

Our first finding is that little more than one-half of patients predominantly fell at the ON-state, confirming earlier work [3]. This may be explained by the fact that patients’ mobility improves at the ON-state. However, because postural instability is generally not improved during the ON-state [3], this combination may lead to falls. Additionally, a novel finding was that many patients fell in the Transition between ON- and OFF-states. During such an end-of- dose state, patients may believe that they can move well and therefore be relatively overconfident, despite a rapid deterioration of their walking ability at that particular time. Improve medication schedule might diminish fluctuations and could tackle falls at Transition-state.

Moreover, we found that FOG is the most frequent cause of falls in patients with PD reporting previous falls, particularly at the OFF- state. Although FOG and falls are likely interconnected, prospective studies only listed FOG as one of many potential causes of falls [3,4,7]. Two reasons may explain this result. First, this high preva- lence of FOG related falls could reflect the much greater awareness for the FOG phenomenon compared to a decade ago, when FOG might have been underestimated and partially overlooked as a cause of falls. For example, falls occurring when turning have usually been considered simply as balance loss, unrelated to FOG [3]. However, in this study, if the patient reported the specific feeling that the feet were glued to the floor at the time of the fall, this fall was determined as a FOG-related fall. Additionally, a second reason may be the bias in patient selection. For instance, by excluding more severely impaired patients and those who had fallen in the previous year, Gazibara et al. (2016) recruited only 40% freezers. Thus, their results showed that only 10% of falls attributed to FOG in single fallers, and 0% of recurrent fallers [12].

A less expected finding was that FOG also contributed to a substantial proportion of the ON-state falls. One reason may be that the high number of FOG-related falls during walking or standing indicates that patients become active during their ON-state, and thus more susceptible to experience FOG and falls [13,14]. In addition, in such occasions, patients are more prone to engage in dual or multiple task activities, increasing the chances of experi- encing FOG and leading to falls.

Not surprising, OFF-state falls were mostly FOG-related. Indeed, FOG occurs most frequently in the OFF-state [15,16] and the high incidence may be explained by the fact that patients still move (for instance due to urgency of going to the toilet) even though the medication effects had worn off.

Finally, only seven falls were related to dyskinesia in two pa- tients. Both patients were ON-fallers who reported balance loss during dual tasking such as dressing while standing, when violent dyskinesias were also present. It is known that both dyskinesias and dual tasking can aggravate balance impairment [17,18], hence this combination likely explained the falls in our patients.

This study has a few short-comings. First, unintentionally, 34 from our 36 patients presented FOG. On one hand, this high number of freezers may have influenced the amount of FOG-related falls. On the other hand, we believe that it represents the greater awareness towards FOG, increasing incidence of FOG with advancing disease (as we included patients who had already fallen), and represents the true picture of PD itself. Additionally, our

continuous education towards the recognition of FOG and dyski- nesia provided the patients with the right knowledge to prevent them from being missed. Second, due to outliers, our fall analysis was limited to the 10 first fall episodes. Although it may not provide the full picture of fall circumstances for a few patients, we believe that the influence of this decision on the results was limited. Third, even though the new-FOG questionnaire [22] is the most used in- strument to evaluate FOG, it may be unsuitable for long-term home based follow-ups. Therefore, we have decided to used the simpli- fied question: “Do you sometimes feel that your feet get glued to the floor while walking, making a turn, or when trying to initiate walking?”. Finally, although our selected sample limits the gener- alization of our conclusions, we are still confident that they are valid for a similar PD population.

In conclusion, this prospective study revealed that most of pa- tients might have a specific pattern of falls, either at the ON, OFF- or Transition-state. In addition, a substantial proportion of falls are due to FOG, in all motor states, with or without dual tasking. This information may be useful for preventing falls in patients with PD. Although challenging, amelioration of FOG at the OFF-, Transition- state, and even at the ON-state is partially possible when recom- mended treatment algorithms are followed [9,23]. Patients and their physicians need to determine in detail, which situations may lead to falls in each specific case and tailor-made strategies might reduce the risk of future falls.

Conflict of interest None.

Acknowledgment

This study was supported by the Block grant from Juntendo University School of Medicine. Ana Lígia Silva de Lima is supported by Coordenaça~o de Aperfeiçoamento de Pessoal de Nível Superior – CAPES [grant number 0428-140].

References

[1] B.R. Bloem, J.P. van Vugt, D.J. Beckley, Postural instability and falls in Parkin- son’s disease, Adv. Neurol. 87 (2001) 209e223.

[2] S.R. Cummings, M.C. Nevitt, S. Kidd, Forgetting falls. The limited accuracy of recall of falls in the elderly, J. Am. Geriatr. Soc. 36 (1998) 613e616.

[3] B.R. Bloem, Y.A.M. Grimbergen, M. Cramer, M. Willemsen, A.H. Zwinderman, Prospective assessment of falls in Parkinson’s disease, J. Neurol. 248 (2001) 950e958.

[4] A. Ashburn, E. Stack, C. Ballinger, L. Fazakarley, C. Fitton, The circumstance of falls among people with Parkinson’s disease and the use of falls diaries to facilitate reporting, Disabil. Rehabil. 30 (2008) 1205e1212.

[5] M.D. Latt, S.R. Lord, J.G. Morris, F.S. Fung, Clinical and physiological assess- ments for elucidating falls risk in Parkinson’s disease, Mov. Disord. 24 (2009) 1280e1289.

[6] G.K. Kerr, C.J. Worringham, M.H. Cole, J.M. Wood, P.A. Silburn, Predictors of future falls in Parkinson’s disease, Neurology 75 (2010) 116e124.

[7] B.R. Bloem, J.M. Hausdorff, J.E. Visser, N. Giladi, Falls and freezing of gait in Parkinson’s disease: a review of two interconnected, episodic phenomena, Mov. Disord. 19 (2004) 871e884.

[8] Y.Okuma,FreezingofgaitandfallsinParkinson’sdisease,J.Park.Dis.4(2014) 255e260.

[9] J. Nonnekes, A.H. Snijders, J.G. Nutt, G. Deuschl, N. Giladi, B.R. Bloem, Freezing of gait : a practical approach to management, Lancet. Neurol. 14 (2015) 768e778.

[10] C.G. Goetz, S. Leurgans, V.K. Hinson, L.M. Blasucci, J. Zimmerman, W. Fan, A. Hsu, Evaluating Parkinson’s disease patients at home: utility of self- videotaping for objective motor, dyskinesia, and ONeOFF assessments, Mov. Disord. 23 (10) (2008) 1479e1482.

[12] T. Gazibara, D.K. Tepavcevic, M. Svetel, A. Tomic, I. Stankovic, V.S. Kostic, T. Pekmezovic, Recurrent falls in Parkinson’s disease after one year of follow- up: a nested case-control study, Arch. Gerontol. Geriatr. 65 (2016) 17e24.

[13] D.M. Tan, J.L. McGinley, M.E. Danoudis, R. Iansek, M.E. Morris, Freezing of gait and activity limitations in people with Parkinson’s disease, Archives Phys. Med. rehabilitation 92 (7) (2011) 1159e1165.

[14] G. Cai, Y. Huang, S. Luo, Z. Lin, H. Dai, Q. Ye, Continuous quantitative

monitoring of physical activity in Parkinson’s disease patients by using

wearable devices: a case-control study, Neurol. Sci. (2017) 1e7.

. [15] J.D. Schaafsma, Y. Balash, T. Gurevich, A.L. Bartels, J.M. Hausdorff, N. Giladi, Characterization of freezing of gait subtypes and the response of each to
levodopa in Parkinson’s disease, Eur. J. Neurol. 10 (2003) 391e398.

. [16] Y. Okuma, N. Yanagisawa, The clinical spectrum of freezing of gait in Par-
kinson’s disease, Mov. Disord. 23 (2008) S426eS430.

. [17] S. Armand, T. Landis, R. Sztajzel, P.R. Burkhard, Dyskinesia-induced postural
instability in Parkinson’s disease, Park. Relat. Disord. 15 (2009) 359e364.

[18] R. Marchese, M. Bove, G. Abbruzzese, Effect of cognitive and motor tasks on postural stability in Parkinson’s disease: a posturographic study, Mov. Disord. 18 (2003) 652e658.

[22] A. Nieuwboer, L. Rochester, T. Herman, W. Vandenberghe, G.E. Emil, T. Thomaes, N. Giladi, Reliability of the new freezing of gait questionnaire: agreement between patients with Parkinson’s disease and their carers, Gait posture 30 (4) (2009) 459e463.

[23] Y. Okuma, Practical approach to freezing of gait in Parkinson’s disease, Pract. Neurol. 14 (2014) 222e230.

Y. Okuma et al. / Parkinsonism and Related Disorders 46 (2018) 30e35 35

Parkinsonism and Related Disorders 46 (2018) 36e40

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Transcranial sonographic findings may predict prognosis of gastroprokinetic drug-induced parkinsonism

Yoon-Sang Oh b, Do-Young Kwon a, *, Joong-Seok Kim b, Moon-Ho Park a, Daniela Berg c, d

a Department of Neurology, Korea University College of Medicine, Ansan-city, South Korea
b Department of Neurology, College of Medicine, The Catholic University of Korea, Seoul, South Korea
c Hertie Institute for Clinical Brain Research, Department of Neurodegeneration, University of Tuebingen, Tuebingen, Germany d Department of Neurology, Christian-Albrechts-University of Kiel, Kiel, Germany

articleinfo

Article history:

Received 21 April 2017 Received in revised form 2 October 2017
Accepted 18 October 2017

Keywords:

Transcranial sonography Echogenicity
Drug-induced parkinsonism Gastroprokinetic drug

abstract

Background: Drug-induced parkinsonism (DIP) is one important cause of parkinsonism and a major cause of misleading diagnosis of Parkinson’s disease (PD). DIP is caused by dopamine receptor blocking agents. Its symptoms will improve after withdrawal of offending drugs. However, parkinsonism does not regress in several individuals. It may persist or exacerbate despite drug withdrawal. Transcranial so- nography (TCS) of the substantia nigra (SN) has been widely used to diagnose PD and differentiate parkinsonism types. The objective of this study was to investigate the value of early TCS findings for predicting clinical outcome of patients with newly diagnosed gastroprokinetic drug-induced parkin- sonism after withdrawal of dopamine receptor blocking agents.

Methods: Fifty PD, 69 DIP, and 74 healthy controls were enrolled in this study. Patients with DIP were categorized into two subgroups: clinically improved after drug withdrawal (pure DIP) and clinically persistent or aggravated parkinsonism after drug withdrawal (unmasked PD). TCS was performed for all individuals to detect echogenicity in the SN.

Results: Transcranial sonographic SN echogenicity was significantly increased in PD while DIP and controls had similar SN echogenicity. In subgroup analysis of DIP, transcranial sonographic SN echoge- nicity was significantly increased in unmasked PD compared to that in pure DIP or healthy controls. Conclusions: SN echogenicity on TCS could be a useful tool to differentiate PD from DIP in clinical situ- ations. Pure DIP and unmasked PD exhibited different SN echogenicity patterns. Early SN echogenicity findings on TCS could be used a biomarker to predict clinical prognosis of DIP.

1. Introduction

Drug-induced parkinsonism (DIP) is one important cause of secondary parkinsonism. The prevalence of DIP among individuals with parkinsonism varies from 15% to 40% [1,2]. Although the pathophysiology of DIP is different from that of PD, clinical differ- entiation of DIP and idiopathic PD may be very difficult for some individuals. Early diagnosis of DIP is important as it allows clini- cians and patients to reduce unnecessary costs and stop the offending drug. In classic concept, DIP is reversible after drug withdrawal for a few days to months as the patient has no

* Corresponding author. Department of Neurology, Korea University College of Medicine, Korea University Ansan Hospital, 516, Gojan-1-dong, Ansan-city, Gyeonggi-do, 425-707, South Korea.

E-mail address: kwondoya@korea.ac.kr (D.-Y. Kwon).

https://doi.org/10.1016/j.parkreldis.2017.10.011

underlying deficits in the dopaminergic pathway. However, many types of DIP inducing drugs, especially neuroleptics, are hard to be withdrawn in practice because they are needed for underlying conditions of the patient. In addition, different studies have shown that drug can only disclose central dopaminergic degeneration in 30e75% of presumed DIP [3e6]. Thus, many DIP patients have persistent or increasing symptoms of parkinsonism following withdrawal of the offending drug [7,8]. Transcranial sonography (TCS) has now been widely accepted to assess various movement disorders, including PD. Advantages of TCS include its non-invasive nature, absence of exposure to radiation or nuclear material, short investigation time, ease of use, good compliance, and low cost compared to dopamine transporter imaging [9]. Presence of hyperechogenic area at the anatomical site of the substantia nigra (SN) in the midbrain has been shown in 68%e99% of PD patients [8,9]. The majority (85e90%) of healthy subjects have no

hyperechogenicity of the SN [9]. TCS has a diagnostic value for early diagnosis of PD, differentiation between atypical parkinsonian syndromes and secondary parkinsonian syndromes, and detection of subjects at risk for PD [10,11]. TCS is also regarded as a useful tool for distinguishing PD from essential tremor, restless legs syndrome, and DIP [7,8,12,13]. The objective of this study was to investigate the value of early TCS findings for predicting the clinical outcome of patients with newly diagnosed gastroprokinetic drug-induced parkinsonism after withdrawal of dopamine receptor blocking agents.

2. Subjects and methods

2.1. Subjects

A total of 196 subjects who visited the movement disorder clinic of a university-affiliated hospital over a period of three years were enrolled in this study. PD was diagnosed based on UK Parkinson’s Disease Society Brain Bank clinical diagnostic criteria. A total of 74 age-matched healthy controls without any notable neurological or psychiatric history were also recruited from an outpatients clinic. Clinical information including age, sex, disease duration, and complete history of medication was obtained from all subjects. Data from complete physical and neurological examinations including Unified Parkinson’s Disease Rating Scale (UPDRS) Part III and modified Hoehn and Yahr (H&Y) stage score were also obtained. Magnetic resonance imaging of the brain was performed for all patients with PD and DIP. General cognitive function was evaluated with the Korean version of Mini Mental Status Examination (K- MMSE) and the Korean version of Montreal Cognitive Assessment (MoCA-K) for all subjects except those in the control group. Amongst patients with PD, those with previous history of stroke, other neurological or psychiatric disorders, a history of medication that might cause parkinsonism, or secondary causes of dementia were excluded. None of these patients with newly diagnosed PD had a previous history of anti-dementia medication use. Informed consent was obtained from all participants. The local ethics com- mittee approved this study. DIP was defined when no parkinsonism was reported before the use of offending prokinetic drugs and signs of parkinsonism appeared following the intake of dopamine re- ceptor blocking agents (gastroprokinetic drug) [14]. Follow-up UPDRS part III motor score was obtained at six months after drug withdrawal. Subjects with almost complete improvement of UPDRS part III motor score without needing symptomatic dopaminergic treatment were categorized as pure DIP. Those with remaining

motor deficits in follow-up UPDRS part III motor score needing anti-parkinsonian medication were categorized as unmasked PD. In our study, only individuals with gastrointestinal prokinetics- induced parkinsonism were included in the DIP group. Cases of DIP caused by antipsychotics, calcium channel antagonists, antiar- rhythmics, or antidepressants were excluded. [18F] FP-CIT scan was performed in 37 (53.6%) of DIP patients.

2.2. TCS

TCS was performed for all participants as part of initial assess- ments before withdrawing offending drugs and starting symp- tomatic treatments. It was performed with a 2.5 MHz transducer (LOGIQ 3pro, GE Healthcare, USA) at depth of 16 cm and dynamic range of 45 dB as described previously [13,15e17]. Individuals with sufficient temporal bone windows were included. An experienced sonographer who was blinded to clinical history of patients measured SN echogenicity through the preauricular temporal acoustic bone window in the mesencephalic axial plane. The examiner modulated the brightness of image as needed to obtain the best quality. Within the butterfly-shaped hypoechogenic midbrain, a constant echogenic signal at the anatomical location of SN ipsilateral to the probe was identified from both sides and stored (Fig. 1). The acquired image was magnified threefold. Another examiner who was also blinded to subject’s clinical information manually outlined the outer circumference of the midbrain and the SN echogenic area as described previously [18,19]. SN echogenic area was classified according to previously determined cut-off values obtained at our lab. Hyperechogenicity was defined as SN echogenic area !0.2 cm2 while normoechogenicity was defined as < 0.2 cm2 [12,13,18,19]. The ratio of the sum of bilateral SN echo- genic areas to total midbrain area was also obtained to compare the relative area of echogenicity in each individual.

2.3. Statistical analysis

Statistical analyses were performed using SPSS version 15.0 (SPSS Inc., Chicago, IL, USA). One-way analysis of variance (ANOVA) with Bonferroni post-hoc testing was used to compare means among groups. Pearson’s c2 test was used to compare frequencies of categorical variables. As SN echogenic areas were not distributed normally, we used natural log transformed values. For between- group comparisons of SN echogenic area and assessment of the ratio of the sum of bilateral SN echogenic areas to the area of the midbrain, analysis of covariance (ANCOVA) with Bonferroni post-

Y.-S. Oh et al. / Parkinsonism and Related Disorders 46 (2018) 36e40 37

Fig. 1. Transcranial sonographic (TCS) images of the midbrain. (A) Image of normal SN echogenicity in healthy control. (B) Image of SN hyperechogenicity in Parkinson’s disease (PD) patient. The butterfly-shaped midbrain is encircled with a line manually. After identifying the cerebral aqueduct (marked with a star), echogenic signals at the area anatomically consistent with the substantia nigra (SN) (dotted line) are then obtained.

38 Y.-S. Oh et al. / Parkinsonism and Related Disorders 46 (2018) 36e40

Table 1

Demographics, clinical characteristics, general cognitive functions and transcranial sonographic findings of each group of subjects.

Controls (N 1⁄4 74)

72.0 ± 8.8 40 (54.1)

e e e e e

0.14 ± 0.05 0.87 ± 0.18 0.03 ± 0.01 7 (9.5)

67 (90.5)

DIP(n1⁄469)

70.5 ± 9.0 23 (33.3)

1.1 ± 1.1 27.1 ± 13.0 2.1 ± 0.6 22.9 ± 5.2 16.2 ± 6.7

0.18 ± 0.06 0.77 ± 0.15 0.04 ± 0.01 24 (34.8)

45 (65.2)

Subgroups Pure DIP

(n1⁄447)

69.3 ± 9.9 12 (25.5)

0.9 ± 1.2 25.3 ± 11.8 2.0 ± 0.5 22.3 ± 5.4 15.6 ± 7.0

0.15 ± 0.04 0.83 ± 0.13 0.04 ± 0.01 6 (12.8)

41 (87.2)

PD(n1⁄450) P

Post hoc test

Subjects
Age, years, mean ± SD

Sex, male (%)**
Disease duration, years UPDRS part III score Hoehn and Yahr stage K-MMSE
MOCA-K

TCS
SN echogenicity (cm2)*

Crude value

Natural-log-transformed value

Sum of SN/midbrain ratio*

SN hyperechogenicity,n (%)** Normal echogenicity, n (%)

Unmasked PD (n1⁄422)

73.0 ± 6.1 11 (50)

1.3 ± 0.9 31.0 ± 14.6 2.3 ± 0.7 24.2 ± 4.7 17.5 ± 5.8

0.24 ± 0.04 0.62 ± 0.08 0.05 ± 0.01 18 (81.8)

4 (18.2)

64.1 ± 10.5 24 (48)

4.5 ± 3.1 20.2 ± 10.6 2.1 ± 0.8 24.2 ± 5.1 18.6 ± 6.9

0.24 ± 0.09 0.64 ± 0.18 0.06 ± 0.02 35 (70)

15 (30)

<0.001 PD < controls 1⁄4 pure DIP 1⁄4 unmasked PD

0.018
<0.001 PD > pure DIP 1⁄4 unmasked PD <0.001 PD < pure DIP 1⁄4 unmasked PD 0.145
0.172
0.096

<0.001 controls 1⁄4 pure DIP < PD 1⁄4 unmasked PD

<0.001

UPDRS, unified Parkinson’s disease rating scale; DIP, drug-induced parkinsonism; MMSE, mini-mental status examination.
MoCA, Montreal cognitive assessment; TCS, transcranial sonography; SN, substantia nigra.
Values represent mean with standard deviation or numbers of patients (percentage).
Analyses were performed by one-way analysis of variance (ANOVA) with Bonferroni post-hoc tests, *one way analysis of covariance (ANCOVA) with Bonferroni post-hoc tests, and **the c2 test.

hoc testing was conducted after controlling for age, UPDRS part III motor score, and disease duration. Statistical significance was defined at p-value < 0.05.

3. Results

A total of 50 PD, 69 DIP patients, and 74 age-matched healthy controls were enrolled in this study. No patients with PD and DIP had evidence of secondary parkinsonism on MRI scan. The mean age of the PD group was lower (p < 0.001) than that of the DIP group or the control group. Female gender was higher (p 1⁄4 0.04) in the DIP group compared to that in the PD or control group. Patients with PD had longer (p < 0.001) disease duration but less pro- nounced motor deficits (UPDRS part III score, p < 0.001; Hoehn and Yahr stage, p < 0.001) than DIP patients. General cognitive function as assessed by K-MMSE or MOCA-K was not significantly (p > 0.05) different between PD and DIP groups (Table 1). Regarding causative prokinetic drugs of DIP, levosulpiride was the most common offending prokinetic drug (n 1⁄4 54, 78.2%), followed by metoclo- pramide (n 1⁄4 8), clebopride (n 1⁄4 5), and itopride (n 1⁄4 2). Trans- cranial sonographic SN echogenicity was significantly (p 1⁄4 0.006) increased in PD patients while DIP patients and controls had similar

SN echogenicity (Fig. 2A). The ratio of the sum of both sides of SN echogenic area to the whole midbrain area was also significantly (p < 0.001) increased in the PD group compared to that in the other two groups. A total of 39 (78%) PD patients showed SN hyper- echogenicity whereas 24 (34.8%) DIP patients and 7 (9.5%) control subjects had SN hyperechogenicity. Subgroup analysis of DIP revealed that the mean age in pure DIP or unmasked PD subgroup was higher than that in the PD group. Female gender was pre- dominant only in the pure DIP subgroup. Disease durations were similar between pure DIP and unmasked PD subgroups. Both were shorter than that in the PD group. UPDRS part III scores were not significantly different between pure DIP and unmasked PD groups. They were higher than the score in the PD group. Modified Hoehn and Yahr stage scores were not different among PD, pure DIP, and unmasked PD groups. General cognitive functions of the three groups were similar to each other (Table 1). Transcranial sono- graphic echogenicity was significantly increased in PD and unmasked PD groups compared to that in pure DIP and control groups (Fig. 2B). The ratio of the sum of both sides of the SN echogenic area to the area of the whole midbrain was also increased in PD or unmasked PD patients compared to that in pure DIP pa- tients or controls. SN hyperechogenicity was predominant in

Fig. 2. Transcranial sonographic substantia nigra (NS) echogenicity in healthy controls, Parkinson’s disease (PD), and drug induced parkinsonism (DIP). A. PD showing increased echogenicity than other groups (**, p 1⁄4 0.006). B. PD and unmasked PD showing increased echogenicity than controls and pure DIP (**, p < 0.001). Data represent mean ± SEM (standard error of the mean).

unmasked PD and PD than pure DIP and controls. SN hyper- echogenicity was found in 6 (12.8%) patients with pure DIP. How- ever, 18 (81.8%) patients with unmasked PD showed SN hyperechogenicity. In the DIP group, sensitivity and specificity of SN hyperechogenicity to detect unmasked PD were 75.0% (18/24) and 91.1%, respectively.

4. Discussion

Transcranial sonographic findings of hyperechogenicity of the SN in patients with PD of this study are in agreement with results of previous studies [13,20e22]. SN normoechogenicity in patients with DIP was also confirmed in this study, in accordance with previous results [7,8]. This indicates that difference in SN echoge- nicity can be used as a marker to distinguish PD from DIP. However, findings concerning SN echogenicity in DIP vary. One study could no detect significant differences in transcranial sonographic results between DIP and PD [23]. The authors of that study still believe that transcranial sonography is a valid technique for the diagnosis of drug-induced parkinsonism [23]. They believe that their results might not have reached significance due to a small sample size [23]. Importantly, one small prospective study has found a correlation between the severity of parkinsonian symptoms and the echoge- nicity of the substantia nigra in DIP, indicating that neuroleptic drugs may unmask the predisposition to PD and result in more severe symptoms [24]. Based on this observation, DIP groups were sub-classified to clarify findings of SN echogenicity. In DIP subgroup analysis, SN echogenicity was significantly different between cases with pure DIP and unmasked PD. Pure DIP presented improvement in parkinsonism. After withdrawal of the offending prokinetic drugs, they did not have impaired activities of daily living. However, unmasked PD was associated with limited improvement in parkinsonism or completely persistent parkinsonism at 6 months after drug withdrawal. SN hyperechogenicity in unmasked PD was similar to that in PD.

TCS hyperechogenicity of the SN is regarded as a risk marker for PD [25]. A large study has shown that SN hyperechogenicity may be detected in about 10% of normal healthy elderly subjects, which may increase the risk of developing clinical features of PD within 5 years by more than 20-fold [20,25]. Therefore, subjects with unmasked PD in this study who had high prevalence of SN hyper- echogenicity might have predisposition to PD due to ongoing neurodegeneration of the dopaminergic system which was unmasked by prokinetic and antidopaminergic drugs. These in- dividuals might have already been in the prodromal phase of PD. Moreover, SN hyperechogenicity besides hyposmia possibly be associated with an increased risk for mild parkinsonian signs in the general elderly community [26]. The strength of this study was that only DIP subjects whose conditions were elicited by gastrointes- tinal prokinetic drugs were enrolled. Gastroprokinetic drugs are potent inducers of DIP, especially in Korea. It is easy to quit these kinds of drugs to check clinical outcomes [27,28]. In clinical prac- tice, it is more difficult for patients to stop taking offending drugs such as neuroleptics, antidepressants, and calcium channel blockers. Discontinuation of antipsychotics is associated with withdrawal symptoms, rebound symptoms, and supersensitivity syndrome. Discontinuation of antidepressant causes various with- drawal symptoms and significant distress to many subjects [29]. Calcium channel blockers are used to treat arrhythmia and hyper- tension. They might be necessary and continuously needed. Moreover, the occurrence of parkinsonism is delayed (after 12 months of treatment with calcium channel blockers) while most other relevant drugs cause parkinsonism within 3 months [30]. Therefore, calcium channel blocker-induced parkinsonism in addition to neuroleptics induced parkinsonism was excluded from

DIP in our study. In addition, age- and sex-matched healthy controls for the DIP group were included to control for the effect of SN hyperechogenicity on normal aged subjects.

This study also has several limitations. Dopamine transporter imaging was not evaluated in all patients of DIP to confirm nigrostriatal dopaminergic degeneration. [18F] FP-CIT scan was performed in only 53.6% (n 1⁄4 37) of DIP patients, which was not sufficient to evaluate the correlation between SN hyper- echogenicity and DAT deficit. However, abnormal DAT scan showed high concordance with unmasked PD (8 unmasked PD out of 9 abnormal scans). Secondly, patients were evaluated at 6 months after drug withdrawal. Parkinsonism might have improved after 6 months. Therefore, patients might have been re-categorized from unmasked PD to pure DIP. A recent study has recommended that the minimal time of neuroleptic withdrawal for a diagnosis of PD should be one year [1]. However, neuroleptic-induced parkin- sonism was not included in this study. However, since most pre- vious studies have evaluated parkinsonian symptoms within 6 months [1], the 6-month time interval in our study might be appropriate. Lastly, the sample size of unmasked the PD group was smaller than that of other groups, making statistical comparison more difficult. Insonation failure owing to relatively advanced age of subjects and the tendency of East Asian people to have thicker temporal bones [31] might have caused such result.

In conclusion, SN echogenicity was generally normal in DIP, similar to that in normal healthy controls. However, in subgroup analysis of DIP, SN echogenicity was significantly different between pure DIP and unmasked PD groups. Pure DIP was associated with nornoechogenicity, similar to that in the control group. Unmasked PD was associated with hyperechogenicity. Therefore, early TCS finding of SN echogenicity might be a useful marker to predict clinical prognosis of DIP.

Authors role

. 1 Research project
A Conception: DY Kwon, MH Park
B Organization: DY Kwon
C Execution: MH Park, YS Oh, JS Kim

. 2 Statistical Analysis
A Design: DY Kwon, MH Park
B Execution: MH Park, YS Oh, JS Kim
C Review and Critique: MH Park, D Berg

. 3 Manuscript Preparation
A Writing of the first draft: DY Kwon, YS Oh B Review and Critique: JS Kim, D Berg

Funding

None.

Conflict of interest None.

Acknowledgement

This study was done in accord with the ethical standards of the Committee on Human Experimentation of the institution (Korea University Ethical Review Board) in which the experiments were done or in accord with the Helsinki Declaration of 1975.

Y.-S. Oh et al. / Parkinsonism and Related Disorders 46 (2018) 36e40 39

40 Y.-S. Oh et al. / Parkinsonism and Related Disorders 46 (2018) 36e40

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.011.

References

. [1] T.T. Lim, A. Ahmed, I. Itin, M. Gostkowski, J. Rudolph, S. Cooper, et al., Is 6 months of neuroleptic withdrawal sufficient to distinguish drug-induced parkinsonism from Parkinson’s disease? Int. J. Neurosci. 123 (2013) 170e174.

. [2] R. Savica, B.R. Grossardt, J.H. Bower, J.E. Ahlskog, M.M. Mielke, W.A. Rocca, Incidence and time trends of drug-induced parkinsonism: a 30-year population-based study, Mov. Disord. 32 (2017) 227e234.

. [3] J.F. Morley, S.M. Pawlowski, A. Kesari, I. Maina, A. Pantelyat, J.E. Duda, Motor and non-motor features of Parkinson’s disease that predict persistent drug- induced Parkinsonism, Park. Relat. Disord. 20 (2014) 738e742.

. [4] S.H. Lee, H.K. Kim, Y.G. Lee, C.H. Lyoo, S.J. Ahn, M.S. Lee, Clinical features indicating nigrostriatal dopaminergic degeneration in drug-induced parkin- sonism, J. Mov. Disord. 10 (2017) 35e39.

. [5] M. Tinazzi, S. Ottaviani, I.U. Isaias, I. Pasquin, M. Steinmayr, C. Vampini, et al., [123I]FP-CIT SPET imaging in drug-induced Parkinsonism, Mov. Disord. 23 (2008) 1825e1829.

. [6] M. Tinazzi, F. Morgante, A. Matinella, T. Bovi, A. Cannas, P. Solla, et al., Imaging of the dopamine transporter predicts pattern of disease progression and response to levodopa in patients with schizophrenia and parkinsonism: a 2- year follow-up multicenter study, Schizophr. Res. 152 (2014) 344e349.

. [7] P. Mahlknecht, H. Stockner, S. Kiechl, J. Willeit, V. Rastner, A. Gasperi, et al., Is transcranial sonography useful to distinguish drug-induced parkinsonism from Parkinson’s disease? Mov. Disord. 27 (2012) 1194e1196.

. [8] W. Poewe, K. Seppi, Diagnosis of drug-induced parkinsonism: can transcranial sonography make the difference? Eur. J. Neurol. 20 (2013) 1429e1430.

. [9] D. Berg, J.D. Steinberger, C. Warren Olanow, T.P. Naidich, T.A. Yousry, Mile- stones in magnetic resonance imaging and transcranial sonography of movement disorders, Mov. Disord. 26 (2011) 979e992.

. [10] A. Berardelli, G.K. Wenning, A. Antonini, D. Berg, B.R. Bloem, V. Bonifati, et al., EFNS/MDS-ES/ENS recommendations for the diagnosis of Parkinson’s disease, Eur. J. Neurol. 20 (2013) 16e34.

. [11] E. Sanzaro, F. Iemolo, Transcranial sonography in movement disorders: an interesting tool for diagnostic perspectives, Neurol. Sci. 37 (2016) 373e376.

. [12] J. Jesus-Ribeiro, A. Freire, J. Sargento-Freitas, M. Sousa, F. Silva, F. Moreira, et al., Transcranial sonography and DaTSCAN in early stage Parkinson’s disease and essential tremor, Eur. Neurol. 76 (2016) 252e255.

. [13] D.Y. Kwon, W.K. Seo, H.K. Yoon, M.H. Park, S.B. Koh, K.W. Park, Transcranial brain sonography in Parkinson’s disease with restless legs syndrome, Mov. Disord. 25 (2010) 1373e1378.

. [14] F.J. Jimenez-Jimenez, M. Orti-Pareja, L. Ayuso-Peralta, T. Gasalla, F. Cabrera- Valdivia, A. Vaquero, et al., Drug-induced parkinsonism in a movement dis- orders unit: a four-year survey, Park. Relat. Disord. 2 (1996) 145e149.

[15] G. Becker, D. Berg, Neuroimaging in basal ganglia disorders: perspectives for transcranial ultrasound, Mov. Disord. 16 (2001) 23e32.

[16] D. Berg, G. Becker, B. Zeiler, O. Tucha, E. Hofmann, M. Preier, et al., Vulnera- bility of the nigrostriatal system as detected by transcranial ultrasound, Neurology 53 (1999) 1026e1031.

[17] U. Walter, S. Behnke, J. Eyding, L. Niehaus, T. Postert, G. Seidel, et al., Trans- cranial brain parenchyma sonography in movement disorders: state of the art, Ultrasound Med. Biol. 33 (2007) 15e25.

[18] D. Berg, C. Siefker, G. Becker, Echogenicity of the substantia nigra in Parkin- son’s disease and its relation to clinical findings, J. Neurol. 248 (2001) 684e689.

[19] U. Walter, M. Wittstock, R. Benecke, D. Dressler, Substantia nigra echogenicity is normal in non-extrapyramidal cerebral disorders but increased in Parkin- son’s disease, J. Neural Transm. (Vienna) 109 (2002) 191e196.

[20] S. Behnke, G. Becker, Sonographic imaging of the brain parenchyma, Eur. J. Ultrasound 16 (2002) 73e80.

[21] D. Berg, In vivo detection of iron and neuromelanin by transcranial sonographyea new approach for early detection of substantia nigra damage, J. Neural Transm. (Vienna) 113 (2006) 775e780.

[22] D. Berg, Substantia nigra hyperechogenicity is a risk marker of Parkinson’s disease: yes, J. Neural Transm. (Vienna) 118 (2011) 613e619.

[23] J. Olivares Romero, A. Arjona Padillo, F.J. Barrero Hernandez, M. Martin Gon- zalez, B. Gil Extremera, Utility of transcranial sonography in the diagnosis of drug-induced parkinsonism: a prospective study, Eur. J. Neurol. 20 (2013) 1451e1458.

[24] D. Berg, B. Jabs, U. Merschdorf, H. Beckmann, G. Becker, Echogenicity of substantia nigra determined by transcranial ultrasound correlates with severity of parkinsonian symptoms induced by neuroleptic therapy, Biol. Psychiatry 50 (2001) 463e467.

[25] D. Berg, S. Behnke, K. Seppi, J. Godau, S. Lerche, P. Mahlknecht, et al., Enlarged hyperechogenic substantia nigra as a risk marker for Parkinson’s disease, Mov. Disord. 28 (2013) 216e219.

[26] P. Mahlknecht, S. Kiechl, H. Stockner, J. Willeit, A. Gasperi, W. Poewe, et al., Predictors for mild parkinsonian signs: a prospective population-based study, Park. Relat. Disord. 21 (2015) 321e324.

[27] H.W. Shin, M.J. Kim, J.S. Kim, M.C. Lee, S.J. Chung, Levosulpiride-induced movement disorders, Mov. Disord. 24 (2009) 2249e2253.

[28] T. Mathew, U.S. Nadimpally, A.D. Prabhu, R. Nadig, Drug-induced Parkin- sonism on the rise: beware of levosulpiride and its combinations with proton pump inhibitors, Neurol. India 65 (2017) 173e174.

[29] G.A. Fava, A. Gatti, C. Belaise, J. Guidi, E. Offidani, Withdrawal symptoms after selective serotonin reuptake inhibitor discontinuation: a systematic review, Psychother. Psychosom. 84 (2015) 72e81.

[30] E. Bondon-Guitton, S. Perez-Lloret, H. Bagheri, C. Brefel, O. Rascol, J.L. Montastruc, Drug-induced parkinsonism: a review of 17 years’ experience in a regional pharmacovigilance center in France, Mov. Disord. 26 (2011) 2226e2231.

[31] J.Y. Kim, S.T. Kim, S.H. Jeon, W.Y. Lee, Midbrain transcranial sonography in Korean patients with Parkinson’s disease, Mov. Disord. 22 (2007) 1922e1926.

Parkinsonism and Related Disorders 46 (2018) 41e46

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Square biphasic pulse deep brain stimulation for essential tremor: The BiP tremor study*

Sol De Jesus a, b, Leonardo Almeida a, *, Leili Shahgholi a, Daniel Martinez-Ramirez a, Jaimie Roper c, Chris J. Hass c, Umer Akbar d, Aparna Wagle Shukla a, Robert S. Raike e, Michael S. Okun a

a University of Florida, Department of Neurology, Center for Movement Disorders and Neurorestoration, Gainesville, FL, USA b Department of Neurology, Pennsylvania State University-Milton S. Hershey Medical Center, Hershey, PA, USA
c University of Florida, Department of Applied Physiology and Kinesiology, Gainesville, FL, USA
d Department of Neurology, Brown University, Providence, RI, USA

e Research and Core Technology, Medtronic Restorative Therapies Group Implantables, USA

articleinfo abstract

Article history:

Received 27 August 2017 Received in revised form 2 October 2017
Accepted 19 October 2017

Keywords:

Deep brain stimulation Essential tremor Neurostimulation
Biphasic pulse stimulation Movement disorders

Background: Conventional deep brain stimulation (DBS) utilizes regular, high frequency pulses to treat medication-refractory symptoms in essential tremor (ET). Modifications of DBS pulse shape to achieve improved effectiveness is a promising approach.
Objectives: The current study assessed the safety, tolerability and effectiveness of square biphasic pulse shaping as an alternative to conventional ET DBS.

Methods: This pilot study compared biphasic pulses (BiP) versus conventional DBS pulses (ClinDBS). Eleven ET subjects with clinically optimized ventralis intermedius nucleus DBS were enrolled. Objective measures were obtained over 3 h while ON BiP stimulation.
Results: There was observed benefit in the Fahn-Tolosa Tremor Rating Scale (TRS) for BiP conditions when compared to the DBS off condition and to ClinDBS setting. Total TRS scores during the DBS OFF condition (28.5 IQR 1⁄4 24.5e35.25) were significantly higher than the other time points. Following active DBS, TRS improved to (20 IQR 1⁄4 13.8e24.3) at ClinDBS setting and to (16.5 IQR 1⁄4 12e20.75) at the 3 h period ON BiP stimulation (p 1⁄4 0.001). Accelerometer recordings revealed improvement in tremor at rest (c2 1⁄4 16.1, p 1⁄4 0.006), posture (c2 1⁄4 15.9, p 1⁄4 0.007) and with action (c2 1⁄4 32.1, p1⁄4<0.001) when comparing median total scores at ClinDBS and OFF DBS conditions to 3 h ON BiP stimulation. There were no adverse effects and gait was not impacted.

Conclusion: BiP was safe, tolerable and effective on the tremor symptoms when tested up to 3 h. This study demonstrated the feasibility of applying a novel DBS waveform in the clinic setting. Larger pro- spective studies with longer clinical follow-up will be required.

1. Introduction

Abbreviations: DBS, deep brain stimulation; ET, essential tremor; BiP, biphasic pulse shape; ClinDBS, conventional DBS pulse; TRS, Fahn-Tolosa Tremor Rating Scale; IQR, interquartile range; VIM, thalamic ventralis intermedius nucleus.

* This manuscript was run through the iThenticate system provided by the University of Florida and the 1st author takes all responsibility for ensuring originality.

* Corresponding author. Department of Neurology, Center for Movement Disorder and Neurorestoration, University of Florida, P.O Box 100236, Gainesville, FL 32610, USA.

E-mail address: l.ameida@ufl.edu (L. Almeida).

https://doi.org/10.1016/j.parkreldis.2017.10.015

Essential tremor (ET) is one of the most common movement disorders encountered in clinical practice with an approximate estimate of seven million cases in the United States [1]. Deep brain stimulation (DBS) targeting the thalamic ventralis intermedius nucleus (VIM) has been demonstrated to be a safe and effective option for a subset of medication-refractory ET patients [2]. Commercially available and FDA approved stimulation devices utilize, high frequency (!100 Hz), regular, rectangular pulse stim- ulation to treat all neuropsychiatric indications for DBS [3]. DBS therapy in clinical practice applies chronic continuous stimulation

42 S. De Jesus et al. / Parkinsonism and Related Disorders 46 (2018) 41e46

(i.e. set it and forget it) [4e6]. DBS programming traditionally in- volves the iterative alteration of lead configuration (contact selec- tion and polarity), amplitude (voltage or current), pulse width, and frequency.

Recently, an improved understanding of underlying brain cir- cuitry and network dysfunction in movement disorders has led to the investigation of new stimulation patterns and pulse shapes that could be employed to more effectively and/or efficiently modulate symptoms [7e9]. Several authors have suggested that newer patterns of stimulation may affect symptom response and neurostimulator battery consumption. These newer patterns may trigger differential neuronal excitation by altering the regularity of pulses [8,10e13] or applying non-rectangular pulse shapes [7,9,14,15]. Non continuous irregular and regular patterns of stim- ulation have been investigated in computer, animal models and in a small number of acutely-implanted human DBS patients [8,10]. Non-rectangular pulse shapes such as sinusoidal or Gaussian waveforms have only been tested in computational models of mammalian axons [9,16,17]. Testing of these unconventional stimulation concepts only recently has been explored in chroni- cally implanted DBS patients [7].

Previously, our group conducted a small safety and feasibility study to test various non-conventional DBS pulse patterns and shapes on tremor in ET and PD patients. Several stimulation set- tings were identified as well-tolerated and effective. Utilizing a charge-balanced biphasic pulse with a square-wave active recharge (BiP), our study revealed promising data on tremor control, how- ever only short stimulation periods of 2 min were employed [7]. The primary aim of the current study was to evaluate the safety and tolerability of longer duration BiP DBS, and to test its effectiveness in a larger cohort of ET subjects.

2. Materials and methods

The protocol was approved by the University of Florida Insti- tutional Review Board and informed consent was obtained from all participants. The study was registered at http://www.ClinicalTrials. gov (identifier: NCT02569021).

2.1. Study design and participants

The study was designed to evaluate biphasic pulse shape (BiP) DBS as compared with conventional pulse DBS (ClinDBS). Both deliver a charge-balanced pulse, however conventional DBS ex- hibits an asymmetric passive recharge whereas BiP exhibits a symmetric active recharge [7].

Eligible subjects were selected during routine DBS program- ming sessions conducted at the University of Florida Health Center for Movement Disorders and Neurorestoration. Inclusion criteria were: 1) a diagnosis of ET by a fellowship trained movement dis- order neurologist using strict published criteria [18], 2) The implanted DBS device is compatible with the research software (Medtronic Activa SC, PC or RC) 3) Subjects were required to have undergone ventralis intermediate nucleus (VIM) stimulation, and 4) Subjects have undergone a minimum of four clinical program- ming sessions over a minimum of four months for optimization of the DBS settings. The exclusion criteria were: 1) Other neurode- generative diagnoses 2) A previously revised DBS device (e.g. infection or suboptimal placement), 3) and less than four DBS programming sessions spread over four months.

Eleven ET DBS subjects were enrolled in the study. All subjects underwent a careful assessment for verification of lead location and optimal device placement prior to testing. Each subject presented for a single office visit in the 12 h off tremor medication state. Objective tremor scores were obtained under the following

conditions: condition 1 (ClinDBS settings); condition 2 (post- 30 min washout/ClinDBS OFF); condition 3 (30 min of ON BiP DBS); condition 4 (1 h of ON BiP DBS); condition 5 (2 h of ON BiP DBS); condition 6 (3 h of ON BiP DBS). These scores were compared with the baseline scores obtained with ClinDBS pulse shape and DBS OFF state.

The study design utilized each patient as their own control, and compared BiP DBS versus ClinDBS, Fig. 1. BiP stimulation was tested on the most affected side, for patients with bilateral DBS stimula- tion the second lead remained on chronic conventional DBS stim- ulation. All assessments were videotaped and videotapes were reviewed by two independent blinded raters.

Throughout the study, subjects remained in the clinic under constant supervision of a movement disorder neurologist to ensure safety and tolerability of the novel DBS setting. Each subject was instructed to report adverse effects or symptom changes (perceived worsening or improvement) during the study period. Each reported adverse effect or symptom change was carefully assessed and a determination was made as to whether it was DBS related or not.

2.2. Implementation of BiP DBS

A charge-balanced biphasic pulse with a square-wave active recharge was used for the study. The parameters conformed to the charge density safety limit of 30 mC/cm2/phase [19]. DBS pulses were delivered utilizing research firmware (FW) developed by Medtronic, Neuromodulation, Minneapolis, MN. The custom firm- ware was available on a Windows-based PC laptop connected to a telemetry head, and it was operated by a trained neurologist who could download and adjust the programing in the clinic setting. The frequency parameters in the DBS settings required slight adjust- ments due to limitations of the software, BiP DBS frequencies comparable to the chronic settings were applied for the duration of the study if an exact match was not available (i.e. subject 1 chronic frequency setting of 135Hz was converted to a frequency of 130Hz ON BiP stimulation). BiP stimulation could be discontinued immediately if there was a reported intolerable side effect. Subjects remained ON BiP stimulation for a maximum of 3 h period. Subjects remained aware of the DBS setting changes. At the completion of the study period all participants were returned to their ClinDBS settings.

2.3. Motor assessments/outcome measures

The effects of BiP DBS were assessed through administration of the following outcome measures: the Fahn-Tolosa-Marin Tremor Rating Scale (TRS), accelerometer recordings (Kinesia Great Lakes NeuroTechnology, Cleveland, OH) and mobility using the GaitRite

Fig. 1. Study protocol. Patients were tested at baseline, after a 30-min washout, and different time points while on BiP settings.

walkway and software suite (GaitRite CIR systems INC. Haver- town, PA). The change in total motor TRS score was evaluated. In addition, accelerometer hand recordings contralateral to the side of DBS ON BiP testing were obtained. Tremor was analyzed with 10 s accelerometry recordings during rest, posture (arms held out in front of the body) and action (finger to nose maneuver). Quantitative tremor analyses was obtained using a previously validated tremor severity scale (0-4) for the Kinesia system [20,21]. The effects of DBS stimulation on balance and gait per- formance in ET patients remain unclear and controversial [22e24]. Gait was quantitatively evaluated with two self-selected speed walks performed across the GaitRite walkway system to evaluate potential balance/gait changes with BiP stimulation in this cohort [25]. The following gait variables were recorded: ve- locity, cadence, step time, step length, stride length, stance time, step width, swing time, single/double limb support and step length asymmetry.

2.4. Data analysis

SPSS version 22.0 statistical package was utilized for the analysis and a defined level of 0.05 employed for statistical significance. Categorical variables were displayed as counts and proportions. Continuous variables were displayed as medians and interquartile ranges. Repeated measures were analyzed via the non-parametric Friedman’s test and pairwise Wilcoxon test with Bonferroni cor- rections used for post-hoc analysis when appropriate. The inter

Table 1

Participant demographics, clinical information and subjective report/stimulation experience.

4 VIM
-Left 160Freq

5 VIM
-Left 135Freq

6 VIM
-Left 200Freq

7 VIM
-Right 150Freq

8 VIM
-Right 180Freq

9 VIM
-Left 180Freq

10 VIM
-Left 135Freq

11 VIM
-Left 135Freq

1-2þ2.8V60PW 2- Cþ 2.1V 120PW

2- 3 þ 3.94V 120PW 1- 2 þ 3.9V 150PW

2- Cþ 2.3V 90PW

1-2- 0 þ 3.4V 120PW

2-1þ3.3V90PW 2- Cþ 2.3V 80PW

S. De Jesus et al. / Parkinsonism and Related Disorders 46 (2018) 41e46 43

rater reliability for blinded tremor assessment was established based on a moderate intraclass correlation coefficient (ICC) of 0.70 or greater. A repeated measures analysis of variance (ANOVA) was conducted to compare the effect of the six stimulation conditions (time) on the gait variables.

3. Results

3.1. Patient characteristics

Eleven ET subjects (6 males and 5 females) consented and completed the study. The disease duration for subjects ranged from 10 years to 60 years with the total lifetime DBS therapy time ranging from 7 months to 11 years. Six patients had bilateral VIM DBS and five patients had unilateral VIM DBS. BiP testing was applied to the most affected motoric side (contralateral brain hemisphere) regardless of handedness based on patient self-report. In cases with bilateral VIM DBS, the chronic clinical DBS settings remained unchanged for the untested lead. All participants had Medtronic Activa SC neurostimulators with the exception of one patient who had a Medtronic Activa PC neurostimulator, and had implanted Medtronic 3387 leads, which were verified to be adequately placed upon inclusion in the study. None of the patients enrolled suffered from stimulation-induced side effects of gait disturbance. Detailed information on the participants’ diagnosis, handedness, disease duration and DBS therapy has been summa- rized in Table 1.

Subject Age

Gender

Male Female Female Female Male Female Female Male Male Male Male

BIP stimulation parameters

Handedness

Unilateral

Unilateral Bilateral Bilateralb Unilateral Bilateral Bilateral Bilateral Unilateral Bilateral Unilateral Unilateral

vs Bilateral DBS

Disease duration

29 years 60 years 25 years
7 years
15 years 14 years 45 years 10 years 25 years 26 þ years 20 years

DBS therapy duration

4 years
9 months 17 months 3 years
10 months 2 years
1 year
7 months 11 years
3 years
2 years

1 2 3 4 5 6 7 8 9 10 11

74 85 73 70 71 62 62 74 71 76 69

Target Chronic stimulation tested parameters

Right Right Right Right Right Right Right Right Right Right Right

1 VIM
-Right 135Freq

2 VIM
-Left 180Freq

3 VIM
-Right 135Freq

1- Cþ 2.0V 90PW 1- Cþ 2.4V 90PW

Subjective BIP stimulation experience

1- Cþ 1.8V 90PW

1- Cþ 1.8V 90PW 130Freq
1- Cþ 2.0V 90PW 190Freq

1- Cþ 2.4V 90PW 130Freq
1- 2 þ 2.8V 60PW 160Freq

2- Cþ 2.1V 120PW 130Freq
2- 3 þ 3.94V 120PW 190Freq

1- 2 þ 3.9V 150PW 160Freq
2- Cþ 2.3V 90PW 190Freq

1-2- 0 þ 3.4V 120PW 190Freq
2- 1 þ 3.3V 90PW 130Freq

2- Cþ 2.3V 90PW 130Freq

Felt improved dexterity. Transienta surge in the left hand. Dull headache at 1hr of BIP stimulation resolved after 20 min (deemed unrelated).
Felt general movements were improved. Transient tingling right hand.

Felt hesitant speech, uncoordinated and more effortful movements.

Felt decreased leg heaviness, improved writing fluidity and improved voice clarity. Transient tingling along of lip and right forearm.
Transient tingling of jaw.

Felt writing was improved. Transient tingling along lips and right hand.
Felt tremor and dexterity were improved. Transient tingling along lips and left hand.
No positive or negative changes reported.
Felt improved tremor and pouring ability however voice felt hoarse.
Felt improved right hand steadiness. Transient tingling along right hand.
Felt improved speed with drawing lines and spirals. Tingling right thumb for 2 h-post (tolerable).

. a Transient 1⁄4 (seconds).

. b Activa PC.

44 S. De Jesus et al. / Parkinsonism and Related Disorders 46 (2018) 41e46

3.2. Safety and tolerability

All subjects tolerated BiP stimulation, there were no intolerable or adverse effects during the 3 h observation period. Transient non- bothersome paresthesia, lasting seconds was reported by n 1⁄4 8 participants at the initiation of the BiP settings. Transient pares- thesia in clinical practice may be seen with IPG device activation on DBS chronic baseline settings in properly placed leads and these effects were not considered an adverse event in this study [5]. In a singular case there was a non-bothersome paresthesia of the thumb lasting for 2 h, this paresthesia was congruent with the subject’s previously reported paresthesia experienced during the chronic baseline DBS settings and was not considered to be an adverse event. The subjective reports of subjects both positive and negative perceptions, the DBS targets, and the stimulation settings are summarized in Table 1.

3.3. Motor assessment

Inter-rater (two raters) blinded assessment of the TRS in eleven subjects by independent videotape review revealed an ICC > 95% throughout the six conditions. There was a statistically significant improvement in motor TRS median scores decreasing from washout OFF DBS (28.5 IQR 1⁄4 24.5e35.25) and ClinDBS (20 IQR 1⁄4 13.8e24.3) to (16.5 IQR 1⁄4 12.5e22.25) at 3 h ON BiP, this was a consistent improvement (15.5IQR 1⁄4 12.5e21.25), (17.5 IQR 1⁄4 11- 22) and (16.5 IQR 1⁄4 12e20.75) at 30min, 1 h, 2 h ON BiP stimulation respectively, (c2 1⁄4 34.8, p 1⁄4 0.00002), Fig. 2. Median specific sub- score post-hoc pairwise comparisons for contralateral arm tremor, tremor at rest/postural/action, TRS drawing A/B/C, pouring and motor contralateral TRS scores all revealed statistical signifi- cance (p 1⁄4 0.007 – <0.001) seen with ON BiP versus washout OFF DBS. TRS total motor contralateral median scores at washout DBS OFF (14.5 IQR 1⁄4 9.50e17.00), ClinDBS (7.0 IQR 1⁄4 4.75e10.25), 30mins ON BiP at (6.0 IQR 1⁄4 4.00e7.25), 1 h ON BiP at (7.0 IQR 1⁄4 3.75e8.25), 2 h ON BiP at (6.0 IQR 1⁄4 4.00e9.25) and 3 h ON BiP DBS (6.0 IQR 1⁄4 4.50e8.25), (c2 1⁄4 37.5, p 1⁄4 <0.001). TRS motor contralateral arm tremor (at rest, posture, action) median scores at washout (DBS OFF: 5.5 (IQR 1⁄4 2.00e6.00)), ClinDBS (2.0 IQR 1⁄4 1.00e4.00), 30mins ON BiP at (1.5 IQR 1⁄4 1.00e2.00), once hour ON BiP at (2.0 IQR 1⁄4 0.00e2.25), 2 h ON BiP at (2.0 IQR 1⁄4 0.00e2.00) and 3 h ON BiP DBS (2.0 IQR 1⁄4 1.00e2.25), (c2 1⁄4 40.0, p 1⁄4 <0.001).

Fig. 2. Median TRS Blinded Scores in 11 ET patients over 6 cinical conditions (opti- mized baseline conventional chronic setting, 30 min DBS OFF washout period, 30-min post, 1 h-post, 2 h-post, 3 h-post ON BIP stimulation). Noted change of TRS median score from 20 at baseline conventional setting to 16.5 after 3 h ON BIP stimulation.

Kinesia accelerometer analysis for tremor at rest, posture and action were available for 10 of 11 ET subjects. One subject had a single missing condition and was excluded from analysis. There was a statistically significant improvement in tremor at rest, posture and with action at the 3 h On BiP stimulation when comparing median total scores to ClinDBS as well as the washout condition (OFF DBS). Rest tremor median scores decreased from ClinDBS (0.26 IQR 1⁄4 0.00e0.71) and washout DBS OFF (0.58 IQR 1⁄4 0.19e0.99) to (0.00 IQR 1⁄4 0.00e0.03) at 30min ON BiP, (0.03 IQR 1⁄4 0.00e0.35) at 1 h ON BiP and (0.00 IQR 1⁄4 0.00e0.06) at 2 h ON BiP and (0.01 IQR 1⁄4 0.00e0.10) at 3 h ON BiP stimulation, (c2 1⁄4 16.1, p 1⁄4 0.006). Postural tremor median scores, decreased from (0.28 IQR 1⁄4 0.04e0.62) and (1.63 IQR 1⁄4 0.86e3.04) at ClinDBS and washout DBS OFF conditions, respectively, to (0.17 IQR 1⁄4 0.00e0.52) at 30min ON BiP, (0.08 IQR 1⁄4 0.00e0.41) at 1 h ON BiP, (0.18 IQR 1⁄4 0.08e0.35) at 2 h ON BiP and (0.20 IQR 1⁄4 0.04e0.37) at 3 h ON BiP stimulation, (c2 1⁄4 15.9, p 1⁄4 0.007). Kinetic tremor median scores, decreased from (0.53 IQR 1⁄4 0.32e0.94) and (1.80 IQR 1⁄4 0.95e3.96) at ClinDBS and washout DBS OFF conditions, respectively, to (0.00 IQR 1⁄4 0.00e0.31) at 30min ON BiP, (0.35 IQR 1⁄4 0.00e0.49) at 1 h ON BiP, (0.13 IQR 1⁄4 0.00e0.50) at 2 h ON BiP and (0.26 IQR 1⁄4 0.00e0.52) at 3 h ON BiP stimulation, (c2 1⁄4 32.1, p1⁄4<0.001). The Kinesia accelerometer assessments have been summarized in Fig. 3.

3.4. Gait assessment

GaitRite analysis was completed for 11 participants. There was not a significant effect of the stimulation conditions on gait pa- rameters, Wilks’ Lambda 1⁄4 0.06, F (100,131.5) 1⁄4 1.065, p 1⁄4 0.37. Individual gait parameters comparing DBS OFF, ClinDBS and 3 h ON BiP stimulation conditions are available as supplementary material.

4. Discussion

The current study demonstrated the safety and tolerability of applying a nonconventional pulse shape (square biphasic DBS pulses) for longer time intervals in the treatment of ET. The primary aims of safety and tolerability were both met by all participating subjects. There were no study withdrawals and no DBS adverse events incurred by using a BiP waveform. There were reports of transient paresthesia, however this symptom is typical and ex- pected when applying DBS in the VIM target [5]. These subjects reported it as a typical transient sensation when changing their DBS group settings at home and found it a non-bothersome issue. One participant reported a prolonged 2 h post BiP mild non-bothersome hand paresthesia, without compromising hand function, which was his typical reactivation phenomenon when turning his device off and on at home.

Clinical effectiveness of using longer periods of BiP DBS was also demonstrated by blinded TRS ratings and by objective and sensitive accelerometry data. BiP DBS reduced total motor TRS median score compared to ClinDBS and washout DBS OFF evaluations by 3.5 points and 12 points respectively. These improvements were sup- ported subjectively by the perceived clinical benefit by the patients and supported objectively by improvements detected by the accelerometer recordings over repeated measures. Another obser- vation across the patients was that the motor symptoms remained consistently improved over the 3 h observation period. Gait did not worsen with unilateral BiP DBS as measured by walking on the GaitRite. Participants fell within “normal” ranges for the gait pa- rameters observed at baseline [26,27]. This is important because previous reports of conventional DBS have revealed changes in gait parameters [22]. Gait was not acutely impacted by unilateral BiP

S. De Jesus et al. / Parkinsonism and Related Disorders 46 (2018) 41e46 45

Fig. 3. Kinesia accelerometer median tremor assessments with interquartile range for postural tremor (A), kinetic tremor (B) and rest tremor (C).

DBS as measured by walking on the GaitRite. Future larger bilateral studies will be needed to confirm this observation and also to document fall frequency.

Current FDA approved DBS devices for movement disorders and epilepsy apply a high-amplitude short stimulation pulse (cathodal phase) and follow this stimulation by a low-amplitude long recharge period (anodal phase) [9,28]. Alternatively, BiP delivers an active symmetric recharge pulse with the identical amplitude and pulse width applied during the initial stimulation phase. Cathodal stimulation has been shown to activate structures by partially depolarizing membranes whereas anodal stimulation may have an inhibitory role by decelerating the depolarizing effect of the cath- odal phase [28]. However, anodal stimulation has been shown to excite neural elements within the cortex and brainstem structures as well as in square biphasic pulse models [28,29]. Yamamoto et al. additionally described effective tremor suppression in ET and post stroke patients via combination of cathodal and anodal stimulation within different regions of the thalamus suggesting a summative effect [30]. Providing an equivalent charge during the anodal phase may potentiate more effective neuromodulation either through providing an additional excitatory effect on the neuronal structures, by delivering more efficient intracellular stimulation, or by another, yet unknown mechanism. The additional stimulus does have a cost of greater battery consumption as compared to conventional DBS which has been detailed by Akbar et al. Though this energy con- sumption issue may become less important as rechargeable neu- rostimulators become more widely available [7]. If our findings can be confirmed in larger studies, this may also be a future alternative programming option for ET patients who may not respond well to conventional DBS pulses.

Strengths of the present study were the use of post-surgical optimized patients as their own controls. This limited the possi- bility of confounding because of the microlesion effect. Addition- ally, the use of two blinded raters and also physiological measures of tremor and gait lessened the inherent risk of using bedside rating scales. The washout period for the original study by Akbar et al. was only 2 min, and this was lengthened in the current study to an observation period of 3 h. There were several potential limitations of the current study. The study had a small sample size and used only unilateral DBS ON BiP stimulation. There were minor software limitations as BiP DBS frequencies had to be matched as closely as possible. This was however easily achievable as the frequency was increased or decreased to the closest available frequency and all remained within high frequency domains. The frequency issue should be addressed in future studies where software adjustments can be made to facilitate exact frequencies. Due to the statistical analysis depending on medians, there was a floor effect on median TRS values when comparing BiP to chronic conventional stimula- tion. Despite the statistical significance in the Friedman’s test, post- hoc pairwise comparisons did not have statistical power to demonstrate a difference in accelerometer scores between BiP and baseline, although the median values did improve. Therefore these findings should limit interpretation of the results to the acute 3 h in-office time period, however setting the stage for a well-powered study to determine clinical efficacy in comparison to conventional DBS, based on a minimally clinical significant change in ET scores and the use of ancillary measures such as accelerometer data.

In conclusion, BiP DBS was safe and tolerable for the treatment of ET. The study results suggest that the effects can suppress tremor, though a larger longer duration study will be required.

46 S. De Jesus et al. / Parkinsonism and Related Disorders 46 (2018) 41e46

Disclosures

RR is a paid employee of Medtronic Neuromodulation Global Research. MSO serves as a consultant for the National Parkinson Foundation, and has received research grants from NIH, NPF, the Michael J. Fox Foundation, the Parkinson Alliance, Smallwood Foundation, the Bachmann-Strauss Foundation, the Tourette Syn- drome Association, and the UF Foundation. MSO’s DBS research is supported by: R01 NR014852. MSO has previously received hono- raria, but in the past >60 months has received no support from industry. MSO has received royalties for publications with Demos, Manson, Amazon, Smashwords, Books4Patients, and Cambridge (movement disorders books). MSO is an associate editor for New England Journal of Medicine Journal Watch Neurology. MSO has participated in CME and educational activities on movement dis- orders (in the last 36) months sponsored by PeerView, Prime, QuantiaMD, WebMD, MedNet, Henry Stewart, and by Vanderbilt University. The institution and not MSO receives grants from Medtronic, Abbvie, Allergan, and ANS/St. Jude, and the PI has no financial interest in these grants. MSO has participated as a site PI and/or co-I for several NIH, foundation, and industry sponsored trials over the years but has not received honoraria.

Acknowledgements

We express gratitude to the UF Foundation and to the patients and caregivers who made this study possible.

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.015.

References

. [1] E.D. Louis, R. Ottman, How many people in the USA have essential tremor? Deriving a population estimate based on epidemiological data, Tremor Other Hyperkinet Mov. (N Y) 4 (2014) 259.

. [2] J.F. Baizabal-Carvallo, M.N. Kagnoff, J. Jimenez-Shahed, R. Fekete, J. Jankovic, The safety and efficacy of thalamic deep brain stimulation in essential tremor: 10 years and beyond, J. Neurol. Neurosurg. Psychiatry 85 (5) (2014) 567e572.

. [3] M.J. Birdno, W.M. Grill, Mechanisms of deep brain stimulation in movement disorders as revealed by changes in stimulus frequency, Neurotherapeutics 5 (1) (2008) 14e25.

. [4] R.J. Elble, Mechanisms of deep brain stimulation for essential tremor, Brain 137 (Pt 1) (2014) 4e6.

. [5] R.E. Gross, J.D. Rolston, The clinical utility of methods to determine spatial extent and volume of tissue activated by deep brain stimulation, Clin. Neu- rophysiol. 119 (9) (2008) 1947e1950.

. [6] C.C. McIntyre, M. Savasta, L. Kerkerian-Le Goff, J.L. Vitek, Uncovering the mechanism(s) of action of deep brain stimulation: activation, inhibition, or both, Clin. Neurophysiol. 115 (6) (2004) 1239e1248.

. [7] U. Akbar, R.S. Raike, N. Hack, C.W. Hess, J. Skinner, D. Martinez-Ramirez, et al., Randomized, blinded pilot testing of nonconventional stimulation patterns and shapes in Parkinson’s disease and essential tremor: evidence for further evaluating narrow and biphasic pulses, Neuromodulation 19 (4) (2016)

343e356.
[8] M.J. Birdno, S.E. Cooper, A.R. Rezai, W.M. Grill, Pulse-to-pulse changes in the

frequency of deep brain stimulation affect tremor and modeled neuronal

activity, J. Neurophysiol. 98 (3) (2007) 1675e1684.
[9] T.J. Foutz, C.C. McIntyre, Evaluation of novel stimulus waveforms for deep

brain stimulation, J. Neural Eng. 7 (6) (2010), 066008.
[10] M.J. Birdno, A.M. Kuncel, A.D. Dorval, D.A. Turner, R.E. Gross, W.M. Grill,

Stimulus features underlying reduced tremor suppression with temporally

patterned deep brain stimulation, J. Neurophysiol. 107 (1) (2012) 364e383. [11] D.T. Brocker, B.D. Swan, D.A. Turner, R.E. Gross, S.B. Tatter, M.M. Koop, et al., Improved efficacy of temporally non-regular deep brain stimulation in Par-

kinson’s disease, Exp. Neurol. 239 (2013) 60e67.
[12] C.W. Hess, D.E. Vaillancourt, M.S. Okun, The temporal pattern of stimulation

may be important to the mechanism of deep brain stimulation, Exp. Neurol.

247 (2013) 296e302.
[13] B.D. Swan, W.M. Grill, D.A. Turner, Investigation of deep brain stimulation

mechanisms during implantable pulse generator replacement surgery, Neu-

romodulation 17 (5) (2014) 419e424 discussion 24.
[14] D.T. Brocker, W.M. Grill, Principles of electrical stimulation of neural tissue,

Handb. Clin. Neurol. 116 (2013) 3e18.
[15] M. Sahin, Y. Tie, Non-rectangular waveforms for neural stimulation with

practical electrodes, J. Neural Eng. 4 (3) (2007) 227e233.
[16] A. Wongsarnpigoon, W.M. Grill, Energy-efficient waveform shapes for neural stimulation revealed with a genetic algorithm, J. Neural Eng. 7 (4) (2010),

046009.
[17] A. Wongsarnpigoon, J.P. Woock, W.M. Grill, Efficiency analysis of waveform

shape for electrical excitation of nerve fibers, IEEE Trans. Neural Syst. Rehabil.

Eng. 18 (3) (2010) 319e328.
[18] G. Deuschl, P. Bain, M. Brin, Consensus statement of the movement disorder

society on tremor, Ad Hoc Sci. Comm. Mov. Disord. 13 (Suppl 3) (1998) 2e23. [19] R.J. Coffey, Deep brain stimulation devices: a brief technical history and re-

view, Artif. Organs 33 (3) (2009) 208e220.
[20] J.P. Giuffrida, D.E. Riley, B.N. Maddux, D.A. Heldman, Clinically deployable

Kinesia technology for automated tremor assessment, Mov. Disord. 24 (5)

(2009) 723e730.
[21] D.A. Heldman, J.P. Giuffrida, R. Chen, M. Payne, F. Mazzella, A.P. Duker, et al.,

The modified bradykinesia rating scale for Parkinson’s disease: reliability and comparison with kinematic measures, Mov. Disord. 26 (10) (2011) 1859e1863.

[22] G.M. Earhart, B.R. Clark, S.D. Tabbal, J.S. Perlmutter, Gait and balance in essential tremor: variable effects of bilateral thalamic stimulation, Mov. Dis- ord. 24 (3) (2009) 386e391.

[23] N. Hwynn, C.J. Hass, P. Zeilman, J. Romrell, Y. Dai, S.S. Wu, et al., Steady or not following thalamic deep brain stimulation for essential tremor, J. Neurol. 258 (9) (2011) 1643e1648.

[24] A. Ramirez-Zamora, H. Boggs, J.G. Pilitsis, Reduction in DBS frequency im- proves balance difficulties after thalamic DBS for essential tremor, J. Neurol. Sci. 367 (2016) 122e127.

[25] G. Baldewijns, G. Verheyden, B. Vanrumste, T. Croonenborghs, Validation of the kinect for gait analysis using the GAITRite walkway, Conf. Proc. IEEE Eng. Med. Biol. Soc. 2014 (2014) 5920e5923.

[26] M. Hoskovcova, O. Ulmanova, O. Sprdlik, T. Sieger, J. Novakova, R. Jech, et al., Disorders of balance and gait in essential tremor are associated with midline tremor and age, Cerebellum 12 (1) (2013) 27e34.

[27] R.T. Roemmich, P.R. Zeilman, D.E. Vaillancourt, M.S. Okun, C.J. Hass, Gait variability magnitude but not structure is altered in essential tremor, J. Biomech. 46 (15) (2013) 2682e2687.

[28] L. Hofmann, M. Ebert, P.A. Tass, C. Hauptmann, Modified pulse shapes for effective neural stimulation, Front. Neuroeng 4 (2011) 9.

[29] C. Cedzich, U. Pechstein, J. Schramm, S. Schafer, Electrophysiological consid- erations regarding electrical stimulation of motor cortex and brain stem in humans, Neurosurgery 42 (3) (1998) 527e532.

[30] T. Yamamoto, Y. Katayama, T. Kano, K. Kobayashi, H. Oshima, C. Fukaya, Deep brain stimulation for the treatment of parkinsonian, essential, and poststroke tremor: a suitable stimulation method and changes in effective stimulation intensity, J. Neurosurg. 101 (2) (2004) 201e209.

Parkinsonism and Related Disorders 46 (2018) 47e55

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

MR planimetry in neurodegenerative parkinsonism yields high diagnostic accuracy for PSP

Stephanie Mangesius a, b, Anna Hussl a, Florian Krismer a, *, Philipp Mahlknecht a,
Eva Reiter a, Susanne Tagwercher a, Atbin Djamshidian a, Michael Schocke b, c,
Regina Esterhammer b, Gregor Wenning a, Christoph Müller a, Christoph Scherfler a, c, Elke R. Gizewski b, c, Werner Poewe a, c, Klaus Seppi a, c, *

a Department of Neurology, Medical University Innsbruck, Innsbruck, Austria
b Department of Neuroradiology, Medical University Innsbruck, Innsbruck, Austria c Neuroimaging Core Facility, Medical University Innsbruck, Innsbruck, Austria

articleinfo abstract

Article history:

Received 28 February 2017 Received in revised form
1 October 2017
Accepted 30 October 2017

Keywords:

Magnetic resonance imaging (MRI) Parkinsonism
Planimetry
Differential diagnosis

Diagnostic accuracy

Introduction: Several previous studies examined different brainstem-derived MR planimetric measures with regards to their diagnostic accuracy in separating patients with neurodegenerative parkinsonian disorders and reported conflicting results. The current study aimed to compare their performance in a well-characterized sample of patients with neurodegenerative parkinsonian disorders.

Methods: MR planimetric measurements were assessed in a large retrospective cohort of 55 progressive supranuclear palsy (PSP), 194 Parkinson’s disease (PD) and 63 multiple system atrophy (MSA) patients. This cohort served as a training set used to build C4.5 decision tree models to discriminate PSP, PD and MSA. The models were validated in two independent test sets. The first test set comprised 84 patients with early, clinically unclassifiable parkinsonism (CUP). A prospective cohort of patients with PSP (n 1⁄4 23), PD (n 1⁄4 40) and MSA (n 1⁄4 22) was exploited as a second test-set.

Results: The pons-to-midbrain diameter ratio, the midbrain diameter, the middle cerebellar peduncle width and the pons area were identified as the most predictive parameters to separate PSP, MSA and PD in C4.5 decision tree models derived from the training set. Using these decision models, AUCs in discriminating PSP, MSA and PD were 0.90, 0.57 and 0.73 in the CUP-cohort and 0.95, 0.61 and 0.87 in the prospective cohort, respectively.

Conclusion: We were able to demonstrate that brainstem-derived MR planimetric measures yield high diagnostic accuracy for the discrimination of PSP from related disorders when decision tree algorithms are applied, even at early, clinically uncertain stages. However, their diagnostic accuracy in discrimi- nating PD and MSA was suboptimal.

* Corresponding authors. Medical University of Innsbruck, Department of Neurology, Anichstraße 35, 6020 Innsbruck, Austria.

E-mail addresses: stephanie.mangesius@tirol-kliniken.at (S. Mangesius), anna. hussl@tirol-kliniken.at (A. Hussl), florian.krismer@i-med.ac.at (F. Krismer), philipp.mahlknecht@i-med.ac.at (P. Mahlknecht), eva-magdalena.reiter@student.i- med.ac.at (E. Reiter), susanne.tagwercher@gmx.at (S. Tagwercher), atbin. djamshidian-tehrani@i-med.ac.at (A. Djamshidian), michael.schocke@rku.de (M. Schocke), regina.esterhammer@i-med.ac.at (R. Esterhammer), gregor. wenning@i-med.ac.at (G. Wenning), christoph.mueller@tirol-kliniken.at (C. Müller), christoph.scherfler@i-med.ac.at (C. Scherfler), elke.gizewski@tirol- kliniken.at (E.R. Gizewski), werner.poewe@i-med.ac.at (W. Poewe), klaus.seppi@ tirol-kliniken.at (K. Seppi).

1. Introduction

Although stringent consensus operational criteria for different neurodegenerative parkinsonian disorders exist [1,2], early differ- ential diagnosis of progressive supranuclear palsy (PSP) and related neurodegenerative parkinsonian disorders including multiple sys- tem atrophy (MSA), corticobasal syndrome (CBS) and Parkinson’s disease (PD) remains challenging. Error rates in initial clinical diagnosis have been as high as one quarter in series with post- mortem confirmation highlighting the need for diagnostic bio- markers [3e5].

Several MR planimetric measures to quantify atrophy of specific

https://doi.org/10.1016/j.parkreldis.2017.10.020
1353-8020/© 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

48 S. Mangesius et al. / Parkinsonism and Related Disorders 46 (2018) 47e55

brainstem structures have been proposed as disease-specific markers [3,5e9]. Currently, it is unclear which of these are most useful in clinical routine.

Hence, the aim of the present study was to evaluate the diag- nostic application of different brainstem-derived MR planimetric measures in a large cohort of patients with neurodegenerative parkinsonism including subjects with early parkinsonism with clinically uncertain classification.

2. Methods

2.1. Study population

Patients included in the prospective cohort were recruited consecutively over a period of 18 months with a predefined patient number per diagnostic entity (2011e2013). Eligible participants for the retrospective cohort were identified on the basis of an MRI database on movement disorders patients followed at our depart- ment from 2000 to 2011 whose cerebral MRI protocol had included a 3-dimensional magnetization-prepared rapid gradient-echo (3D- MPRAGE) (T1) sequence. Clinical diagnoses of PSP, PD and MSA were established according to consensus research criteria by 2 movement disorder experts. Criteria for PSP [2] were modified to include recent criteria for PSP-P [1]. Because a substantial number of patients with CBS show a neuropathological pattern of PSP [10], patients with clinically defined CBS in the retrospective cohort were combined with the PSP group. We have also included MSA-C patients, if they had parkinsonism. Patients of the retrospective cohort who e at the time of MRI e were considered indeterminate by their treating neurologist were classified as clinically unclassi- fiable parkinsonism (CUP). Reasons for uncertainty were assessed by movement disorder specialists. For qualification as CUP, patients had to meet at least one of the following criteria: presence of two cardinal signs of parkinsonism if the patient presented without bradykinesia [11]; parkinsonism of mild intensity; newly diagnosed untreated parkinsonian patients, or disease duration of less than 18 months. Moreover, they had to fulfill step two of the UK Parkinson’s Disease Society Brain Bank Diagnostic Criteria [12]. All patients were regularly followed-up for at least 24 months and the clinical diagnosis at last visit was considered as final diagnosis [1,2]. All patients from the retrospective cohort had at least a clinical follow- up after one year. However, patients with CUP were followed for at least 2 years, and a final diagnosis of PD in the CUP cohort required confirmation after at least 4 years.

The study was approved by the local Ethics Committee of the Medical University Innsbruck and participants enrolled provided written informed consent.

A subset of patients of our study has been reported previously in the context of two studies of our group [9,13].

2.2. Magnetic resonance imaging protocol and image analysis

All subjects received high-resolution MR images on three different 1.5 T Siemens (Erlangen, Germany) MR Scanners comprising a T1-weighted 3D-MPRAGE sequence and a transversal double-echo fast spin echo sequence with T2 and proton density contrast. Further information is given in Supplementary Methods.

These sequences were visually assessed by experienced neuro- radiologists (MS, RE) to exclude symptomatic parkinsonism [3].

Midsagittal midbrain and pontine areas (MA and PA) and di- ameters (Md and Pd), middle and superior cerebellar peduncle di- ameters (MCPd and SCPd) were measured according to methods described previously (supplementary methods and supplementary figure) [6,8,9].

All readers of MR images were blinded to any subject

information.

2.3. Statistical analysis

Parametric, nonparametric, or chi-square tests were used for group comparisons depending on the scale type of the variables. One-way analyses of variance (ANOVA) with post-hoc Bonferroni correction were performed to calculate differences regarding dis- ease duration, age, Hoehn and Yahr Scale (H&Y) and Unified Par- kinson’s Disease Rating Scale (UPDRS) part II and III, and to evaluate differences in planimetric measurements between groups.

Receiver Operating Characteristic (ROC) curves statistics with their area under the curve (AUCs) for Pd, Md, Md/Pd, MA, PA, MA/PA, MRPI, SCPd and MCPd were calculated for the discrimination of PSP from non-PSP, PD from non-PD and MSA from non-MSA parkin- sonism in the retrospective cohort. Based on the same MR mea- sures, a MRI-based decision tree exploiting the C4.5 classifier algorithm [14] with reduced error pruning (3 folds were used for reduced error pruning) implemented by the Waikato Environment of Knowledge Analysis machine learning software (WEKA 3.8.1) was developed to classify clinically diagnosed MSA, PSP, and PD patients. The retrospective cohort was used as training set to establish the classification model, and two independent cohorts were used to validate classification performance of the decision tree.

The first test set comprised 84 patients with CUP and the pro- spective cohort was exploited as a second test-set. The classifica- tion performance of the C4.5 classifier algorithm is expressed as AUC, predictive accuracy as well as sensitivity and specificity. Diagnostic accuracy values were also calculated for the single MR measures that reached an AUC > 0.800 in the training set. Optimal cut-off values were determined by the Youden Index. These cut-off values were also applied to both test sets and their diagnostic ac- curacy values are also given.

3. Results

3.1. Clinical and demographic data

Demographic, clinical and MRI data of the study participants are shown in Table 1. Three patients of the prospective cohort/test set, who were diagnosed as having PD at the time of scanning, were reclassified as MSA (1 case) or PSP (2 cases) during clinical follow- up. Patients with PSP or MSA had significantly shorter disease duration and higher H&Y stage as compared to PD in the retro- spective and prospective cohort.

3.2. MRI data

Results of brainstem-derived MR planimetric measures are summarised in Table 1.

It took significantly longer to assess MRPI (161 ± 9.1 s) and MA/ PA (67 ± 6.1 s) [9] than to perform measurements of Md (20.1 ± 3.4 s), Pd (19.6 ± 1.9 s) and Md/Pd (39.7 ± 4.6 s) as assessed on 3 PSP, 3 MSA and 4 PD patients.

3.3. C4.5 decision tree calculation

Three disease-specific decision trees were developed to classify patients as PSP, MSA or PD (Fig. 1). Pd/Md and Md were identified as the most significant parameters to construct the prediction model for the discrimination of PSP, MCPd and Pd/Md for the discrimina- tion of MSA, and PA, Md, Pd/Md and MCPd for the discrimination of PD. In the training (Table 2, Fig. 1) and both test sets (Table 3), all predictive accuracies were higher for the disease-specific decision

S. Mangesius et al. / Parkinsonism and Related Disorders 46 (2018) 47e55 49

Table 1

Demographic, clinical and MRI data of patients in the retrospective cohort/training set, prospective cohort/re-test set, and CUP cohort/test set. Results are reported as means ± SD.

n Gender

55
(33 RS, 12 PSP-P, 10 CBS)
28/27

194 122/72

63
(44 MSA-P, 19 MSA-C*) 28/35

0.028/ 0.606/ 1.000/ 0.036 <0.001/ 0.108/ <0.001/ 0.008 <0.001/ <0.001/ 1.000/ <0.001 <0.001/ <0.001/ 1.000/ <0.001

17
(12 RS, 5 PSP- P)
10/7

55 36/19

12
(11 MSA-P, 1 MSA-C*)
7/5

0.826/ 1.000/ 1.000/ 1.000 0.596/ 0.970/ 1.000/ 1.000 0.003/ <0.001/ 0.300/ 0.522 0.640/ 1.000/ 1.000/ 1.000

23
(15 RS, 8 PSP- P)
16/7

40 25/15

22
(16 MSA-P, 6 MSA-C*) 12/10

0.582/ 1.000/ 1.000/ 1.000 0.165/ 0.636/ 0.181/ 1.000 <0.001/ 0.006/ 0.909/ <0.001 <0.001/ <0.001/ 1.000/ <0.001 <0.001/ <0.001/ 0.177/ <0.001 <0.001/ 0.396/ 0.010/ <0.001 <0.001/ 0.012/ 0.010/ <0.001

distributiona (m/f)

Age at MRI mean ± SD b

69.82 ± 10.30

66.88 ± 8.67

62.98 ± 8.14

66.79 ± 9.59

64.29 ± 9.21

65.54 ± 7.33

68.21 ± 5.77

65.41 ± 9.41

63.38 ± 9.09

Hoehn and Yahr ± SD a

3.51 ± 0.88

2.40 ± 0.76

3.42 ± 1.01

2.44 ± 0.50

1.65 ± 0.63

2.13 ± 0.64

3.14 ± 0.74

2.40 ± 0.63

3.39 ± 0.596

Disease
Duration ± SD b

3.72 ± 1.64

8.09 ± 6.50

4.07 ± 2.40

0.93 ± 0.39

0.83 ± 0.41

0.90 ± 0.43

2.07 ± 1.53

5.96 ± 4.17

2.13 ± 1.51

UPDRS II c

17.95 ± 5.90

11.61 ± 5.37

21.38 ± 5.40

UPDRS III c

31.77 ± 10.09

27.76 ± 11.71

42.19 ± 9.11

UPDRS total c

49.72 ± 17.67

39.37 ± 14.33

63.57 ± 11.91

Midbrain diameter

mean ± SDb

7.78 ± 1.11

10.16 ± 0.81

9.82 ± 0.74

<0.001/ <0.001/ <0.001/ 0.018

7.76 ± 1.11

10.20 ± 1.04

9.66 ± 0.81

<0.001/ <0.001/ <0.001/ 0.308

7.58 ± 0.78

9.98 ± 0.73

9.61 ± 0.82

<0.001/ <0.001/ <0.001/ 0.226

Pontine diameter

mean ± SDb

16.75 ± 1.35

16.97 ± 1.55

14.63 ± 2.45

<0.001/ 1.000/ <0.001/ <0.001

17.30 ± 1.39

16.72 ± 1.44

16.48 ± 1.80

0.272/ 0.482/ 0.446/ 1.000

17.50 ± 1.11

16.81 ± 1.49

15.83 ± 2.23

<0.001/ 0.325/ 0.003/ 0.082

Midbrain area

mean ± SDb

80.80 ± 16.32

116.05 ± 19.76

107.69 ± 15.53

<0.001/ <0.001/ <0.001/ 0.006

77.75 ± 16.53

115.86 ± 18.59

106.00 ± 8.91

<0.001/ <0.001/ <0.001/ 0.226

71.75 ± 14.59

112.66 ± 12.40

112.77 ± 17.54

<0.001/ <0.001/ <0.001/ 1.000

Pontine area

mean ± SDb

503.39 ± 62.11

542.83 ± 60.61

427.24 ± 105.44

<0.001/ 0.002/

509.63 ± 65.81

532.69 ± 53.02

505.06 ± 49.58

0.148/ 0.412/

523.08 ± 38.11

544.50 ± 49.16

451.62 ± 89.07
(continued on next page)

retrospective cohort/training set

CUP cohort/test set

prospective cohort/re-test

set PD

PSP

MSA

P value for group comparisons**

PSP

MSA

P value for group comparisons**

PSP

MSA

P value for group comparisons**

<0.001/ 0.524/

50 S. Mangesius et al. / Parkinsonism and Related Disorders 46 (2018) 47e55

Table 1 (continued )

Mean MCPd mean ± SDb

8.73 ± 1.05

9.51 ± 1.08

6.83 ± 1.90

<0.001/ <0.001/ <0.001/ <0.001

8.41 ± 1.24

9.79 ± 1.32

8.67 ± 1.11

<0.001/ 0.001/ 1.000/ 0.021

9.62 ± 1.06

10.48 ± 0.92

8.41 ± 1.92

<0.001/ 0.039/ 0.007/ <0.001

Mean SCPd mean ± SDb

2.87 ± 0.55

3.40 ± 0.39

2.87 ± 0.63

<0.001/ <0.001/ 1.000/ <0.001

2.82 ± 0.40

3.40 ± 0.38

3.46 ± 0.44

<0.001/ <0.001/ <0.001/ 1.000

3.02 ± 0.56

3.63 ± 0.43

3.44 ± 0.63

<0.001/ <0.001/ 0.026/ 0.481

M d/Pd mean ± SDb

0.47 ± 0.07

0.60 ± 0.06

0.69 ± 0.14

<0.001/ <0.001/ <0.001/ <0.001

0.45 ± 0.06

0.61 ± 0.07

0.59 ± 0.08

<0.001/ <0.001/ <0.001/ 1.000

0.43 ± 0.05

0.60 ± 0.06

0.62 ± 0.10

<0.001/ <0.001/ <0.001/ 0.710

MA/PA mean ± SDb

0.16 ± 0.03

0.21 ± 0.03

0.27 ± 0.07

<0.001/ <0.001/ <0.001/ <0.001

0.15 ± 0.03

0.22 ± 0.04

0.21 ± 0.02

<0.001/ <0.001/ <0.001/ 1.000

0.14 ± 0.03

0.21 ± 0.02

0.26 ± 0.07

<0.001/ <0.001/ <0.001/ <0.001

MRPI

mean ± SDb

20.04 ± 4.89

13.43 ± 2.52

9.70 ± 3.21

<0.001/ <0.001/ <0.001/ <0.001

20.29 ± 5.26

13.65 ± 2.99

12.07 ± 1.96

<0.001/ <0.001/ <0.001/ 0.464

25.04 ± 8.01

14.15 ± 2.02

10.24 ± 3.61

<0.001/ <0.001/ <0.001/ 0.008

retrospective cohort/training set

CUP cohort/test set

prospective cohort/re-test set

PSP

MSA

P value for group comparisons**

PSP

MSA

P value for group comparisons**

PSP

MSA

P value for group comparisons**

<0.001/ <0.001

1.000/ 0.363

<0.001/ <0.001

* MSA-C presenting with parkinsonism (requiring dopaminergic treatment).
** p-value for overall group comparison/PSP versus PD/PSP versus MSA/MSA versus PD.
The significance level is set at P < 0.05. P-values of post-hoc comparisons are adjusted by Bonferroni correction for multiple comparisons.
Abbreviations: CUP 1⁄4 clinically uncertain parkinsonian syndromes; PD 1⁄4 Parkinson disease; MSA 1⁄4 multiple system atrophy; MSA-P 1⁄4 parkinsonian variant of multi system atrophy; MSA-C 1⁄4 cerebellar variant of multi system atrophy; PSP 1⁄4 progressive supranuclear palsy; PSP-P 1⁄4 parkinsonian variant of progressive supranuclear palsy; RS 1⁄4 Richardson syndrome; CBS 1⁄4 corticobasal degeneration; MRI 1⁄4 magnet resonance imaging; m 1⁄4 male; f 1⁄4 female; SD 1⁄4 standard deviation; H&Y 1⁄4 Hoehn and Yahr Scale; UPDRS 1⁄4 Unified Parkinson’s Disease Rating Scale; SCPd 1⁄4 superior cerebellar peduncle; MCPd 1⁄4 middle cerebellar peduncle; Md/Pd-ratio 1⁄4 midbrain-to- pontine-diameter-ratio; MA/PA-ratio 1⁄4 midbrain-to-pons-area-ratio; MRPI 1⁄4 Magnetic Resonance Parkinsonism Index.

. a Chi-square test.

. b Parametric tests (unpaired t-test; univariate one-way analysis of variance, ANOVA).

. c Nonparametric tests (Mann-Whitney U test; Kruskal-Wallis one-way analysis of variance).

Fig. 1. Decision algorithms derived from the training set (i.e. retrospective cohort) for the differential diagnosis of PSP vs non-PSP, MSA vs non-MSA and PD vs non-PS; values give optimal cut-offs for the differential diagnosis and the ratio of correctly classified/ misclassified. Sensitivity, specificity and predictive accuracy of the decision algorithms are 81.8% (95%CI 69.1e90.9%), 98.4% (95%CI 96.1e99.6%), 95.5% (95%CI 92.6e97.5%) for PSP from non-PSP, 54.0% (95%CI 40.9e66.6%), 100.0% (95%CI 98.5e100.0%), 90.7% (95% CI 86.9e93.7%) for MSA from non-MSA and 92.8% (95%CI 88.2e96.0%), 76.3% (95%CI 67.6e83.6%), 86.5% (95%CI 82.2e90.1%) for PD from non-PD parkinsonism.

legend: PSP 1⁄4 progressive supranuclear palsy; MSA 1⁄4 multiple system atrophy; PD 1⁄4 Parkinson disease; Pd/Md 1⁄4 pons-to- midbrain -diameter-ratio; Md 1⁄4 midbrain diameter; MCPd 1⁄4 middle cerebellar peduncle diameter; PA 1⁄4 pons area.

trees versus the ROC-derived diagnostic accuracy values except for the detection of MSA where the discrimination with MCPd and MRPI was superior to the C4.5 algorithm based discrimination in the prospective cohort. The supplementary table summarizes the discriminative measures including number of false negatives and positives of the classification performance obtained by the decision trees relative to the clinical diagnoses in the two test sets. Using the decision models, AUCs in discriminating PSP from non-PSP, MSA from non-MSA and PD from non-PD parkinsonism were 0.90, 0.57 and 0.73 in the CUP-cohort and 0.95, 0.61 and 0.87 in the pro- spective cohort, respectively.

Diagnostic accuracy of decision tree algorithms, for the discrimination of PSP from non-PSP parkinsonism was high, even at early, clinically uncertain stages, whereas it was suboptimal in discriminating PD and MSA. Although none of the patients with PSP and PD were falsely diagnosed as MSA (specificity 100%), diagnoses of MSA could only be established in 8.3% (sensitivity) in the CUP- cohort and 22.7% in the prospective cohort.

4. Discussion

Discrimination of atypical parkinsonism (APD) and PD is

important for a number of reasons, including differences in natural history, patient counselling and treatment response [4]. Further- more, interventional treatment trials for degenerative parkinsonian disorders require a correct diagnosis avoiding inclusion of patients with misdiagnoses, especially because there is an increase of disease-modifying trials with disease-specific candidate in- terventions for PD, MSA and PSP [4,15e17].

In the present study, we applied previously described plani- metric measurements, including Md, MA, PA, Pd, MA/PA, Md/Pd, Pd/ Md, PA/MA, SCPd, MCPd and MRPI, to evaluate and compare their discriminative potential in the differential diagnosis of degenera- tive parkinsonism in one of the largest cohorts of neurodegenera- tive parkinsonism studied so far. To maximize the diagnostic potential of different planimetric measurements, we took a novel approach by developing disease-specific decision tree algorithms for PSP, MSA and PD. We further validated the performance of the decision trees in two independent cohorts: a prospectively recruited cohort of patients with degenerative parkinsonism, and a population of early stage clinically unclassifiable parkinsonism. Overall, we were able to demonstrate that brainstem-derived MR planimetric measures yield high diagnostic accuracy for the discrimination of PSP when used along our decision algorithms, even at early, clinically uncertain stages. However, diagnostic ac- curacy in discriminating PD and MSA was suboptimal. This pro- cedure of assessing the performance of the prediction models with patient data not used in the development process is a further strength of our study, especially because they also included a cohort of early-stage parkinsonian patients.

Nevertheless, all but one predictive accuracy were higher for the disease-specific decision trees compared to the ROC-derived diag- nostic accuracy values in the training and both test sets supporting the use of MRI-based decision trees. Only for the detection of MSA, the MCPd and MRPI were slightly better than the disease-specific decision tree in the prospective cohort. While the MCPd has been shown to separate optimally between MSA and PD in a previous study [18], the decision algorithm for the detection of MSA involves both the MCPd and the Pd/Md. The latter is decreased in PSP and therefore helps to exclude these patients. Because predictive ac- curacy values in detecting MSA were similar between the applica- tion of the MCPd and the C4.5 algorithm in both test sets, one could favor the use of MCPd for the detection of MSA. The decision tree for PSP involves structures typically reported useful for the diagnosis of PSP, namely Md and Pd/Md and that for the diagnosis of PD includes parameters pathognomonic for APD, i.e. measures for the detection of PSP (Md and Pd/Md) and further of MSA (PA and MCPd), respec- tively. Area- and diameter-derived parameters, by contrast to ratio parameters, are subject to individual variations and normalization to intracranial volumes could likely cancel out. Nevertheless, some of the area- and diameter-derived parameters, namely MCPd and Md as well as MA and PA showed either high AUCs similar to pre- vious studies [3,5,8] or were included in our decision tree models, although not normalized to intracranial volumes similar to all previous studies using these planimetric parameters [3,5,8].

Overall diagnostic accuracy for PSP versus non-PSP as deter- mined with the AUC was 0.90 for the CUP cohort and 0.95 for the prospective cohort, while overall diagnostic accuracy for PD and MSA based on the decision algorithms was less accurate than for PSP, suggesting that brainstem-derived MR planimetry is primarily useful for a differential diagnosis of PSP. Of interest, Pd/Md was identified as the most important parameter to construct the pre- diction models, being involved in all three decision trees. Diameter- derived measurements were initially developed in histologically proven disease and then replicated in clinically diagnosed patients with degenerative parkinsonism, showing high specificity and sensitivity for PSP, and performing better than qualitative visual

S. Mangesius et al. / Parkinsonism and Related Disorders 46 (2018) 47e55 51

S. Mangesius et al. / Parkinsonism and Related Disorders 46 (2018) 47e55

Table 2

ROC-derived AUCs with diagnostic accuracy* of the different MR planimetric measurements and diagnostic accuracy* of the decision algorithms for the classification of degenerative parkinsonism in the training set.

PSP vs. non-PSP MSA vs. non-MSA

ROC-derived AUCsa of the different MR measures with cut-off values and diagnostic accuracya

PD vs. non-PD

AUC (95% CI)a 0.78 (0.73e0.83)

0.62 (0.56e0.67) 0.77 (0.72e0.82)

0.73 (0.68e0.78) 0.79 (0.74e0.83)c > 8.8
76.7 (56.6e92.4) 68.7 (49.6e86.4) 73.2 (66.7e78.8) 0.75 (0.70e0.80) 0.63 (0.57e0.70) 0.63 (0.57e0.70)

0.59 (0.52e0.65) 0.59 (0.52e0.65)

0.54 (0.48e0.61)

92.8 (88.2e96.0)/ 76.3 (67.6e83.6)/ 86.5 (82.2e90.1)

MR Measure Midbrain diameter

cut-off-value Diagnostic accuracy

Pontine diameter Midbrain area cut-off-value

Diagnostic accuracy

Pontine area Mean MCPd

cut-off-value Diagnostic accuracy

Mean SCPd Pd/Md Md/Pd

cut-off-value Diagnostic accuracy

PA/MA MA/PA

cut-off-value Diagnostic accuracy

MRPI

cut-off-value Diagnostic accuracy

Diagnostic accuracy of the decision algorithms

Diagnostic accuracy

AUC (95% CI)a 0.94 (0.91e0.98) <8.9 mm
90.0 (81.4e97.1)/ 90.2 (84.7e94.8)/ 90.0 (85.9e93.7) 0.55 (0.48e0.62) 0.90 (0.86e0.94) < 98.1 mm2
90.0 (80.0e97.1)/ 81.6 (72.7e88.4)/ 82.8 (76.3e88.1) 0.60 (0.53e0.67) 0.61 (0.54e0.67)

0.74 (0.68e0.80) 0.93 (0.89e0.97) 0.93 (0.89e0.97) < 0.54

88.6 (80.0e95.7)/ 89.0 (84.0e95.4)/ 89.1 (84.9e93.9) 0.89 (0.84e0.94) 0.89 (0.84e0.94) < 0.18

81.4 (70.0e91.4)/ 87.1 (82.8e95.4)/ 86.4 (82.3e92.2) 0.91 (0.87e0.95) > 15.62

82.9 (72.9e92.9)/ 86.2 (76.4e92.3)/ 85.9 (78.5e90.2)

81.8 (69.1e90.9)/ 98.4 (96.1e99.6)/ 95.5 (92.6e97.5)

AUC (95% CI)a 0.50 (0.44e0.57)

0.73 (0.66e0.80) 0.52 (0.46e0.58)

0.75 (0.69e0.82) 0.84 (0.78e0.89) < 8.0 mm
69.3 (56.0e85.3)/ 88.8 (70.7e94.0)/ 84.6 (72.7e89.1) 0.65 (0.58e0.73) 0.71 (0.65e0.78) 0.71 (0.65e0.78)

0.75 (0.69e0.81) 0.75 (0.69e0.81)

0.83 (0.78e0.88) < 11.14
70.7 (60.0e89.3)/ 84.1 (62.9e88.8)/ 81.3 (66.9e85.4)

54.0 (40.9e66.6)/ 100 (98.5e100)b/ 90.7 (86.9e93.7)

* sensitivity/specificity/predictive accuracy in % (95%CI).
Abbreviations: PSP 1⁄4 progressive supranuclear palsy; MSA 1⁄4 multiple system atrophy; PD 1⁄4 Parkinson disease; CUP 1⁄4 clinically uncertain parkinsonian syndromes; MR 1⁄4 magnetic resonance (imaging); SCPd 1⁄4 diameter of the superior cerebellar peduncle; MCPd 1⁄4 diameter of the middle cerebellar peduncle; Md/Pd 1⁄4 midbrain-to- pontine-diameter-ratio; Pd/Md 1⁄4 pons-to-midbrain-diameter-ratio; MA/PA-ratio 1⁄4 midbrain-to-pons-area-ratio; PA/MA-ratio 1⁄4 pons-to-midbrain-area-ratio; MRPI 1⁄4 Magnetic Resonance Parkinsonism Index; AUC 1⁄4 area under the curve; ROC 1⁄4 receiver operating characteristic curve; CI 1⁄4 confidence interval.

a AUC > 0.800 are highlighted in bold. For these AUCs, optimal cut-off values determined by the Youden Index and their diagnostic accuracy values (sensitivity, specificity and predictive accuracy) to separate groups are given.

. b Denominator is 0; no false positives (95% CI of specificity was calculated with https://www.medcalc.org/calc/diagnostic_test.php).

. c Diagnostic accuracy values are given, because this MR measure has the highest AUC to separate PD vs. non-PD parkinsonism.

assessment [8]. It is interesting to note, that while the MRPI yielded a high AUC in the ROC-analyses for classification of PSP versus non- PSP (0.91), this measure is not relevant for the decision algorithm. Diameter-derived measurements not only performed better in this study compared to area-derived measurements, but are also faster to perform.

Diagnosis of degenerative parkinsonism can be especially chal- lenging in early disease stages when symptoms have not fully emerged or patients present with atypical signs, but evidence of the use of MR-based neuroimaging studies in early disease stages of these patients is scarce [3e5]. By assessing a group of operationally defined subjects with CUP in early disease stages, we could show high sensitivity and specificity, as well as high overall accuracy of our calculated decision tree in predicting PSP also in these patients, but discrimination of MSA and PD was less satisfactory. To our knowledge, there is only one previous study assessing the useful- ness of MR planimetric measurements in predicting the clinical

evolution of CUP to defined clinical phenotypes [19]. In that study an abnormal MRPI had a diagnostic accuracy of 92.9% in predicting the evolution of CUP toward PSP phenotypes [19]. However, pa- tients with CUP in our study differed considerably from those in the previous study, who were more advanced and presented with atypical signs (falls in the first year, slowness of vertical saccades, freezing in the first 3 years of disease) [19], already suggesting high pre-test probability of a diagnosis of APD.

For application of the established decision trees in clinical routine, automated MR planimetric measurements would be practical. A recent study presented an automated method for MRPI calculation which was highly accurate in distinguishing PSP from PD, and showed good agreement with manual values [20]. Such an automatic approach would allow a widespread use of planimetric measurements in clinical practice and in longitudinal research studies. Since our decision trees showed the diagnostic potential of diameter-derived measurements developing an automatic method

Table 3

Diagnostic accuracy* of the different MR planimetric measurements using the ROC- derived cut-off values of the training set and diagnostic accuracy* of the decision algorithms for the classification of degenerative parkinsonism in the validation sets.

for calculating midbrain and pontine diameter seems a worthwhile effort [20].

A prediction model showing acceptable or good performance based on internal validation in the development data set, will not necessarily behave similarly in a different group of individuals [21,22]. This might explain why initial studies introducing brainstem-derived planimetric measures had very high diagnostic accuracy in separating PSP from non-PSP parkinsonism [6,7] or MSA from PD [18], while newer studies differentiating PSP from non-PSP patients did not [23,24]. In the present study, we have therefore assessed the performance of the prediction models with patient data not used in the development process and were able to obtain further support for the diagnostic performance of the model [22].

In general, the improved diagnostic accuracy of our decision tree algorithm may have significant implications for both, future research studies and patient counselling, in PSP and, to a lesser degree, for PD and MSA. Due to the higher sensitivity of our approach, the number of false negative findings is reduced and, thus, more patients are amenable to future research studies including interventional trials. On the other hand, the high speci- ficity e reflecting a high number of true negative and a low number of false positive cases e is imperative when considering a test for patient counselling. In fact, PSP and MSA are fatal disorders and clinicians require a diagnostic test to yield high specificity to avoid misdiagnosis and resulting impacts on the scheme of life of in- dividuals suffering from parkinsonism. Despite the improved diagnostic accuracy of the decision tree algorithm presented in this paper, there is still room for improvement. Previous diffusion- imaging studies (reflecting microstructural damage) as well as iron-sensitive MRI approaches were shown to be helpful in discriminating patients with APD from PD patients [3,5]. Because different MR sequences provide unique kinds of information on tissue changes, approaches using multimodal MR imaging to assess complementary tissue characteristics may be the most promising approach to improve diagnostic accuracy in the differential diag- nosis of neurodegenerative parkinsonian disorders [4,5,25], particularly early in the disease course. This is especially relevant for the diagnostic discrimination of MSA and PD, where diagnostic accuracy of brainstem-derived planimetric measures remains suboptimal and multimodal MR imaging has the potential to further improve the diagnostic accuracy. Moreover, because PSP is known to have an underneath tauopathy [26], MR methods to detect white matter microstructural damage such as diffusion im- aging methods might likely better lead to detect specific differences between PSP and non-PSP parkinsonism [27]. Moreover, novel tau markers detected in the cerebrospinal fluid (CSF) or by PET imaging will likely become available in the near future for this reason [17,28].

A few limitations of our study warrant further discussion. MRI images in our retrospective cohort were obtained on different 1.5 T Siemens Scanners, which potentially might be a source for a methodological bias and increased variability. There is, however, evidence, that different scanner do not influence brainstem- derived planimetric measures [13,24]. Indeed, we have shown that different scanners produce similar results of quantifiable, infratentorial changes using MRI planimetry, and that these changes on 1.5 T MRI can be reproduced at 3 T MRI planimetry concluding that there was no methodological bias derived from the use of different scanners in the retrospective cohort in the present study [13,29]. In addition, misdiagnosis in some of the clinically diagnosed patients cannot be excluded in the absence of post- mortem verification, particularly in those in early disease stages [4]. However, previous studies suggested that diagnostic certainty in- creases with disease progression [3,4] and, hence, one of the

second test setd
Diagnostic accuracy of the decision algorithms

first test setc second test setd

82.4 (56.6e96.2) 97.0 (89.6e99.6) 94.1 (86.7e98.0) 91.3 (72.0e98.9) 98.4 (91.3e100) 96.5 (90.0e99.3)

8.3 (0.2e38.5) 100 (95.1e100)b 86.9 (77.8e93.3) 22.7 (7.8e45.4) 100 (94.3e100)b 80.0 (69.9e87.9)

83.6 (71.2e92.2) 55.2 (35.7e73.6) 73.8 (63.1e82.8) 100 (91.2e100)b 75.6 (60.5e87.1) 87.1 (78.0e93.4)

S. Mangesius et al. / Parkinsonism and Related Disorders 46 (2018) 47e55 53

Cut-off valuesa and MR Measure Midbrain diameter Diagnostic accuracy

first test setc

second test setd

Midbrain area

Diagnostic accuracy first test setc

second test setd

Mean MCPd Diagnostic accuracy

first test setc

second test setd

Md/Pd
Diagnostic accuracy

first test setc

second test setd

MA/PA
Diagnostic accuracy

first test setc

second test setd

MRPI

Diagnostic accuracy first test setc

PSP vs. non-PSP MSA vs. non-MSA PD vs. non-PD

diagnostic accuracy of the different MR measures

cut-off-valuea <8.9 mm

88.2 (63.6e98.5) 86.6 (76.0e93.7) 86.9 (77.8e93.3) 95.7 (78.1e99.9) 83.9 (72.3e92.0) 87.1 (78.0e93.4) <98.1 mm2

94.1 (71.3e99.9) 82.1 (70.8e90.4) 84.5 (75.0e91.5) 91.3 (72.0e98.9) 83.9 (72.3e92.0) 85.9 (76.6e92.5)

<0.54

94.1 (71.3e99.9) 80.6 (69.1e89.2) 83.3 (73.6e90.6) 100 (85.2e100)b 77.4 (65.0e87.1) 83.5 (73.9e90.7) <0.18

76.5 (50.1e93.2) 86.6 (76.0e93.7) 84.5 (75.0e91.5) 95.7 (78.1e99.9) 88.7 (78.1e95.3) 90.6 (82.3e95.9) >15.62

88.2 (63.6e98.5) 85.1 (74.3e92.6) 85.7 (76.4e92.4) 100.0 (85.2e100)b 87.1 (76.2e94.3) 90.6 (82.3e95.9)

cut-off-valuea

<8.0 mm

25.0 (5.5e57.2) 86.1 (75.9e93.1) 77.4 (67.0e85.8) 45.5 (24.4e67.8) 98.4 (91.5e100) 84.7 (75.3e91.6)

>8.8 mm

72.7 (59.0e83.9) 51.7 (32.5e70.6) 65.5 (54.3e75.5) 97.5 (86.8e99.9) 44.4 (29.6e60.0) 69.4 (58.5e79.0)

<11.14

58.3 (27.7e84.8) 15.3 (7.9e25.7) 21.4 (13.2e31.7) 63.6 (40.7e82.8) 95.2 (86.7e99.0) 87.1 (78.0e93.4)

* sensitivity/specificity/predictive accuracy in % (95%CI).
Abbreviations: PSP 1⁄4 progressive supranuclear palsy; MSA 1⁄4 multiple system at- rophy; PD 1⁄4 Parkinson disease; CUP 1⁄4 clinically uncertain parkinsonian syndromes; MR 1⁄4 magnetic resonance (imaging); SCPd 1⁄4 diameter of the superior cerebellar peduncle; MCPd 1⁄4 diameter of the middle cerebellar peduncle; Md/Pd 1⁄4 midbrain- to-pontine-diameter-ratio; Pd/Md 1⁄4 pons-to-midbrain-diameter-ratio; MA/PA- ratio 1⁄4 midbrain-to-pons-area-ratio; PA/MA-ratio 1⁄4 pons-to-midbrain-area-ratio; MRPI 1⁄4 Magnetic Resonance Parkinsonism Index; AUC 1⁄4 area under the curve; ROC 1⁄4 receiver operating characteristic curve; CI 1⁄4 confidence interval.

a ROC-derived cut-off values of the training set.

b denominator is 0; no false positives (95% CI of specificity was calculated with https://www.medcalc.org/calc/diagnostic_test.php).

c CUP-cohort.
d Prospective cohort.

54 S. Mangesius et al. / Parkinsonism and Related Disorders 46 (2018) 47e55

patient selection criteria delineated in our study plan was a clinical follow-up of at least 2 years [6e9]. Optimally, novel diagnostic markers established in clinical cohorts similar to our study should be validated in neuropathologically confirmed cohorts, which is, however, not practical in life. Therefore, the development and validation of highly sensitive and specific imaging and CSF or even plasma biomarker is a high priority objective in the research of degenerative parkinsonism.

In conclusion, we could show that brainstem-derived MR planimetric measures yield high diagnostic accuracy for the discrimination of PSP when used along our decision algorithms, even at early, clinically uncertain stages, while diagnostic accuracy in discriminating PD and MSA was suboptimal. The combination of brainstem-derived MR planimetry with other MR techniques may represent a promising approach to improve diagnostic accuracy in discriminating especially MSA and PD, and should be explored in future studies.

Authors’ contributions

1. Research Project: A. Conception and Design, B. Organization, C. Acquisition of Data; 2. Statistical Analysis: A. Design, B. Execu- tion, C. Review/Critique/Interpretation; 3. Manuscript Preparation: A. Drafting the Article, B. Critical Review for Important Intellectual Content; 4. Final Approval of the submitted version.

S.M.: 1A, 1B, 1C, 2A, 2B, 2C, 3A, 3C, 4. A.H.: 1A, 1C, 2C, 3B, 4.
F.K.: 1A, 2A, 2C, 3B, 4.
P.M.: 1A, 2A, 3B, 4.

E.R.: 1C, 3B, 4.
S.T.: 1C, 3B, 4.
A.D.: 2A, 3B, 4.
M.S.: 1A, 1C, 3B, 4.
R.E.: 1C, 3B, 4.
G.W.:1A, 1C, 3B, 4.
C.M.: 1C, 3B, 4.
E.G.: 1A, 1B, 3B, 4.
W.P.: 1A, 1B, 2C, 3B, 4.
K.S.: 1A, 1B, 1C, 2A, 2B, 2C, 3B, 4.

Funding

Financial disclosure/conflict of interest concerning the research related to the manuscript

None of the authors has conflicts of interest to disclose, which may be relevant for this work.

Funding sources for study

The project was supported by funds of the Oesterreichische Nationalbank (Austrian Central Bank, Anniversary Fund, project number: 14174), the Austrian Science Fund (FWF: Der Wissen- schaftsfonds, projet number: KLI82-B00), the MSA Coalition (MSA Coalition 2015-#15), the Integrated Research and Therapy Center (IFTZ: Integrierte Forschungs-und Therapiezentrum, project num- ber: 2007152), and the medical research grant Innsbruck (MFI: Medizinische Forschungsfo€rderung Innsbruck, project number: 6169).

These funding sources had no involvement in study design, collection, analysis and interpretation of data, writing of the report, or in the decision to submit the article for publication.

Financial disclosures of all authors (for the preceding 12 months)

Stephanie Mangesius reports a grant from the Austrian Parkin- son’s Disease Society.

Anna Hussl, Eva Reiter, Susanne Tagwercher, Regina Ester- hammer, Christoph Müller and Christoph Scherfler have nothing to report.

Florian Krismer reports grants from MSA Coalition, the Inter- national Parkinson’s Disease and Movement Disorder Society, and the Austrian Parkinson’s Disease Society, and non-financial support from Fight MSA and Ipsen Pharma.

Philipp Mahlknecht was supported by grants from the Austrian Parkinson’s Disease Society.

Gregor Wenning reports consulting fees and lecture fees from AstraZeneca and lecture fees from Abbvie.

Atbin Djamshidian received travel grants from Medtronic.

Michael Schocke is Co-founder of the academic spin-off com- pany Ergospect GmbH.

Elke R.Gizewski reports personal lecture fees from Bracco and Bayer.

Werner Poewe reports personal fees from AbbVie, Allergan, AstraZeneca, BIAL, Biogen, Boehringer-Ingelheim, Boston Scientific, GlaxonSmithKline, Ipsen, Lundbeck, Medtronic, MSD, Merck- Serono, Merz Pharmaceuticals, Novaris, Orion Pharma, Teva, UCB and Zambon (consultancy and lecture fees in relation to clinical drug development programmes für PD) and royalties from Thieme, Wiley Blackwell, Oxford University Press and Cambridge University press.

Klaus Seppi reports grants from Medical University Innsbruck, grants from Oesterreichische Nationalbank Nationalbank (Austrian Central Bank, Anniversary Fund; project no.: 14174), grants from Austrian Science Fund (FWF: Der Wissenschaftsfonds; project no.: KLI82-B00), grants from the Michael J. Fox Foundation, grants and personal fees from International Parkinson and Movement Disorder Society, personal fees from Teva, personal fees from UCB, personal fees from Lundbeck, personal fees from AOP Orphan, personal fees from Roche and personal fees from Abbvie outside the submitted work.

Acknowledgment

We thank all patients who volunteered to participate in our study.

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.020.

References

[1] D.R. Williams, A.J. Lees, What features improve the accuracy of the clinical diagnosis of progressive supranuclear palsy-parkinsonism (PSP-P)? Mov. Disord. 25 (3) (2010) 357e362.

[2] I. Litvan, K.P. Bhatia, D.J. Burn, C.G. Goetz, A.E. Lang, I. McKeith, N. Quinn, K.D. Sethi, C. Shults, G.K. Wenning, C. Movement Disorders Society Scientific Issues, Movement disorders society scientific issues committee report: SIC task force appraisal of clinical diagnostic criteria for parkinsonian disorders, Mov. Disord. 18 (5) (2003) 467e486.

[3] P. Mahlknecht, A. Hotter, A. Hussl, R. Esterhammer, M. Schocke, K. Seppi, Significance of MRI in diagnosis and differential diagnosis of Parkinson’s disease, Neurodegener. Dis. 7 (5) (2010) 300e318.

[4] W. Poewe, K. Seppi, C.M. Tanner, G.M. Halliday, P. Brundin, J. Volkmann, A.E. Schrag, A.E. Lang, Parkinson disease, Nat. Rev. Dis. Prim. 3 (2017) 17013. [5] B. Heim, F. Krismer, R. De Marzi, K. Seppi, Magnetic resonance imaging for the diagnosis of Parkinson’s disease, J. Neural Transm. (Vienna) 124 (8) (2017)

915e964.
[6] A. Quattrone, G. Nicoletti, D. Messina, F. Fera, F. Condino, P. Pugliese, P. Lanza,

P. Barone, L. Morgante, M. Zappia, U. Aguglia, O. Gallo, MR imaging index for

differentiation of progressive supranuclear palsy from Parkinson disease and the Parkinson variant of multiple system atrophy, Radiology 246 (1) (2008) 214e221.

. [7] H. Oba, A. Yagishita, H. Terada, A.J. Barkovich, K. Kutomi, T. Yamauchi, S. Furui, T. Shimizu, M. Uchigata, K. Matsumura, M. Sonoo, M. Sakai, K. Takada, A. Harasawa, K. Takeshita, H. Kohtake, H. Tanaka, S. Suzuki, New and reliable MRI diagnosis for progressive supranuclear palsy, Neurology 64 (12) (2005) 2050e2055.

. [8] L.A. Massey, H.R. Jager, D.C. Paviour, S.S. O’Sullivan, H. Ling, D.R. Williams, C. Kallis, J. Holton, T. Revesz, D.J. Burn, T. Yousry, A.J. Lees, N.C. Fox, C. Micallef, The midbrain to pons ratio: a simple and specific MRI sign of progressive supranuclear palsy, Neurology 80 (20) (2013) 1856e1861.

. [9] A. Hussl, P. Mahlknecht, C. Scherfler, R. Esterhammer, M. Schocke, W. Poewe, K. Seppi, Diagnostic accuracy of the magnetic resonance Parkinsonism index and the midbrain-to-pontine area ratio to differentiate progressive supra- nuclear palsy from Parkinson’s disease and the Parkinson variant of multiple system atrophy, Mov. Disord. 25 (14) (2010) 2444e2449.

. [10] P.M. Wadia, A.E. Lang, The many faces of corticobasal degeneration, Park. Relat. Disord. 13 (Suppl 3) (2007) S336eS340.

. [11] A.M. Catafau, E. Tolosa, DaTSCAN Clinically Uncertain Parkinsonian Syn- dromes Study Group, Impact of dopamine transporter SPECT using 123I- Ioflupane on diagnosis and management of patients with clinically uncertain Parkinsonian syndromes, Mov. Disord. 19 (10) (2004) 1175e1182.

. [12] W.R. Gibb, A.J. Lees, The relevance of the Lewy body to the pathogenesis of idiopathic Parkinson’s disease, J. Neurol. Neurosurg. Psychiatry 51 (6) (1988) 745e752.

. [13] S. Mangesius, F. Krismer, E.R. Gizewski, C. Muller, A. Hussl, M. Schocke, C. Scherfler, W. Poewe, K. Seppi, 1.5 versus 3 tesla magnetic resonance planimetry in neurodegenerative parkinsonism, Mov. Disord. 31 (12) (2016) 1925e1927.

. [14] C. Scherfler, G. Gobel, C. Muller, M. Nocker, G.K. Wenning, M. Schocke, W. Poewe, K. Seppi, Diagnostic potential of automated subcortical volume segmentation in atypical parkinsonism, Neurology 86 (13) (2016) 1242e1249.

. [15] F. Krismer, G.K. Wenning, Multiple system atrophy: insights into a rare and debilitating movement disorder, Nat. Rev. Neurol. 13 (4) (2017) 232e243.

. [16] M. Stamelou, K.P. Bhatia, Atypical parkinsonism – new advances, Curr. Opin. Neurol. 29 (4) (2016) 480e485.

. [17] A.L. Boxer, J.T. Yu, L.I. Golbe, I. Litvan, A.E. Lang, G.U. Hoglinger, Advances in progressive supranuclear palsy: new diagnostic criteria, biomarkers, and therapeutic approaches, Lancet Neurol. 16 (7) (2017) 552e563.

. [18] G. Nicoletti, F. Fera, F. Condino, W. Auteri, O. Gallo, P. Pugliese, G. Arabia, L. Morgante, P. Barone, M. Zappia, A. Quattrone, MR imaging of middle cerebellar peduncle width: differentiation of multiple system atrophy from Parkinson disease, Radiology 239 (3) (2006) 825e830.

. [19] M. Morelli, G. Arabia, F. Novellino, M. Salsone, L. Giofre, F. Condino, D. Messina, A. Quattrone, MRI measurements predict PSP in unclassifiable parkinsonisms:

a cohort study, Neurology 77 (11) (2011) 1042e1047.
[20] S. Nigro, G. Arabia, A. Antonini, L. Weis, A. Marcante, A. Tessitore, M. Cirillo,

G. Tedeschi, S. Zanigni, G. Calandra-Buonaura, C. Tonon, G. Pezzoli, R. Cilia, M. Zappia, A. Nicoletti, C.E. Cicero, M. Tinazzi, P. Tocco, N. Cardobi, A. Quattrone, Magnetic Resonance Parkinsonism Index: diagnostic accuracy of a fully automated algorithm in comparison with the manual measurement in a large Italian multicentre study in patients with progressive supranuclear palsy, Eur. Radiol. 27 (6) (2017) 2665e2675.

[21] D.G. Altman, Y. Vergouwe, P. Royston, K.G. Moons, Prognosis and prognostic research: validating a prognostic model, BMJ 338 (2009) b605.

[22] J.M. Hendriksen, G.J. Geersing, K.G. Moons, J.A. de Groot, Diagnostic and prognostic prediction models, J. Thromb. Haemost. 11 (Suppl 1) (2013) 129e141.

[23] S. Zanigni, G. Calandra-Buonaura, D.N. Manners, C. Testa, D. Gibertoni, S. Evangelisti, L. Sambati, M. Guarino, P. De Massis, L.L. Gramegna, C. Bianchini, P. Rucci, P. Cortelli, R. Lodi, C. Tonon, Accuracy of MR markers for differenti- ating progressive supranuclear palsy from Parkinson’s disease, Neuroimage Clin. 11 (2016) 736e742.

[24] L. Moller, J. Kassubek, M. Sudmeyer, R. Hilker, E. Hattingen, K. Egger, F. Amtage, E.H. Pinkhardt, G. Respondek, M. Stamelou, F. Moller, A. Schnitzler, W.H. Oertel, S. Knake, H.J. Huppertz, G.U. Hoglinger, Manual MRI morphom- etry in Parkinsonian syndromes, Mov. Disord. 32 (5) (2017) 778e782.

[25] G. Barbagallo, M. Sierra-Pena, F. Nemmi, A.P. Traon, W.G. Meissner, O. Rascol, P. Peran, Multimodal MRI assessment of nigro-striatal pathway in multiple system atrophy and Parkinson disease, Mov. Disord. 31 (3) (2016) 325e334.

[26] G.U. Hoglinger, G. Respondek, M. Stamelou, C. Kurz, K.A. Josephs, A.E. Lang, B. Mollenhauer, U. Muller, C. Nilsson, J.L. Whitwell, T. Arzberger, E. Englund, E. Gelpi, A. Giese, D.J. Irwin, W.G. Meissner, A. Pantelyat, A. Rajput, J.C. van Swieten, C. Troakes, A. Antonini, K.P. Bhatia, Y. Bordelon, Y. Compta, J.C. Corvol, C. Colosimo, D.W. Dickson, R. Dodel, L. Ferguson, M. Grossman, J. Kassubek, F. Krismer, J. Levin, S. Lorenzl, H.R. Morris, P. Nestor, W.H. Oertel, W. Poewe, G. Rabinovici, J.B. Rowe, G.D. Schellenberg, K. Seppi, T. van Eime- ren, G.K. Wenning, A.L. Boxer, L.I. Golbe, I. Litvan, P.S.P.S.G. Movement Dis- order Society-endorsed, Clinical diagnosis of progressive supranuclear palsy: the movement disorder society criteria, Mov. Disord. 32 (6) (2017) 853e864.

[27] A. Tessitore, A. Giordano, G. Caiazzo, D. Corbo, R. De Micco, A. Russo, S. Liguori, M. Cirillo, F. Esposito, G. Tedeschi, Clinical correlations of microstructural changes in progressive supranuclear palsy, Neurobiol. Aging 35 (10) (2014) 2404e2410.

[28] D.J. Irwin, Tauopathies as clinicopathological entities, Park. Relat. Disord. 22 (Suppl 1) (2016) S29eS33.

[29] L. Moller, J. Kassubek, M. Sudmeyer, R. Hilker, E. Hattingen, K. Egger, F. Amtage, E.H. Pinkhardt, G. Respondek, M. Stamelou, F. Moller, A. Schnitzler, W.H. Oertel, S. Knake, H.J. Huppertz, G.U. Hoglinger, Manual MRI morphom- etry in Parkinsonian syndromes, Mov. Disord. 32 (5) (2017) 778e782.

S. Mangesius et al. / Parkinsonism and Related Disorders 46 (2018) 47e55 55

Parkinsonism and Related Disorders 46 (2018) 56e61

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Grey matter volume loss is associated with specific clinical motor signs in Huntington’s disease

Emma M. Coppen a, *, 1, Milou Jacobs a, 1, Annette A. van den Berg-Huysmans b, Jeroen van der Grond b, Raymund A.C. Roos a

a Department of Neurology, Leiden University Medical Center, PO BOX 9600, 2300 RC Leiden, The Netherlands b Department of Radiology, Leiden University Medical Center, PO BOX 9600, 2300 RC Leiden, The Netherlands

articleinfo

Article history:

Received 22 May 2017 Received in revised form 12 September 2017 Accepted 1 November 2017

Keywords:

Huntington’s disease Structural MRI Chorea
Motor cortex

Basal ganglia

abstract

Background: Motor disturbances are clinical hallmarks of Huntington’s disease (HD) and involve chorea, dystonia, hypokinesia and visuomotor dysfunction. Investigating the association between specific motor signs and different regional volumes is important to understand the heterogeneity of HD.
Objective: To investigate the motor phenotype of HD and associations with subcortical and cortical grey matter volume loss.

Methods: Structural T1-weighted MRI scans of 79 HD patients and 30 healthy controls were used to calculate volumes of seven subcortical structures including the nucleus accumbens, hippocampus, thalamus, caudate nucleus, putamen, pallidum and amygdala. Multiple linear regression analyses, cor- rected for age, gender, CAG, MRI scan protocol and normalized brain volume, were performed to assess the relationship between subcortical volumes and different motor subdomains (i.e. eye movements, chorea, dystonia, hypokinesia/rigidity and gait/balance). Voxel-based morphometry analysis was used to investigate the relationship between cortical volume changes and motor signs.

Results: Subcortical volume loss of the accumbens nucleus, caudate nucleus, putamen, and pallidum were associated with higher chorea scores. No other subcortical region was significantly associated with motor symptoms after correction for multiple comparisons. Voxel-based cortical grey matter volume reductions in occipital regions were related with an increase in eye movement scores.

Conclusion: In HD, chorea is mainly associated with subcortical volume loss, while eye movements are more related to cortical volume loss. Both subcortical and cortical degeneration has an impact on motor impairment in HD. This implies that there is a widespread contribution of different brain regions resulting in the clinical motor presentation seen in HD patients.

1. Introduction

Huntington’s disease (HD) is an autosomal-dominant, neuro- degenerative disorder characterized by progressive motor distur- bances, cognitive impairment and psychiatric symptoms. The clinical diagnosis of HD is based on the presence of motor signs, and can involve chorea, dystonia and/or hypokinesia [1]. Oculomotor dysfunction, such as saccadic eye movements or gaze paralysis, can

* Corresponding author. Department of Neurology (J3-R-162), Leiden University Medical Center, PO Box 9600, 2300 RC Leiden, The Netherlands.

E-mail addresses: e.m.coppen@lumc.nl (E.M. Coppen), m.jacobs@lumc.nl (M. Jacobs), a.a.van_den_berg-huysmans@lumc.nl (A.A. van den Berg-Huysmans), j.van_der_grond@lumc.nl (J. van der Grond), r.a.c.roos@lumc.nl (R.A.C. Roos).

1 These authors contributed equally to the manuscript.

https://doi.org/10.1016/j.parkreldis.2017.11.001

also be prominent in premanifest and early HD [2]. The clinical HD phenotype is heterogeneous and different motor signs can also co- exist [3]. Longitudinal analysis of motor signs showed that chor- eatic movements decrease over time, whereas hypokinetic-rigid signs slightly increase [4]. This suggests that different motor symptoms can be more pronounced during different disease stages. The Unified HD Rating Scale Total Motor Score (UHDRS-TMS) [5] is the gold standard to evaluate motor functioning in HD and estab- lish the clinical diagnosis. Here, several motor domains including, chorea, dystonia, gait, rigidity, and eye movements are examined, with higher total scores indicating more motor dysfunction.

Although striatal atrophy is the main neuropathological finding in HD, neuronal loss has been identified in many other extrastriatal brain regions [6]. In these regions, it has been shown that grey matter volume reductions may also be associated with decreased

global motor and functional scores [7e11].
Instead of focusing on global motor functioning, we aimed to

investigate associations between separate motor domains and grey matter volume changes. To monitor HD signs in clinical practice and intervention trials, it is important to further understand the pathophysiology underlying the HD phenotype, because this can vary among patients.

2. Methods

2.1. Participants

A total of 79 patients with manifest HD and 30 healthy controls who visited the outpatient clinic at the department of Neurology of the Leiden University Medical Center (LUMC) between January 2008 and June 2016 were included. All manifest HD had a geneti- cally confirmed CAG repeat length of !39 and an UHDRS-TMS of more than 5, confirming the diagnosis and clinical motor presence of HD. The local ethical committee approved this study and written informed consent was obtained from all participants.

Distinctive items of the UHDRS motor scale were added for each participant to establish total scores per motor subdomain based on previous studies [4,12,13], representing five domains of motor functioning. For further details, see supplementary Table S1.

2.2. MRI image acquisition

All participants underwent MRI scanning on a 3 T MRI scanner (Philips Achieva, Best, the Netherlands). For each participant, a structural three-dimensional T1-weighted image was acquired. Imaging parameters of the scan protocols were: TR 1⁄4 7.7 ms, TE 1⁄4 3.5 ms, flip angle 1⁄4 8 , FOV 24 cm, matrix size 224 224 cm and 164 sagittal slices to cover the entire brain with a slice thick- ness of 1.0 mm with no gap between slices. This resulted in a voxel size of 1,07 mm x 1,07 mm x 1,0 mm.

2.3. Image post-processing

Functional Magnetic Resonance Imaging of the Brain (FMRIB) Software Library (FSL, version 5.0.8, Oxford, United Kingdom) was used for data analysis of all structural T1-weighted images [14]. Brain tissue volume, normalized for individual head size, was estimated with SIENAX [15]. Using SIENAX, brain and skull images were extracted from the single whole-head input data. Then, the brain image is affine-registered to Montreal Neurological Institute (MNI) 152-space standard image [16], using the skull image to determine the registration scaling. This volumetric scaling factor was used to normalize for head size. Next, tissue-type segmentation with partial volume estimation was performed in order to calculate the total vol- ume of normalized brain tissue, including separate estimates of vol- umes of grey matter, white matter, peripheral grey matter and ventricular CSF for each HD patient. Visual inspection of the registra- tion and segmentation was performed for each brain-extracted image.

2.4. Subcortical volumes

Absolute volumes of seven subcortical structures (i.e. nucleus accumbens, hippocampus, thalamus, caudate nucleus, putamen, pallidum and amygdala) were measured using FMRIB’s integrated registration and segmentation tool (FIRST) [17]. Here, all non-brain tissue was removed from the T1-weighted images using a semi- automated brain extraction tool that is implemented in FSL [18]. After registration of the images to the MNI 152-standard space image, using linear registration with 12 of freedom, segmentation of the subcortical regions was carried out using mesh models that

were constructed from manually segmented images provided by the Center for Morphometric Analysis (CMA), Massachusetts Gen- eral Hospital, Boston. Then, the volume for each structure was separately estimated. Visual inspection was performed for each output image during the registration and segmentation steps.

2.5. Voxel-based morphometry

To investigate voxel-wise differences in grey matter volume between HD patients and controls, voxel-based morphometry (VBM) analysis was performed as implemented in FSL [19].

First, brain extracted T1-weighted images were segmented into different tissue types (i.e. grey matter, white matter or cerebro- spinal fluid). Each segmented image has values that indicate the probability of a given tissue type. Then, the grey matter images were aligned to the 2 mm MNI-152 standard space image using non-linear registration. The resulting images were averaged to create a study-specific grey matter template. Subsequently, all native grey matter images were non-linearly registered to this study-specific template and ‘modulated’ to correct for local en- largements and contractions due to the non-linear component of the spatial transformation [20]. The modulated grey matter images were finally smoothed with an isotropic Gaussian kernel with a sigma of 3 mm and analyzed using a general linear model in FSL for statistical inference.

Brain structures that showed a significant difference between groups were identified using the Harvard-Oxford atlas integrated in FSL.

2.6. Statistical analyses

Group differences between HD patients and controls were analyzed using parametric (independent sample t-test) and non- parametric tests (c2-test) when applicable. To analyze group dif- ferences in the VBM output, a general linear model was constructed

Table 1

Clinical and volumetric group differences between HD patients and controls.

E.M. Coppen et al. / Parkinsonism and Related Disorders 46 (2018) 56e61 57

Clinical characteristics

Age
Gender m/f (%m)
CAG
Disease duration
Disease burden UHDRS-TMS
UHDRS chorea
UHDRS hypokinetic-rigid UHDRS dystonia
UHDRS eye movements UHDRS gait/balance

Subcortical structures

Accumbens nucleus Caudate nucleus Amygdala
Putamen

Pallidum Thalamus Hippocampus

HD (n 1⁄4 79)

46.5 (9.7; 28e65) 30/49 (38.0%)
44.1 (2.4; 40e51)
3.3 (3.0; 0e13)
382.1 (77.8; 234e551) 17.8 (10.8; 6e45)

5.2 (4.8; 0e18) 4.6 (3.2; 0e12) 0.2 (0.6; 0e3) 4.9 (3.2; 0e13) 1.8 (1.4; 0e6)

732.0 (188.0) 4942.2 (997.5) 2208.0 (528.5) 7093.0 (1229.1) 2749.8 (555.8) 13958.0 (1551.3) 7195.4 (1016.0)

Controls (n 1⁄4 30)

48.9 (8.4; 35e65) 14/16 (46.7%)
NA
NA

NA
2.6 (2.4; 0e7) NA
NA
NA
NA
NA

930.5 (207.0) 6695.4 (839.0) 2163.4 (379.4) 9280.0 (1289.7) 3338.5 (471.4) 14844.2 (1383.7) 7682.1 (818.3)

p-value

0.229 0.409 NA
NA
NA <0.001 NA

NA NA NA NA

<0.001 <0.001 0.673 <0.001 <0.001 <0.005 0.021

Data are mean (SD; range) or number (%) for gender. Volumes of subcortical structures are expressed in mm3. Mean disease duration is based on a smaller sample size (n 1⁄4 65) due to missing data. Independent sample t-test was used to compare groups, except for gender (c2-test). Statistically significant p-values are highlighted in bold (p < 0.05). NA 1⁄4 Not applicable; CAG 1⁄4 Cytosine-Adenine- Guanine; HD 1⁄4 Huntington’s Disease; UHDRS 1⁄4 Unified Huntington’s Disease Rat- ing Scale; TMS 1⁄4 Total Motor Score.

58 E.M. Coppen et al. / Parkinsonism and Related Disorders 46 (2018) 56e61

in FSL to compare controls with manifest HD using two-tailed t- statistics with age, gender, normalized brain volume and MRI scan protocol as covariates. Voxel-wise non-parametric permutation testing with 5000 permutations was performed using FSL ran- domise [21]. The Threshold-Free Cluster Enhancement (TFCE) technique was used to correct for multiple comparisons with family wise error [22], with a p-value < 0.05 as significant threshold. The regions that showed significant differences between HD patients and controls were selected for further analyses in the HD group only.

The following analyses, investigating the relationship between separate motor subdomains, subcortical and cortical brain volumes, were performed in HD patients only. Multiple linear regression analyses were used to investigate the relationship between the separate motor subdomains and subcortical brain volumes. Ana- lyses were accounted for age, gender, CAG repeat length, normal- ized brain volume, and MRI scan protocol. To correct for multiple comparisons the p-value for statistical significance was set at p < 0.008 (0.05/6) for analyses of subcortical volumes. To assess the relationship between clinical motor scores and cortical grey matter changes in HD patients, a general linear model was constructed using a design matrix in FSL with each clinical motor domain separately, correcting for age, gender, CAG repeat length,

Table 2

Relationship between UHDRS motor subdomains and subcortical brain volumes.

normalized brain volume, and MRI scan protocol.
FSL-Randomise was used for voxel-wise non-permutation

testing [21], using the regions that showed significant grey matter changes between controls and HD patients as a grey matter mask. Again, the TFCE technique was used to correct for multiple com- parisons with family wise error [22], with a p-value < 0.05 as sig- nificant threshold. Statistical analyses were performed using IBM SPSS 23.0 for Windows.

3. Results

Group characteristics and comparisons between HD patients and controls are reported in Table 1. There were no significant differences in age and gender between both groups. HD patients had a signifi- cantly higher mean UHDRS-TMS compared to the control group.

3.1. Subcortical volumes

The mean volumes of the accumbens nucleus, caudate nucleus, putamen, pallidum, thalamus and hippocampus were significantly lower in manifest HD compared to controls (Table 1). Since the mean volume of the amygdala did not differ between HD patients

UHDRS-TMS
UHDRS chorea
UHDRS hypokinetic-rigid UHDRS dystonia
UHDRS eye movements UHDRS gait/balance

Accumbens nucleus

¡0.283 ¡0.260 0.180 0.075 0.188 0.097

Caudate nucleus

¡0.316 ¡0.346 0.118 0.033 0.212 0.056

Putamen Pallidum

¡0.279 ¡0.312 ¡0.275 ¡0.273 0.175 0.156 0.012 0.076 0.171 0.239 0.112 0.214

Thalamus Hippocampus

0.096 0.265 0.045 0.172 0.052 0.179 0.056 0.047 0.169 0.240 0.057 0.245

Reported data are standardized coefficients (standardized beta) from the multiple linear regression analysis. Analyses were accounted for age, gender, CAG, MRI scan protocol, and normalized brain volume. Statistically significant values are printed in bold (corrected for multiple comparisons, p < 0.008). UHDRS 1⁄4 Unified Huntington’s Disease Rating Scale; TMS 1⁄4 Total Motor Score.

Fig. 1. Voxel based morphometry analysis between manifest HD and controls.
Brain regions that showed significant differences in grey matter volume in manifest HD compared to controls by means of voxel-based morphometry (VBM) are presented. Age, gender, MRI study protocol and normalized brain volume were included as covariates in the statistical model. Identified grey matter regions are overlaid on sagittal, transversal and coronal slices of Montreal Neurological Institute (MNI)-152 standard space T1-weighted images. Corresponding MNI x-, y-, z- coordinates are displayed. A threshold of p < 0.05 (corrected with TFCE family wise error) is used.

E.M. Coppen et al. / Parkinsonism and Related Disorders 46 (2018) 56e61 59

Fig. 2. Correlations between clinical motor scores and grey matter loss in manifest HD.
VBM analyses showing significant correlations between increased motor scores and reduction in grey matter volume. A threshold of p < 0.05 is used. Brain regions in blue are uncorrected for multiple testing and red-yellow brain regions are corrected with TFCE family wise error. Results are overlaid on sagittal, transversal and coronal slices of Montreal Neurological Institute (MNI)-152 standard space T1-weighted images. Corresponding MNI x-, y-, z- coordinates are displayed. UHDRS e TMS 1⁄4 Unified Huntington’s Disease Rating Scale e Total Motor Score. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

60 E.M. Coppen et al. / Parkinsonism and Related Disorders 46 (2018) 56e61

and controls, this structure was not included in further analyses in HD patients only.

After correction for multiple comparisons, there was a signifi- cant association between the UHDRS chorea score and UHDRS-TMS with the accumbens nucleus, caudate nucleus, putamen and pal- lidum in HD patients (Table 2). Thalamus and hippocampus vol- umes did not show any association with UHDRS motor subdomains.

3.2. Cortical grey matter volume

To assess differences in cortical grey matter volume between HD patients and controls, regional volumetric VBM analysis was per- formed. Significant grey matter volume reduction in HD patients was found in the motor cortex, visual cortex, and in the frontal and temporal lobes (Fig. 1 and supplementary Table S2).

In HD patients, VBM analysis showed that after correction for covariates and multiple comparisons, higher eye movement scores and UHDRS-TMS were associated with cortical volume loss of oc- cipital regions (Fig. 2 and supplementary Table S3).

4. Discussion

Our study showed that specific clinical motor signs in manifest HD are related to volume loss in different grey matter brain regions. Higher UHDRS chorea scores were particularly related to volume loss of subcortical structures, especially the accumbens nucleus, caudate nucleus, putamen and pallidum, whereas cortical brain regions did not. These findings suggest that volume loss in the subcortical regions are more involved in the development of chorea than cortical atrophy. It is well known that the medium-sized spiny neurons located in the striatum, that comprises of the caudate nucleus and putamen, are the most affected cells in HD [23]. As these neurons are involved in motor control, this might explain the association we found between striatal volume loss and the UHDRS chorea score.

In premanifest HD, general motor functioning is related to vol- ume loss of the putamen, caudate nucleus and pallidum [7,10,11]. Increased choreatic movements have been associated with striatal atrophy in premanifest HD [24]. However, to our knowledge, no studies have been performed that examined motor domains sepa- rately in relation with both subcortical and cortical changes. In addition to striatal volume loss, we observed a correlation between volume loss of the pallidum and higher UHDRS chorea scores. It is suggested that changes in the pallidum might be due to the loss of striato-pallidal fibers projecting from striatal medium spiny neu- rons, implying that volume loss of the pallidum is not due to cell loss within the pallidum [11].

Besides subcortical grey matter volume changes, we also investigated the association with cortical regions in patients with HD. Here, cortical grey matter volume loss was particularly asso- ciated with oculomotor dysfunction, but not with choreatic signs. Our findings are in contrast with results reported in a previous study where no correlations were found between cortical grey matter and motor functioning in premanifest HD [11]. A possible explanation might be that this previous study calculated lobular cortical volumes instead of investigating relationships with cortical volumes using a voxel-based technique. Another explanation could be that the HD patients included in our study were in a more advanced disease stage with more motor impairments, suggesting that involvement of cortical regions is more pronounced later in the disease. Still, UHDRS dystonia and hypokinetic-rigid scores did not show any significant correlations with subcortical volumes in our study.

The motor cortex, visual cortex, and cortical regions in the frontal and temporal lobes showed significant decrease in grey

matter volume in manifest HD compared to controls by means of voxel-based morphometry. These identified regions are consistent with findings in previous voxel-based studies [11,25e28]. Addi- tionally, we observed volume loss in visual cortical regions, which were associated with higher eye movement scores in HD gene carriers. It is known that fronto-striatal and occipital regions are important for oculomotor control and visual processing [29,30], providing a possible explanation for the observed correlations in these specific motor domains. These results are comparable to other studies observing associations between volume changes and quantitative motor functioning [27,28].

It has also been reported that more prominent bradykinesia and dystonia are related to cortical thinning of the anterior frontal re- gions, including the premotor and supplementary motor cortex [9,28]. In addition, finger tapping has been related to striatal and cortical atrophy [24,28]. Although we investigated changes in subcortical and cortical regions separately, there is a known inter- play between the basal ganglia and cerebral cortex. Especially changes of the basal ganglia-thalamo-frontal circuits are known to contribute to hyperkinetic movements such as chorea [11,23].

We did not find an association between some of the motor domains and grey matter regions, such as the cingulate gyrus. Since we aimed to focus on the clinical hallmark of HD, which is the presence of motor signs, this absent association might be caused by the fact that these brain regions are also involved in other domains than motor control. It has been reported that cortical brain atrophy, specifically in frontal, parietal and occipital lobes is related to a decline in cognitive functioning [9,27,30]. Future studies investi- gating the relationship between cognitive and psychiatric symp- toms of HD and volume reductions of the brain are necessary to further understand the pathogenesis of HD.

The lack of a relationship between dystonia and subcortical volumes in our study might also be caused by the relatively low scores on this item in our cohort of early stage HD patients. A further limitation of this study is the relatively smaller sample size of the control group, which could potentially influence the results. A larger sample size of the control group is preferred in future studies.

In conclusion, patients with HD can present with a heteroge- neous motor phenotype, consisting of chorea, dystonia, hypo- kinesia and/or balance disturbances. Our results demonstrate that chorea, which is the clinical hallmark of HD, is strongly associated with subcortical volume loss of the striatum and pallidum and not with cortical atrophy. Oculomotor dysfunction, however, seems to be more related to cortical volume changes, especially in occipital regions. Thus, there is a widespread contribution of different brain regions resulting in the overall clinical motor presentation seen in HD patients. We showed that not only subcortical volume loss is involved in the expression of motor disturbances, but also, although to a much lesser extent, cortical degeneration.

Authors’ roles

E.M. Coppen, study concept and design, data acquisition, anal- ysis and interpretation of data, drafting the manuscript.

M. Jacobs, study concept and design, data acquisition, analysis and interpretation of data, drafting the manuscript.

A.A. van den Berg-Huysmans, analysis and interpretation of MRI data, critical revision of the manuscript for intellectual content.

J. van der Grond, study concept and design, analysis and inter- pretation of data, critical revision of the manuscript for intellectual content.

R.A.C. Roos, study concept and design, analysis and interpreta- tion of data, critical revision of the manuscript for intellectual content.

All authors have approved the final version of the manuscript. Study funding

No funding.

Financial disclosures

E.M. Coppen, M. Jacobs, A.A. van den Berg-Huysmans and J. van der Grond report no conflict of interest.

R.A.C. Roos receives research grants from TEVA pharmaceuticals and is advisor for UniQure.

Acknowledgements

The authors would like to thank all patients and their relatives who participated in this study. Also, we would like to thank the study investigators for collecting all the data that was used in this study.

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.11.001.

References

. [1] R.A.C. Roos, Huntington’s disease: a clinical review, Orphanet J. Rare Dis. 5 (2010) 40.

. [2] T.M. Blekher, R.D. Yee, S.C. Kirkwood, A.M. Hake, J.C. Stout, M.R. Weaver, T.M. Foroud, Oculomotor control in asymptomatic and recently diagnosed individuals with the genetic marker for Huntington’s disease, Vis. Res. 44 (2004) 2729e2736.

. [3] P. Thompson, A. Berardelli, J. Rothwell, B. Day, J.P. Dick, R. Benecke, C. Marsden, The coexistence of bradykinesia and chorea in Huntington’s dis- ease and its implications for theories of basal ganglia control of movement, Brain 111 (1988) 223e244.

. [4] M. Jacobs, E.P. Hart, E.W. van Zwet, A.R. Bentivoglio, J.M. Burgunder, D. Craufurd, R. Reilmann, C. Saft, R.A.C. Roos, Progression of motor subtypes in Huntington’s disease: a 6-year follow-up study, J. Neurol. 263 (2016) 2080e2085.

. [5] Huntington Study Group, Unified Huntington’s disease rating scale: reliability and consistency, Mov. Disord. 11 (1996) 136e142.

. [6] S. De la Monte, J. Vonsattel, E. Richardson, Morphometric demonstration of atrophic changes in cerebral cortex, white matter and neostriatum in Hun- tington’s disease, J. Neuropathol. Exp. Neurol. 47 (1988) 516e525.

. [7] C.K.Jurgens,L.VanDeWiel,A.C.G.M.VanEs,Y.M.Grimbergen,M.N.W.Witjes- Ane , J. Van Der Grond, H.A.M. Middelkoop, R.A.C. Roos, Basal ganglia volume and clinical correlates in “preclinical” Huntington’s disease, J. Neurol. 255 (2008) 1785e1791.

. [8] B. Go mez-Anso n, M. Alegret, E. Mun~oz, G.C. Monte , E. Alayrach, A. S anchez, M. Boada, E. Tolosa, Prefrontal cortex volume reduction on MRI in preclinical Huntington’s disease relates to visuomotor performance and CAG number, Park. Relat. Disord. 15 (2009) 213e219.

. [9] H.D. Rosas, D.H. Salat, S.Y. Lee, A.K. Zaleta, V. Pappu, B. Fischl, D. Greve, N. Hevelone, S.M. Hersch, Cerebral cortex and the clinical expression of Huntington’s disease: complexity and heterogeneity, Brain 131 (2008) 1057e1068.

. [10] S.J.A. van den Bogaard, E.M. Dumas, T.P. Acharya, H. Johnson, D.R. Langbehn, R.I. Scahill, S.J. Tabrizi, M.A. Van Buchem, J. Van Der Grond, R.A.C. Roos, Early atrophy of pallidum and accumbens nucleus in Huntington’s disease, J. Neurol. 258 (2011) 412e420.

. [11] E.H. Aylward, D.L. Harrington, J.A. Mills, P.C. Nopoulos, S. Diego, H. System, S. Diego, Regional atrophy associated with cognitive and motor function in prodromal Huntington disease, J. Huntingt. Dis. 2 (2013) 477e489.

. [12] K.Marder,H.Zhao,R.H.Myers,M.Cudkowicz,E.Kayson,K.Kieburtz,C.Orme,

J. Paulsen, J. Penney, E. Siemers, I. Shoulson, Rate of functional decline in

Huntington’s disease, Neurology 54 (2000) 452e458.
[13] N. Mahant, E. Mccusker, K. Byth, S. Graham, Huntington’s disease: clinical

correlates of disability and progression, Neurology 61 (2003) 1085e1092. [14] S.M. Smith, M. Jenkinson, M.W. Woolrich, C.F. Beckmann, T.E.J. Behrens, H. Johansen-Berg, P.R. Bannister, M. De Luca, I. Drobnjak, D.E. Flitney, R.K. Niazy, J. Saunders, J. Vickers, Y. Zhang, N. De Stefano, J.M. Brady, P.M. Matthews, Advances in functional and structural MR image analysis and

implementation as FSL, Neuroimage 23 (2004) S208eS219.
[15] S.M. Smith, Y. Zhang, M. Jenkinson, J. Chen, P.M. Matthews, A. Federico, N. De Stefano, Accurate, robust, and automated longitudinal and cross-sectional

brain change analysis, Neuroimage 17 (2002) 479e489.
[16] M. Jenkinson, P. Bannister, M. Brady, S. Smith, Improved optimization for the

robust and accurate linear registration and motion correction of brain images,

Neuroimage 17 (2002) 825e841.
[17] B. Patenaude, S.M. Smith, D.N. Kennedy, M. Jenkinson, A bayesian model of

shape and appearance for subcortical brain segmentation, Neuroimage 56

(2011) 907e922.
[18] S.M. Smith, Fast robust automated brain extraction, Hum. Brain Mapp. 17

(2002) 143e155.
[19] J. Ashburner, K.J. Friston, Voxel-based morphometrydthe methods, Neuro-

image 11 (2000) 805e821.
[20] C. Good, I. Johnsrude, J. Ashburner, R. Henson, K. Friston, R. Frackowiak,

A voxel-based morphometric study of ageing in 465 normal adult human

brains, Neuroimage 14 (2001) 21e36.
[21] A.M. Winkler, G.R. Ridgway, M.A. Webster, S.M. Smith, T.E. Nichols, Permu-

tation inference for the general linear model, Neuroimage 92 (2014) 381e397. [22] S.M. Smith, T.E. Nichols, Threshold-free cluster enhancement: addressing problems of smoothing, threshold dependence and localisation in cluster

inference, Neuroimage 44 (2009) 83e98.
[23] A. Reiner, R.L. Albin, K.D. Anderson, C.J. D’Amato, J.B. Penney, A.B. Young,

Differential loss of striatal projection neurons in Huntington disease, Proc.

Natl. Acad. Sci. U. S. A. 85 (1988) 5733e5737.
[24] K.M. Biglan, C.A. Ross, D.R. Langbehn, E.H. Aylward, J.C. Stout, S. Queller,

N.E. Carlozzi, K. Duff, L.J. Beglinger, J.S. Paulsen, Motor abnormalities in pre- manifest persons with Huntington’s disease: the PREDICT-HD study, Mov. Disord. 24 (2009) 1763e1772.

[25] J. Kassubek, F.D. Juengling, T. Kioschies, K. Henkel, J. Karitzky, B. Kramer, D. Ecker, J. Andrich, C. Saft, P. Kraus, A.J. Aschoff, A.C. Ludolph, G.B. Landwehrmeyer, Topography of cerebral atrophy in early Huntington’s disease: a voxel based morphometric MRI study, J. Neurol. Neurosurg. Psy- chiatry 75 (2004) 213e220.

[26] S.J. Tabrizi, D.R. Langbehn, B.R. Leavitt, R.A.C. Roos, A. Durr, D. Craufurd, C. Kennard, S.L. Hicks, N.C. Fox, R.I. Scahill, B. Borowsky, A.J. Tobin, H.D. Rosas, H. Johnson, R. Reilmann, B. Landwehrmeyer, J.C. Stout, Biological and clinical manifestations of Huntington’s disease in the longitudinal TRACK-HD study: cross-sectional analysis of baseline data, Lancet Neurol. 8 (2009) 791e801.

[27] R.I. Scahill, N.Z. Hobbs, M.J. Say, N. Bechtel, S.M.D. Henley, H. Hyare, D.R. Langbehn, R. Jones, B.R. Leavitt, R.A.C. Roos, A. Durr, H. Johnson, S. Lehe ricy, D. Craufurd, C. Kennard, S.L. Hicks, J.C. Stout, R. Reilmann, S.J. Tabrizi, Clinical impairment in premanifest and early Huntington’s disease is associated with regionally specific atrophy, Hum. Brain Mapp. 34 (2013) 519e529.

[28] N. Bechtel, R.I. Scahill, H.D. Rosas, T. Acharya, S.J.A. Van Den Bogaard, C. Jauffret, M.J. Say, A. Sturrock, H. Johnson, C.E. Onorato, D.H. Salat, A. Durr, B.R. Leavitt, R.A.C. Roos, G.B. Landwehrmeyer, D.R. Langbehn, J.C. Stout, S.J. Tabrizi, R. Reilmann, Tapping linked to function and structure in pre- manifest and symptomatic Huntington disease, Neurology 75 (2010) 2150e2160.

[29] E. Lobel, P. Kahane, U. Leonards, M. Grosbras, S. Lehericy, D. Bihan, A. Berthoz, Localization of human frontal eye fields: anatomical and functional findings of functional magnetic resonance imaging and intracerebral electrical stimula- tion, J. Neurosurg. 95 (2001) 804e815.

[30] E.B. Johnson, E.M. Rees, I. Labuschagne, A. Durr, B.R. Leavitt, R.A.C. Roos, R. Reilmann, H. Johnson, N.Z. Hobbs, D.R. Langbehn, J.C. Stout, S.J. Tabrizi, R.I. Scahill, E. Axelson, N. Bechtel, S.J.A. Van Bogaard, S. Bohlen, J. Callaghan, C. Campbell, M. Campbell, D. Cash, A. Coleman, D. Craufurd, R.D. Santos, J. Decolongon, E. Dumas, N. Fox, C. Frost, J. Van der Grond, E. T’Hart, S. Hicks, N. Hobbs, C. Jauffret, R. Jones, D. Justo, C. Kennard, N. Lahiri, B. Landwehmeyer, S. Lehericy, I. Malone, C. Marelli, C. Milchman, K. Nigaud, G. Owen, T. Pepple, S. Queller, J. Read, M. Say, A. Sturrock, R. Valabre gue, C. Wang, The impact of occipital lobe cortical thickness on cognitive task performance: an investi- gation in Huntington’s Disease, Neuropsychologia 79 (2015) 138e146.

E.M. Coppen et al. / Parkinsonism and Related Disorders 46 (2018) 56e61 61

Parkinsonism and Related Disorders 46 (2018) 62e68

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Cortical thinning correlates of changes in visuospatial and visuoperceptual performance in Parkinson’s disease: A 4-year follow- up

A.I. Garcia-Diaz a, b, B. Segura a, b, H.C. Baggio a, b, C. Uribe a, b, A. Campabadal a, b, c, A. Abos a, b, M.J. Marti c, d, e, F. Valldeoriola c, d, e, Y. Compta b, c, d, e, N. Bargallo b, c, f,

C. Junque a, b, c, d, e, *

a Department of Medicine, Faculty of Medicine and Health Science, University of Barcelona, Barcelona, Catalonia, Spain
b Institute of Neurosciences, University of Barcelona, Barcelona, Catalonia, Spain
c Institut D’Investigacions Biom ediques August Pi I Sunyer (IDIBAPS), Barcelona, Catalonia, Spain
d Centro de Investigacio n Biom edica en Red Sobre Enfermedades Neurodegenerativas (CIBERNED), Hospital Clinic de Barcelona, Barcelona, Catalonia, Spain e Movement Disorders Unit, Neurology Service, Hospital Clínic de Barcelona, Barcelona, Catalonia, Spain

f Centre de Diagnostic per La Imatge, Hospital Clinic, Barcelona, Catalonia, Spain

articleinfo

Article history:

Received 23 July 2017 Received in revised form 26 October 2017
Accepted 6 November 2017

Keywords:

Parkinson’s disease Neuropsychology
MRI
Cortical thickness Longitudinal data Visuospatial functions

abstract

Background: Growing evidence highlights the relevance of posterior cortically-based cognitive deficits in Parkinson’s disease (PD) as possible biomarkers of the evolution to dementia. Cross-sectional correla- tional studies have established a relationship between the degree of atrophy in posterior brain regions and visuospatial and visuoperceptual (VS/VP) impairment. The aim of this study is to address the pro- gressive cortical thinning correlates of VS/VP performance in PD.

Methods: Forty-four PD patients and 20 matched healthy subjects were included in this study and fol- lowed for 4 years. Tests used to assess VS/VP functions included were: Benton’s Judgement of Line Orientation (JLOT), Facial Recognition (FRT), and Visual Form Discrimination (VFDT) Tests; Symbol Digit Modalities Test (SDMT); and the Pentagon Copying Test (PCT). Structural magnetic resonance imaging data and FreeSurfer were used to evaluate cortical thinning evolution.

Results: PD patients with normal cognition (PD-NC) and PD patients with mild cognitive impairment (PD-MCI) differed significantly in the progression of cortical thinning in posterior regions. In PD-MCI patients, the change in VS/VP functions assessed by PCT, JLOT, FRT, and SMDT correlated with the symmetrized percent change of cortical thinning of occipital, parietal, and temporal regions. In PD-NC patients, we also observed a correlation between changes in FRT and thinning in parieto-occipital regions.

Conclusion: In this study, we establish the neuroanatomical substrate of progressive changes in VS/VP performance in PD patients with and without MCI. In agreement with cross-sectional data, VS/VP changes over time are related to cortical thinning in posterior regions.

1. Introduction

Parkinson’s disease (PD) is a heterogeneous neurodegenerative disorder that manifests with a wide range of nonmotor symptoms. Recent initiatives have aimed to depict the features and evolution

* Corresponding author. Department of Medicine, Faculty of Medicine and Health Sciences, University of Barcelona, Casanova 143, 08036, Barcelona, Spain.

E-mail address: cjunque@ub.edu (C. Junque).

of cognitive decline in PD [1e4].
Impairment in specific cognitive domains has been associated

with a differential risk of cognitive decline. While executive func- tions are widely recognized to be impaired in PD even at early disease stages [1,5,6], interest in the role of posterior cortically- based functions as biomarkers of the cognitive evolution to de- mentia (PDD) has increased [1,7,8].

Several cross-sectional structural MRI correlational studies have established a relationship between the degeneration of posterior

https://doi.org/10.1016/j.parkreldis.2017.11.003
1353-8020/© 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

brain regions and cognitive impairment [9e12]. Specifically, pre- vious studies by our group showed that visuospatial and visuo- perceptual (VS/VP) tests are suitable to reflect cortical thinning in lateral temporo-parietal regions in PD patients [13,14].

Longitudinal studies have assessed structural gray matter dif- ferences over time in PD [15,16], and the progression of cognitive impairment has been related to degeneration of several cortical regions, including bilateral frontal and temporoparietal areas [16e18]. Progressive atrophy in widespread brain regions, such as the bilateral temporal and right occipital medial lobes, left superior frontal gyrus, and inferior parietal cortex, has been related to worsening in measures of global cognition [17,18]. Also, volumetric studies have associated the decline in executive functions with mainly bilateral frontal areas [19,20]. However, to the best of our knowledge, the relationship between the impairment of specific VS/VP functions and cortical thinning over time has yet to be studied. The aims of this study are (1) to address differential pro- gressive gray matter loss between PD patients and healthy controls (HC), as well as (2) to investigate the changes over time in VS/VP functions in PD patients grouped according to cognitive status and their relationship with progressive cortical degeneration.

2. Methods

2.1. Participants

The cohort of this study was recruited from an outpatient movement disorders clinic (Parkinson’s Disease and Movement Disorders Unit, Service of Neurology, Hospital Clínic, Barcelona, Spain), and HC were recruited from Institut de l’Envelliment (Bar- celona, Spain). All participants are part of an ongoing longitudinal study, composed of 121 PD patients and 48 healthy subjects in the initial screening phase. Both groups were matched for age, sex, and years of education.

Inclusion criteria for participants consisted of fulfilling the diagnostic criteria for PD established by the UK PD Society Brain Bank [21]. Exclusion criteria consisted of: presence of dementia according to the Movement Disorder Society criteria [22], Hoehn and Yahr scale (H&Y) score >3, juvenile-onset PD, presence of psychiatric and/or neurologic comorbidity, low global IQ score estimated by the Vocabulary subtest of the Wechsler Adult Intel- ligence Scale, 3rd edition (scalar score 7 points), Mini-Mental State Examination (MMSE) score 25, claustrophobia, imaging findings on MRI not compatible with PD other than mild white matter hyperintensities in the FLAIR sequence, and MRI artifacts. The final sample at the baseline assessment consisted of 92 PD patients and 36 controls. A follow-up assessment was pursued after approximately four years (see Table 1), with a sample of 20 HC and 44 PD patients. Only subjects with baseline and follow-up assess- ments were included in this study (see Supplementary Fig. 1).

Motor symptoms were assessed with the Unified Parkinson’s Disease Rating Scale, motor section (UPDRS-III). All PD patients were taking antiparkinsonian drugs, consisting of different com- binations of L-DOPA, cathecol-O-methyltransferase inhibitors, monoamine oxidase inhibitors, dopamine agonists, and amanta- dine. In order to standardize doses, the L-DOPA equivalent daily dose (LEDD) [23] was calculated. All assessments were done while patients were under the effect of their usual medication (“on” state).

In line with the PD-MCI Movement Disorder Society Task Force (MDSTF) recommendations [24], we assessed five cognitive do- mains as previously described [12]. We divided the subjects into three groups: HC, PD patients without MCI (PD-NC), and PD pa- tients with MCI (PD-MCI) at baseline. Expected z scores adjusted for age, sex, and education for each test and each subject were

calculated based on a multiple regression analysis performed in the HC group [3]. As in previous studies [12,25], the presence of MCI was established if the z score for a given test was at least 1.5 lower than the expected score in at least two tests in one domain, or in at least one test per domain in at least two domains.

Written informed consent was obtained from all study partici- pants after full explanation of the procedures. The study was approved by the institutional Ethics Committee from the University of Barcelona (IRB00003099).

2.2. Visuospatial and visuoperceptual assessment

All participants underwent a comprehensive neuropsychologi- cal assessment with VS/VP tests usually employed to evaluate the cognitive status of PD patients. The battery of tests chosen in this study is the same as that used in a previous cross-sectional study that addressed the neuroanatomical correlates of VS/VP deficits in PD [14]. The tests included were the pentagon copying test (PCT) from the MMSE, scored according to the Modified Mini-Mental State criteria (3MS); Benton’s Judgment of Line Orientation test (JLOT), Visual Form Discrimination test (VFDT), and Facial Recog- nition test (FRT); and Symbol Digits Modalities test (SDMT).

2.3. MRI acquisition

Magnetic resonance images (MRI) were acquired with a 3T scanner (MAGNETOM Trio, Siemens, Germany) at baseline and follow-up. The scanning protocol included high-resolution 3- dimensional T1-weighted images acquired in the sagittal plane (TR 1⁄4 2300 ms, TE 1⁄4 2.98 ms, TI 1⁄4 900 ms, 240 slices, FOV 1⁄4 256 mm; matrix size 1⁄4 256 256; 1 mm isotropic voxel and an axial FLAIR sequence (TR 1⁄4 9000 ms, TE 1⁄4 96 ms).

2.4. Longitudinal cortical thickness

FreeSurfer software (version 5.1; available at http://surfer.nmr. harvard.edu) was used to obtain structural measures as previ- ously described [13]. After processing each subject cross- sectionally, in order to perform the longitudinal analyses of the data, within-subject templates [26] and corresponding longitudinal files were created for each time point for each subject. Briefly, a template volume for each subject using information from all of their time points and an average image were created using robust, inverse, consistent registration [27]. All time points were con- structed through unbiased mean images and later aligned. After registration and creation of the templates, images from all time points are mapped to the template location and averaged, and processed with the default cross-sectional stream. The symme- trized percent change was used for longitudinal analyses of cortical thickness: [(Thickness at time point 1 e Thickness at time point 2)/ Interval between assessments)]/[0.5*(Thickness at time point 1 þ Thickness at time point 2)].

Comparisons between groups and regressions were assessed using vertex-by-vertex general linear models. Multiple contrasts were carried out to assess differences between all study subgroups (HC vs. all PD patients; HC vs. PD-NC; HC vs. PD-MCI; and PD-NC vs. PD-MCI). Regression models included symmetrized percent change as an independent factor and cognitive scores as dependent factors. In order to avoid clusters appearing significant purely by chance (i.e., false positives), Monte Carlo null-Z simulation with 10,000 iterations was applied to cortical thickness maps to provide clus- terwise correction for multiple comparisons. Results were thresh- olded at a corrected p value of 0.05.

A.I. Garcia-Diaz et al. / Parkinsonism and Related Disorders 46 (2018) 62e68 63

64 A.I. Garcia-Diaz et al. / Parkinsonism and Related Disorders 46 (2018) 62e68 Table 1

Demographic and clinical data of the participants at baseline. HC(n1⁄420)

PD-NC (n 1⁄4 28)

3.89 ± 0.41 59.50 ± 9.58 20/8
12.96 ± 4.87 29.54 ± 0.69 6.50 ± 3.87 53.00 ± 10.21 700.79 ± 470.61 17/28 (60.71%) 13.93 ± 9.19

1: 11 1,5: 1 2: 12 2,5: 2 3:2

PD-MCI (n 1⁄4 16)

3.94 ± 0.59 64.63 ± 9.67 10/6
11.25 ± 5.94 28.69 ± 1.54 8.03 ± 6.73 56.91 ± 12.22 675.63 ± 535.21 9/16 (56.25%) 11.75 ± 11.01 1:6

2:8 2.5: 1 3:1

F,x2,t,U

0.065a 3.010a 2.286b 1.045a 6.481a* 0.814c 1.136c 0.162c 1.206b 185.000d 219.000d

Interval (years) Age
Sex (male/female) Education

MMSE
Evolution (years) Age at onset
LEDD
Dopamine agonists UPDRS-III
H&Y

3.90 ± 0.32 65.50 ± 8.00 10/10
11.10 ± 4.13 29.75 ± 0.44

HC: Healthy controls; PD-NC: Parkinson’s disease patients without mild cognitive impairment; PD-MCI: Parkinson’s disease patients with mild cognitive impairment; MMSE: Mini-Mental State Examination; LEDD: Levodopa Equivalent Daily Dose; UPDRS-III: Unified Parkinson’s disease Rating Scale; H&Y: Hoehn and Yahr scale. Values are presented as mean ± standard deviation.
*significant at p < 0.01.

a F ANOVA statistics.
b Pearson’s c2 statistics.
c Student t-test statistics.
d Mann-Whitney U statistics.

2.5. Global atrophy measures

Gray matter and lateral ventricular volumes were obtained automatically via whole brain segmentation procedures performed with FreeSurfer (version 5.1; available at http://surfer.nmr.harvard. edu). Intracranial volume (ICV) was entered as a covariate of no interest in comparisons of global atrophy measures. Mean thick- ness for both hemispheres was calculated as follows: [(left hemi- sphere thickness * left hemisphere surface area) þ (right hemisphere thickness * right hemisphere surface area)]/(left hemisphere surface area þ right hemisphere surface area).

2.6. Statistical analyses

Statistical analyses of neuropsychological, demographic, clinical, and MRI volumetric data variables were carried out using the sta- tistical package SPSS-20 (2011; Armonk, NY: IBM Corp.). For the baseline analysis of demographic variables, Student t tests, ANOVA, Pearson’s c2 statistics, and Mann-Whitney’s U were used as appropriate.

A longitudinal variable was created for each test used to pair neuropsychological data with the structural longitudinal measure of symmetrized percent change, and was used in all statistical and structural analyses of the study.

For longitudinal clinical, neuropsychological, and structural variables, repeated measures general linear model was used to assess group differences over time in quantitative variables; and post-hoc tests were performed using Bonferroni correction for multiple comparisons. To address group and time effects in quali- tative variables, Kruskal-Wallis H, Friedman’s F, or Pearson’s c2 statistics were used as appropriate.

3. Results

3.1. Sociodemographic and clinical data

Demographic and clinical data of the participants at baseline are summarized in Table 1. No significant differences were found be- tween study groups in age, sex, education, clinical variables asso- ciated with PD, or the interval between assessments. The

characteristics of the subjects who remained as study participants and those who dropped out are summarized in Supplementary Table 1.

The longitudinal evolution of clinical variables in all PD patients is summarized in Supplementary Table 2. Medication and motor measures showed no significant progression in this follow-up period, and did not differ between PD-NC subjects and patients with impaired cognition. MMSE showed significant group differ- ences at baseline as well as group and time effects in the longitu- dinal analysis.

The progression of the detailed neuropsychological evaluation can be found in Table 2 and Supplementary Table 3. Aside from VS/ VP measures, significant group-by-time interactions were seen in tests of attention and working memory. At follow-up, 17 patients remained as PD-NC (60.71%), 9 remained as PD-MCI (56.25%), 5 PD- MCI patients reverted to PD-NC (31.25%), 11 PD-NC patients pro- gressed to PD-MCI (39.29%) and 2 PD-MCI patients fulfilled criteria for PDD (3.1%).

3.2. Visuospatial and visuoperceptual performance

All VS/VP tests showed significant group differences (see Table 2). Significant time and group-by-time interaction effects were observed for the SDMT. Post-hoc analyses evidenced that differences were found between HC and PD-MCI in all contrasts.

PCT differed between groups at baseline and follow-up when scored according to the original MMSE criteria (c2 1⁄4 12.800, p 1⁄4 0.002; c2 1⁄4 8.957, p 1⁄4 0.011 respectively) as well as according to Williams-Gray et al. criteria [1,7,8] (c2 1⁄4 9.295, p 1⁄4 0.010; c2 1⁄4 8.987, p 1⁄4 0.011 respectively); however, no significant time effects were observed for any groups.

3.3. MRI evolution

Imaging analyses revealed that, compared with PD-NC patients, PD-MCI patients exhibited significantly greater progressive cortical thinning in left lateral occipital and inferior parietal regions, and in right medial temporal regions (see Fig. 1 and Supplementary Table 4). Cortical thinning differences between HC and PD-NC, and between HC and all PD patients, were not significant.

A.I. Garcia-Diaz et al. / Parkinsonism and Related Disorders 46 (2018) 62e68

Table 2

Group comparison of VS/VP performance between healthy subjects, PD patients without MCI, and PD patients with MCI.

HC PD-NC

PD-MCI Baseline

9.19 ± 1.11 21.06 ± 5.74 26.88 ± 3.52 20.44 ± 3.39 36.38 ± 18.91

HC: Healthy controls;
Pentagon Copying Test; JLOT: Judgment of Line Orientation Test; VFDT: Visual Form Discrimination Test; FRT: Facial Recognition Test; SDMT: Symbol Digit Modalities Test. Values are presented as mean ± standard deviation.

Baseline

PCT 9.70 ± 0.48 JLOT 23.80 ± 2.91 VFDT 30.00 ± 2.25 FRT 22.70 ± 1.92 SDMT 44.67 ± 8.40

Follow-up Baseline

9.65 ± 0.59 9.79 ± 0.57 25.00 ± 3.45 24.07 ± 3.81 29.40 ± 2.26 29.54 ± 2.25 22.85 ± 1.87 22.14 ± 2.48 46.90 ± 7.51 48.54 ± 10.61

Follow-up

9.64 ± 0.68 25.21 ± 3.24 29.64 ± 2.53 21.79 ± 2.73 43.75 ± 12.68

Follow-up

8.56 ± 2.73 19.56 ± 8.27 27.81 ± 4.20 20.07 ± 3.26 31.31 ± 19.87

F (Group)

4.428 (p 1⁄4 0.016) 6.311 (p 1⁄4 0.003) 6.028 (p 1⁄4 0.004) 4.992 (p 1⁄4 0.010) 5.109 (p 1⁄4 0.009)

F (Time)

2.749 (p 0.292 (p 0.130 (p 0.447 (p 7.552 (p

1⁄4 0.102) 1⁄4 0.591) 1⁄4 0.720) 1⁄4 0.506) 1⁄4 0.008)

F (Group by time) Post-hoc P

1.022 (p 1⁄4 0.366) HC/PD-MCI: 0.047 2.597 (p 1⁄4 0.083) HC/PD-MCI: 0.013 1.024 (p 1⁄4 0.365) HC/PD-MCI: 0.008 0.346 (p 1⁄4 0.709) HC/PD-MCI: 0.008 6.574 (p 1⁄4 0.003) HC/PD-MCI: 0.029

PD-NC: Parkinson’s disease patients without mild cognitive impairment; PD-MCI: Parkinson’s disease patients with mild cognitive impairment. PCT:

Group comparison of global MRI atrophy parameters evidenced that mean thickness differed between groups and had a time effect (F(Group) 1⁄4 7.711; p 1⁄4 0.001; F(Time) 1⁄4 9.891, p 1⁄4 0.003; Post-hoc P: PD-MCI vs HC 1⁄4 0.001; PD-MCI vs PD-NC 1⁄4 0.016), whereas the increase in the volume of the lateral ventricular system achieved statistical significance for time and the interaction between group and time (F(Time): 88.596; p < 0.0001; F(GroupxTime) 1⁄4 4.745; p 1⁄4 0.012) (see Supplementary Fig. 2).

3.4. Cortical thickness correlates of visuospatial and visuoperceptual changes

Whole-brain imaging analyses showed significant correlations between changes in VS/VP measures and cortical thinning over time. In the PD-NC group, FRT also correlated with cortical thinning in the left lateral occipital area.

In PD-MCI patients, changes in PCT scores over time showed a significant cluster in the left entorhinal region that involved the middle and inferior temporal gyri, the medial temporal pole, and the parahippocampal, fusiform, lingual, and lateral occipital cortices. JLOT was significantly related to cortical atrophy in clus- ters located in the left insula, inferior and superior temporal areas, and the right fusiform gyrus, which extended to the left temporal pole, entorhinal, fusiform, and lingual cortices. FRT scores corre- lated significantly with cortical thinning in the left lingual gyrus. SDMT showed significant correlations with reductions in the left superior temporal, parahippocampal and lingual, as well as the right parahippocampal cortices (see Fig. 2, Supplementary Table 5). This pattern of anatomical correlates was maintained when only subjects with sustained cognitive diagnosis at follow-up were included in the analysis (see Supplementary Fig. 3).

We performed complementary analyses to study the cross- sectional correlates of the tests used in this study and we observed a pattern of posterior atrophy more pronounced in PD-

MCI patients (see Supplementary Figs. 4a and 4b and Supplementary Table 6). We analyzed the association between cortical thinning over time and the significant longitudinal differ- ences found in neuropsychological measures relative to other cognitive domains. In PD-NC patients, the Stroop colors test correlated significantly with left superior parietal and frontal re- gions. In PD-MCI patients, a non-specific widespread pattern of anterior and posterior regions correlated bilaterally with TMT-A and Stroop colors tests (see Supplementary Fig. 5 and Supplementary Table 7).

4. Discussion

In the present study, we aimed to investigate the longitudinal differences in cortical thinning between PD patients and healthy subjects, as well as the relationship between the progressive loss of VS/VP functions and the cortical degeneration underlying these changes in PD patients, grouped according to their cognitive status using the Movement Disorder Society Task Force criteria.

Our results evidence that all the neuropsychological tests with a posterior cortically-based component used in this work are sensi- tive to detect VS/VP impairment in MCI patients. However, among the five VS/VP tests used, only the SDMT showed a significant time effect as well as a significant group-by-time interaction, indicating that it may be useful for the evaluation of progressive cognitive impairment in PD. Previous research in PD cognitive deterioration has also described the progressive decline of visuospatial and visuoconstructive functions [1,4,7,8,28,29]. In longitudinal studies, an important issue is the distinction between cognitive and motor deficits, as there are several VS/VP tests, such as the clock drawing, the pentagon test drawing or the block design, that have a strong motor component. By contrast, in the SDMT, the motor component is very low, mainly involving eye tracking. It thus seems to be a suitable test for PD follow-up studies. In agreement with our

Fig. 1. Vertex-wise symmetrized percent change in cortical thickness differences between study groups. The scale bar shows P values. PD-NC: Parkinson’s disease patients without mild cognitive impairment; PD-MCI: Parkinson’s disease patients with mild cognitive impairment.

66 A.I. Garcia-Diaz et al. / Parkinsonism and Related Disorders 46 (2018) 62e68

Fig. 2. Vertex-wise symmetrized percent change in cortical thickness correlations with VS/VP measures in PD patients. The scale bar shows P values. PD-NC: Parkinson’s disease patients without mild cognitive impairment; PD-MCI: Parkinson’s disease patients with mild cognitive impairment; FRT: Facial Recognition Test; SDMT: Symbol Digit Modalities Test; PCT: Pentagon Copying Test; JLOT: Judgment of Line Orientation Test.

findings, a 3-year multi-center follow-up of a large sample of PD patients, using short versions of the JLOT and the SDMT, found statistically significant effects for both tests, but the differences

were stronger for the SDMT [29].
In our previous cross-sectional studies, we demonstrated a

relationship between visuospatial and visuoperceptual

performance and cortical thickness in bilateral temporo-parietal- occipital areas, and widespread posterior-anterior white matter microstructure alterations [13,14]. Interestingly, in the present study, we have established a relationship between the progressive worsening in VS/VP performance and bilateral degeneration of posterior cortical regions. In PD-MCI patients, PCT, JLOT, FRT, and SDMT evidenced significant correlates with temporal, occipital, and parietal cortices. In PD-NC we also observed a relationship between decreases in FRT scores and the rate of thinning in the occipito- parietal cortex. We highlight the emergence of specific neuroana- tomical correlates in PD-MCI patients, in absence of a significant time effect in neuropsychological performance for most VS/VP tests. This finding reflects that, although performance in these tests did not change significantly over time at the group level, there was a variable progressive loss of visuospatial and visuoperceptual functions in some PD patients that was explained by the variability in thinning of specific posterior cortical brain regions. This notion is supported by the finding that PD-MCI patients exhibited extensive progressive reductions in posterior parieto-temporal cortical re- gions in comparison with their cognitively unimpaired PD patient peers, which is in agreement with recent findings using the same technique in large study samples [17,18].

The neurobiological basis for cognitive dysfunction in PD is unclear, and several factors have been implicated, including loss of dopaminergic, noradrenergic, serotonergic, and cholinergic pro- jections to limbic and cortical areas, as well as AD-type pathology [30]. Enhanced a-synuclein pathology, together with lysosomal deficits, have been linked to poor cognitive evolution in PD patients [31]. Functional cross-sectional studies with dopamine tracers and metabolic parameters have established the relevance of posterior regions in cognitive decline in PD [32,33], as well as the relationship between visuospatial impairment, posterior cortical regions, pre- dominant a-synucleinopathy, and worse cognitive evolution [34,35].

The strengths of our study are that we applied validated criteria and tests to determine cognitive diagnoses [24], established a considerable follow-up interval, used a sensitive technique to identify regional gray matter changes associated with PD [36], and used the same MRI scanner, avoiding the variability of multi-center data. Our study is limited by the size of the sample due to the considerable attrition, which could in turn affect the results observed. However, cross-sectional as well as large-scale longitu- dinal studies of other groups are in line with our current findings [11,17,18]. In our study, mean group ages could appear as relatively low considering the epidemiological data of PD patents. This might be due to the exclusion of demented PD patients, who tend to be older. In fact, the mean age of our sample is similar to those in the abovementioned studies that also focused on non-demented PD patients using larger cohorts [17,18].

In conclusion, the present study establishes the neuroanatom- ical substrate of the progressive deterioration of visuospatial and visuoperceptual performance in PD patients with and without mild cognitive impairment. This study reinforces previous findings on the differential progression of atrophy in patients with MCI, thus supporting the validity of this construct. These findings give evi- dence to the notion that the progression of posterior-cortically based cognitive tests is indicative of progressive cortical thinning in posterior brain regions.

Disclosure

Authors AIGD, BS, HCB, CU, AC, AA, MJM, FV, NB and CJ report no disclosure. YC has received funding, research support and/or hon- oraria in the last 5 years from Union Chimique Belge (UCB pharma), Lundbeck, Medtronic, Abbvie, Novartis, GSK, Boehringer, Pfizer,

Merz, Piramal Imaging and Esteve.

Author roles

1. Research project: A. Conception, B. Organization, C. Execution; 2. Statistical Analysis: A. Design, B. Execution, C. Review and Critique; 3. Manuscript: A. Writing of the first draft, B. Review and Critique.

Garcia-Diaz: 1C, 2A, 2B, 3A; Segura: 1B, 1C, 2A, 2B, 3B; Uribe: 1C; Campabadal: 1C; Abos: 1C; Baggio: 1B, 1C, 2B, 2C, 3B; Marti: 1B, 1C, 2C, 3B; Valldeoriola: 1B, 1C, 2C, 3B; Compta: 1B, 1C, 2C, 3B; Bargallo: 1B, 1C, 2C,3B; Junque: 1A, 1B, 1C, 2A, 2B, 2C, 3B.

Acknowledgements

This study was sponsored by Spanish Ministry of Economy and Competitiveness (PSI2013-41393-P), by Generalitat de Catalunya (2014SGR 98) and by Fundacio la Marato de TV3 in Spain (20142310).

CU was supported by a 2014 fellowship, Spanish Ministry of Economy and Competitiveness (BES-2014-068173) and co-financed by the European Social Fund (ESF).

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.11.003.

References

[1] C.H. Williams-Gray, T. Foltynie, C.E. Brayne, T.W. Robbins, R.A. Barker, Evo- lution of cognitive dysfunction in an incident Parkinson’s disease cohort, Brain 130 (2007) 1787e1798.

[2] Hely AM, Reid WGJ, Adena MA, Halliday GM, Morris JGL. The sydney multi- center study of Parkinson’s disease: the inevitability of dementia at 20 years. Mov. Disord.;23:837e844.

[3] D. Aarsland, K. Brønnick, J.P. Larsen, O.B. Tynes, G. Alves, Cognitive impaired in incident, untreated Parkinson disease. The Norwegian ParkWest Study, Neurology 72 (2009) 121e126.

[4] M. Broeders, R.M.A. de Bie, D.C. Velseboer, J.D. Speelman, D. Muslimovic , B. Schmand, Evolution of mild cognitive impairment in Parkinson disease, Neurology 81 (2013) 346e352.

[5] D.Muslimovi c,B.Post,J.D.Speelman,B.Schmand,Cognitiveprofileofpatients with newly diagnosed Parkinson disease, Neurology 65 (2005) 1239e1245.

[6] B.J. Lawrence, N. Gasson, A.M. Loftus, Prevalence and subtypes of mild cognitive impairment in Parkinson’s disease, Sci. Rep. 6 (2016) 33929.

[7] C.H.Williams-Gray,J.R.Evans,A.Goris,T.Foltynie,M.Ban,T.W.Robbins,etal., The distinct cognitive syndromes of Parkinson’s disease: 5 year follow-up of the CamPaIGN cohort, Brain 132 (2009) 2958e2969.

[8] C.H. Williams-Gray, S.L. Mason, J.R. Evans, T. Foltynie, C. Brayne, T.W. Robbins, R.A. Barker, The CamPaIGN study of Parkinson’s disease: 10-year outlook in an incident population-based cohort, J. Neurol. Neurosurg. Psychiatry 84 (2013) 1258e1264.

[9] Pagonabarraga J, Corcuera-Solano I, Vives-Gelabert Y, Llebaria G, García- S anchez C, Pascual-Sedano B, et al. Patter of regional cortical thinning asso- ciated with cognitive deterioration in Parkinson’s disease. PLoS One; 8: e54980.

[10] J.V. Filoteo, J.D. Reed, I. Litvan, D.L. Harrington, Volumetric correlates of cognitive functioning in nondemented patients with Parkinson’s disease, Mov. Disord. 3 (2014) 360e367.

[11] J.B.Pereira,P.Svenningsson,D.Weintraub,K.Brønnick,A.Lebedev,Westman, D. Aarsland, Initial cognitive decline is associated with cortical thinning in early Parkinson disease, Neurology 82 (2014) 2017e2025.

[12] Segura B, Baggio HC, Marti MJ, Valldeoriola F, Compta Y, Garcia-Diaz AI, et al. Cortical thinning associated with mild cognitive impairment in Parkinson’s disease. Mov. Disord., 29:1495e1503.

[13] A.I.Garcia-Diaz,B.Segura,H.C.Baggio,M.J.Marti,F.Valldeoriola,Y.Compta,et al., Structural MRI correlates of the MMSE and pentagon copying test in Parkinson’s disease, Park. Relat. Disord. 12 (2014) 1405e1410.

[14] A.I.Garcia-Diaz,B.Segura,H.C.Baggio,M.J.Marti,F.Valldeoriola,Y.Compta,et al., Structural brain correlations of visuospatial and visuoperceptual tests in Parkinson’s disease, J. Int. Neuropsychol. Soc. 17 (2017) 1e12.

[15] N. Ibarretxe-Bilbao, C. Junque, B. Segura, H.C. Baggio, M.J. Marti, F. Valldeoriola, et al., Progression of cortical thinning in early Parkinson’s disease, Mov. Disord. 27 (2012) 1746e1754.

A.I. Garcia-Diaz et al. / Parkinsonism and Related Disorders 46 (2018) 62e68 67

68 A.I. Garcia-Diaz et al. / Parkinsonism and Related Disorders 46 (2018) 62e68

. [16] Y. Compta, J.B. Pereira, J. Rios, Junque C. Ibarretxe-Bilbao, N. Bargallo, et al., Combined dementia-risk biomarkers in Parkinson’s disease: a prospective longitudinal study, Park. Relat. Disord. 19 (2013) 717e724.

. [17] A. Hanganu, C. Bedetti, C. Degroot, B. Mejia-Constain, A.L. Lafontaine, V. Soland, et al., Mild cognitive impairment is linked with faster rate of cortical thinning in patients with Parkinson’s disease longitudinally, Brain 137 (2014) 1120e1129.

. [18] E. Mak, L. Su, G.B. Williams, M.J. Firbank, R.A. Lawson, A.J. Yarnall, et al., Baseline and longitudinal grey matter changes in newly diagnosed Parkin- son’s disease: ICICLE-PD study, Brain 138 (2015) 2974e2986.

. [19] J.E. Lee, K.H. Cho, S.K. Song, H.J. Kim, H.S. Lee, Y.H. Sohn, P.H. Lee, Exploratory analysis of neuropsychological and neuroanatomical correlates of progressive mild cognitive impairment in Parkinson’s disease, J. Neurol. Neurosurg. Psy- chiatry 85 (2014) 7e16.

. [20] M.C. Wen, A. Ng, R.J. Chander, W.L. Au, L.C.S. Tan, N. Kandiah, Longitudinal brain volumetricl changes and their predictive effects on cognition among cognitively asymptomatic patients with Parkinson’s disease, Park. Relat. Dis- ord. 21 (2015) 483e488.

. [21] S.E. Daniel, A.J. Lees, Parkinson’s disease society brain bank, London: overview and research, J. Neural Transm. Suppl. 39 (1993) 165e172.

. [22] B. Dubois, D. Burn, C. Goetz, D. Aarsland, R.G. Brown, G.A. Broe, et al., Diag- nostic procedures for Parkinson’s disease dementia: recommendations from the movement disorder society task force, Mov. Disord. 22 (2007) 2314e2324.

. [23] C.L. Tomlinson, R. Stowe, S. Patel, C. Rick, R. Gray, C.E. Clarke, Systematic re- view of levodopa dose equivalency reporting in Parkinson’s disease, Mov. Disord. 25 (2010) 2649e2653.

. [24] I. Litvan, J. Goldman, A. Tro€ster, B. Schmand, D. Weintraub, R. Petersen, et al., Diagnostic criteria for mild cognitive impairment in Parkinson’s disease: movement disorder society task force guidelines, Mov. Disord. 27 (2012) 349e356.

. [25] H.C. Baggio, B. Segura, R. Sala-Llonch, M.J. Marti, F. Valldeoriola, Y. Compta, et al., Cognitive impairment and resting-state network connectivity in Parkin- son’s disease, Hum. Brain Mapp. 36 (2015) 199e212.

[26] M. Reuter, B. Fischl, Avoiding asymmetry-induced bias in longitudinal image processing, NeuroImage 57 (2011) 19e21.

[27] M. Reuter, H.D. Rosas, B. Fischl, Highly accurate inverse consistent registra- tion: a robust approach, NeuroImage 53 (2010) 1181e1196.

[28] D. Muslimovi c, b Schmand, J.D. Speelman, R.J. de Haan, Course of cognitive decline in Parkinson’s disease: a meta-analysis, J. Int. Neuropsychol. Soc. 13 (2007) 920e932.

[29] C. Caspell-Garcia, T. Simuni, D. Tosun-Turgut, I.W. Wu, Y. Zhang, M. Nalls, et al., Multiple modality biomarker prediction of cognitive impairment in pro- spectively followed de novo Parkinson disease, PLoS One 17 (12) (2017) e0175674.

[30] G.M. Halliday, J.B. Leverenz, J.S. Schneider, C.H. Adler, The neurobiological basis of cognitive impairment in Parkinson’s disease, Mov. Disord. 29 (2014) 634e650.

[31] R.N. Alcalay, E. Caccappolo, H. Mejia-Santana, M. Tang, L. Rosado, M. Orbe Reilly, et al., Cognitive performance of GBA mutation carriers with early-onset PD: the CORE-PD study, Neurology 78 (2012) 1434e1440.

[32] D. Arnaldi, C. Campus, M. Ferrara, F. Fam a, A. Picco, F. De Carli, et al., What predicts cognitive decline in de novo Parkinson’s disease? Neurobiol. Aging 33 (1127) (2012) e11e20.

[33] Y. Nishio, K. Yokoi, M. Uchiyama, Y. Mamiya, H. Watanabe, M. Gang, et al., Deconstructing psychosis and misperception symptoms in Parkinson’s dis- ease, J. Neurol. Neurosurg. Psychiatry 88 (2017) 722e729.

[34] C. Nombela, J.B. Rowe, S.E. Winder-Rhodes, A. Hampshire, A.M. Owen, D.P. Breen, et al., Genetic impact on cognition and brain function in newly diagnosed Parkinson’s disease: ICICLE-PD study, Brain 137 (2014) 2743e2758.

[35] T. Baba, Y. Hosokai, Y. Nishio, A. Kikuchi, K. Hirayama, K. Suzuki, et al., Lon- gitudinal study of cognitive and cerebral metabolic changes in Parkinson’s disease, J. Neurol. Sci. 273 (2017) 288e293.

[36] J.B. Pereira, N. Ibarretxe-Bilbao, M.J. Marti, Y. Compta, C. Junque, N. Bargallo, et al., Assessment of cortical degeneration in patients with Parkinson’s disease by voxel-based morphometry, cortical Folding, and cortical thickness, Hum. Brain Mapp. 33 (2012) 2521e2534.

Parkinsonism and Related Disorders 46 (2018) 69e73

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Olfactory dysfunctions in drug-naïve Parkinson’s disease with mild cognitive impairment*

Jin-Woo Park a, Do-Young Kwon a, *, Ji Ho Choi b, Moon-Ho Park a, Ho-Kyoung Yoon c

a Department of Neurology, Korea University Ansan Hospital, Ansan, South Korea
b Department of Otorhinolaryngology, Soonchunhyang University College of Medicine, Bucheon, South Korea c Department of Psychiatry, Korea University Ansan Hospital, Ansan, South Korea

articleinfo

Article history:

Received 3 July 2017 Received in revised form
21 September 2017 Accepted 13 November 2017

Keywords:

Mild cognitive impairment Parkinson’s disease Olfactory dysfunction

abstract

Background: Evaluation of olfactory function is valuable for the detection of pre-motor state of Parkin- son’s disease (PD). PD patients have an increased risk of associated dementia and one-third of PD pa- tients have mild cognitive impairment (MCI) at the time of diagnosis. However, the characteristics of olfactory dysfunction in PD-MCI patients are unclear. This study examined the relationship between olfactory dysfunction and cognitive function in drug-naïve PD at the time of diagnosis with the patterns of olfactory function in PD-MCI patients using the Korean version of the Sniffin’ stick test II (KVSS II). Methods: A total of 66 drug-naïve PD patients were enrolled. A neuropsychiatric assessment battery and KVSS II were performed. For the statistical analyses, univariate, multivariable linear regression and Student’s t-test were used to determine the relationship between the variables and olfactory function. Results: Olfactory dysfunction was more prevalent in the PD-MCI group than in the PD-normal cognition (PD-CN) group. Each domains of odor threshold, discrimination, identification and total olfactory score were more impaired in the PD-MCI group than the PD-CN group. Whether cognitive impairment was single or multiple domain was not affected.

Conclusion: PD-MCI is more likely to be associated with severe olfactory impairment than PD-CN. There may be more extensive neurodegenerative processes affecting olfaction in PD-MCI patients. With further investigation and validation using neuropathological data, an objective olfactory function test could be used as a tool to evaluate disease progression. Further studies with prospective design investigating the prognostic value of olfactory dysfunction in PD-MCI patients are essential.

1. Introduction

Parkinson’s disease (PD) is the second most common neurode- generative disorder. It affects about 1% of people over 60 years of age with a recently increasing prevalence [1]. Motor symptoms of resting tremor, rigidity, bradykinesia and gait disorder are recog- nized as typical clinical manifestations of PD. Non-motor symptoms

* All authors who described above have contributed significantly, and that all authors are in agreement with the content of the manuscript. Statement of all funding source: None. Any conflict of interest disclosures: None. This study was done in accord with the ethical standards of the Committee on Human Experi- mentation of the institution (Korea University Ethical Review Board) in which the experiments were done or in accord with the Helsinki Declaration of 1975.

* Corresponding author. Department of Neurology, Korea University College of Medicine, Korea University Ansan Hospital, 15355, Gojan-1-dong, Ansan, Gyeonggi- do, South Korea.

E-mail address: kwondoya@korea.edu (D.-Y. Kwon).

https://doi.org/10.1016/j.parkreldis.2017.11.334

such as cognitive impairment, olfactory dysfunction, mood disorder and psychosis are also considered very important features of early PD [2]..

Mild cognitive impairment (MCI) in PD is known somewhat distinct from PD without cognitive impairment [3,4]. In 2012, The Movement Disorder Society established diagnostic criteria for PD- MCI. The criteria revealed the presence of PD-MCI in up to 34% of patients with early-stage PD [5]. PD-MCI may present with different manifestations than MCI evident in the course of Alz- heimer’s disease (AD-MCI) [6]. The most important feature of PD- MCI is its risk of progressing to PD-related dementia (PD-D) [7]. The Norwegian Park West study suggested that more than 25% of patients with MCI at diagnosis of PD developed dementia within 3 years [8]. Secondarily, while AD-MCI is associated with marked impairments in language and memory, PD-MCI features decreased cognition especially in executive function, visuospatial function and attention [9]. PD-MCI may be a heterogeneous condition with a

70 J.-W. Park et al. / Parkinsonism and Related Disorders 46 (2018) 69e73

complicated pathomechanism.
Olfactory dysfunction is a representative non-motor symptom

of PD that is especially evident in the early stage of PD and even used as a marker detecting premotor stages of PD [10]. PD-MCI patients can display greater loss of awareness of hyposmia compared to PD cognitively normal (PD-CN) patients [11]. Thus, subjective measurement of hyposmia in PD-MCI patients could mislead undetection of olfactory dysfunction.

Olfactory dysfunction in PD-MCI patients has not been studied to any extent. In this study, we objectively measured and compared olfactory function between PD-MCI and PD-CN subjects using the Korean version of the Sniffin’ stick test II (KVSS-II) [12,13], and examined the clinical factors affecting the degree of olfactory dysfunction in both groups.

2. Subjects and methods

2.1. Subjects

Patients who visited the movement disorder clinic of a univer- sity affiliated hospital and were diagnosed with PD for the first time (drug-naïve state of PD) according to the clinical diagnostic criteria of the United Kingdom Parkinson’s Disease Society Brain Bank were enrolled [14]. All participating patients provided signed informed consent. The study was approved by the hospital’s ethics commit- tee. The participants underwent neurological examinations by an experienced neurologist specialized in movement disorders. The severity of PD was assessed using the Unified Parkinson’s Disease Rating Scale (UPDRS) [15]. Magnetic resonance imaging (MRI) of the brain was performed on all participants. PD-MCI was diagnosed using the MDS task force criteria [5,16]. All participants were examined by an oto-rhino-laryngologist before evaluation of ol- factory function to exclude the possibility of secondary anosmia due to obstruction and/or sensorineural etiologies [17]. Participants with depression based on criteria from the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) for major depressive disorder (MDD) were excluded because severe depression might affect the results of the examinations. Patients with a Mini-Mental State Examination (MMSE) score less than 24, disease duration over 5 years, current smokers also excluded as smoking significantly affect olfactory function. Focal or diffuse brain lesions on MRI were also excluded to remove other factors that might affect the reliability of the study results.

2.2. Neuropsychological assessment battery

All participants underwent detailed neuropsychological testing by a psychologist. The battery of tests included evaluation of speech and language, memory functioning (verbal, visual), visuoperceptual and visuoconstructional function, frontal and executive functions, as well as the Korean version of MMSE (K-MMSE) and the Korean version of the Montreal Cognitive Assessment (K-MOCA). Korean version neuropsychiatric inventory scale, anxiety, depression scale, apathy scale was also included to evaluate underlying affective symptoms. Based on these results, diagnosis of PD-MCI was con- ducted according to the Movement Disorder Society task force criteria. The PD-MCI patients were categorized into two subgroups: single domain amnestic (only memory function impairment) and multiple domain amnestic (more than one domain of dysfunction) according to previous criteria [18].

2.3. KVSS II

KVSS-II was conducted after excluding patients who had the aforementioned secondary etiologies of anosmia by

otolaryngologist [13]. KVSS II is a validated test instrument for Koreans, which uses odors that are familiar to Koreans [12]. The test consists of each 16 different items. It is performed in the same way as the Sniffin’ stick test. The three components of olfactory function are the olfactory threshold (detecting the lowest concentration of odor), odor discrimination (differentiating certain odors from several other odors), and odor identification (naming a certain odor). The olfactory threshold and odor discrimination tests were carried out using 16 pens that were each filled with a different liquid odorant. For odor discrimination, three pens (two with identical odors) were presented to the patient. The odor identifi- cation test utilized all 16 odors. KVSS II testing was done at the Department of Otolaryngology by a trained specialist. The olfactory threshold test was performed first, followed by the odor discrimi- nation test and the odor identification test, with a break of 3 min between tests. The total score of the three tests (T.D.I score) was interpreted as anosmia (0e20), hyposmia (20.25e27) and normal (27.25e48) [12,13,19].

2.4. Statistical analysis

Comparisons between the two groups for normally distributed continuous data were done using Student’s t-test and one-way analysis of variance. Categorical data was analyzed using the chi- square test. Univariate and multivariable linear regression anal- ysis were performed to identify predictors of total KVSS scores after adjusting for sex, age, disease stage (H&Y), disease severity by UPDRS part III score, disease duration, educational year, K-MMSE score and MOCA score. We further analyzed each component of the olfactory score (threshold, discrimination and identification) and MCI subgroups using multivariable linear regression analysis to determine the variables associated with each component of olfac- tory function as well as the associations between MCI subgroups and olfactory dysfunction. A p-value < 0.05 was considered to be significant. All statistical analyses were performed using SPSS version 20.0 for Windows (SPSS Inc., Chicago, IL, USA).

3. Results

3.1. Baseline and clinical characteristics

Among 106 patients newly diagnosed as PD, 40 were excluded according to the exclusion criteria. The remaining 66 patients ful- filled the study criteria and were included in the final analysis. The baseline characteristics of the patients are summarized in Table 1. The number of PD-CN (n 1⁄4 37) was slight more than PD-MCI pa- tients (n 1⁄4 29). There were no differences in age, sex ratio, disease stage (H&Y), UPDRS score, disease duration, sex ratio, educational years and MMSE between the two groups. Some differences were

Table 1

Baseline characteristics of PD patients.
PD-CN (n 1⁄4 37)

PD-MCI (n 1⁄4 29)

20: 9
64.79 ± 11.42 2.02 ± 0.55 26.59 ± 8.22 2.38 ± 1.40 10.40 ± 3.46 27.03 ± 1.89 22.07 ± 3.57

p-value

0.098 0.995 0.104 0.132 0.998 0.840 0.242 0.036**

Sex (men: women) Age (years)
H&Y stage
UPDRS

Disease duration (years) Education level (years) MMSE
MOCA

18: 19
64.81 ± 12.42 1.81 ± 0.45 23.41 ± 8.3 2.38 ± 1.40 10.62 ± 5.54 27.54 ± 1.54 24.00 ± 3.74

Data 1⁄4 mean ± SD; *significant p < 0.05, sexual difference- Fisher extract. Abbre- viations: H&Y (Hoehn & Yahr); UPDRS, Unified Parkinson’s Disease rating scale; MMSE, Mini-Mental State Examination; MOCA, Montreal Cognitive Assessment; KVSS II, Korean version of Sniffin’ sticks test II.

found between the two groups in MOCA score, and total KVSS II score, as well as each olfactory component (threshold, discrimi- nation and identification) (Table 1).

3.2. Relationship between mild cognitive impairment and olfactory dysfunction

The box plot in Fig. 1 shows the difference in total KVSS II score and olfactory components (threshold, discrimination and identifi- cation) between the PD-CN and PD-MCI groups. The average total KVSS II score and score of each olfactory component was signifi- cantly lower in the PD-MCI group than in the PD-CN group. Uni- variate and multivariate regression analysis revealed that the presence of PD-MCI was the only factor that affected the total KVSS II score, olfactory threshold and odor discrimination (Table 2). In the odor identification model, PD-MCI was also included (p 1⁄4 0.000) but sex (female) wasalso included in the regression model (p 1⁄4 0.013) (Table 2). Adjusted R2 of multivariated regression analysis was 0.12, 0.08, 0.31 and 0.278 respectively for olfactory threshold, odor discrimination, odor identification and total KVSS II score.

4. Discussion

Non-motor symptoms of PD are common and significant for the diagnosis of PD in its early premotor stages, as well as in the pre- diction of symptom progression [20]. These are also important to understand the pathophysiology of PD. Most of the PD patients have olfactory dysfunction as a prodromal symptom [20,21]. However, it is not clinically easy to identify olfactory dysfunction due to PD because there are many other causes of hyposmia in elderly patients, including normal aging, naso-pharyngeal pathol- ogy and other neurodegenerative disorders [17]. In addition, PD-

MCI patients may neglecting their potential olfactory dysfunction [11]. Therefore, olfactory dysfunction in PD-MCI patients should be measured more precisely and objectively than in cognitively normal PD patients, especially in the very early stage of the disease.

Another problem in identifying olfactory dysfunction in PD is that it is difficult to quantify the degree of olfactory dysfunction because individual subjective symptoms do not always match the actual degree of olfactory dysfunction. Many developed methods are currently used which include the University of Pennsylvania smell identification test (UPSIT) [22], Connecticut Chemosensory Clinical Research Center Test [23] and the T & T Olfactometer. However, assessments of olfactory function should be individual- ized based on nationality and cultural difference because of the need to designate familiar odors for the test. Therefore, we used KVSS-II to evaluate olfactory function in the Korean participants.

The KVSS II was developed by modifying the versions of the ‘Sniffin’ sticks’ [19,24] with odors that are more familiar to Korean populations such as soy sauce and sesame oil. Although it is not easy to perform the KVSS II test in clinics, its advantages include its objectivity and its ability to evaluate the three different compo- nents of olfactory function more precisely than other simple methods.

In this study, we compared the KVSS II results between the PD- MCI and PD-CN groups. The findings that the total score and all of the components in the olfactory function test were decreased in the drug-naïve PD-MCI group and regression analysis revealed that the presence of MCI as the only independent predictor of anosmia means the severity of olfactory loss may reflect the degree of un- derlying neurodegeneration in PD disease progression. Baba et al. presented that severe anosmia can be a prognostic marker of developing dementia in PD (PDD) [25]. Having a different olfactory function according to the cognitive status even at the time of PD diagnosis suggest that PD-MCI may already have more

J.-W. Park et al. / Parkinsonism and Related Disorders 46 (2018) 69e73 71

Fig. 1. Box plots of KVSS II test results between PD-CN and PD-MCI.

72 J.-W. Park et al. / Parkinsonism and Related Disorders 46 (2018) 69e73

Table 2

Risk factors associated with components of olfactory dysfunction in PD patients.

Variable

PD-MCI
Sex (Female) Age
H&Y Stage Disease duration

Variable

PD-MCI
Sex (Female) Age
H&Y Stage Disease duration

Olfactory Threshold Univariate
Beta p-value

¡1.603 0.002* 0.398 0.455 0.035 0.112 0.301 0.573 0.038 0.840

Odor Identification Univariate
Beta p-value

¡2.537 0.000* 1.904 0.001* 0.034 0.192 0.925 0.133 0.004 0.986

aMultivariatea
Beta p-value

¡1.632 0.002* 0.014 0.979 0.036 0.091 0.149 0.776 0.046 0.794

aMultivariatec
Beta p-value

¡2.248 0.000* 1.364 0.013* 0.028 0.202 0.057 0.915 0.009 0.961

Odor Discrimination Univariate
Beta p-value

¡2.873 0.002* 0.526 0.580 0.009 0.819 0.079 0.934 0.048 0.888

Total KVSS II score Univariate
Beta p-value

¡7.013 0.000* 2.829 0.080 0.093 0.175 1.146 0.483 0.004 0.995

aMultivariateb
Beta p-value

¡3.019 0.002* 0.055 0.954 0.013 0.733 0.760 0.425 0.051 0.874

aMultivariated
Beta p-value

¡6.913 0.000* 1.400 0.333 0.091 0.122 0.889 0.540 0.030 0.950

*p < 0.05.
a Adjusted R2 for multivariate regression model; a 1⁄4 0.12, b 1⁄4 0.08, c 1⁄4 0.31, d 1⁄4 0.278.

neurodegenerative process across the brain and possibly progress more rapidly as time passes. The most interesting part in our pre- sented study was the highest R2 of multivariate regression model of odor identification compared with olfactory threshold and odor discrimination. As odor identification is most related with higher- order order quality perception [26], including semantic memory, our result highlights that there could be relation between dysfunction in odor identification and presence of MCI in PD based on the spread of pathology.

Measuring olfactory dysfunction in PD-MCI has not been well- studied. A recent study described that olfactory dysfunction pre- dicts early cognitive decline in PD [25], which is similar to the present finding. Although there are no suggestions of generalized profile in olfactory dysfunction in aging and types of PD, it was observed that deficits in odor threshold and identification can be prodromal symptoms in patients with AD [27]. This study sug- gested that specific profiles of olfactory functions may be present in different types of dementia [28]. As the potential areas responsible for olfactory dysfunction in patients with PD include the entorhinal cortex, perirhinal cortex, CA1/subcular area of the hippocampus, amygdala and the association cortices, memory functions may also be significantly related with olfactory processes [29]. Although it remains unknown to detect the specific profile of the olfactory function in neurodegenerative disorders because olfactory function may worsen with age and disease progression, however, our study result that implying the possible relation between odor identifi- cation dysfunction and PD-MCI could be a clue for the future studies.

The study results indicate that odor threshold, discrimination and identification are more impaired in PD-MCI than PD-CN pa- tients, and that PD-MCI is a very important factor predicting ol- factory dysfunction. This might imply that PD-MCI might be a manifestation of a more extensive pathological degeneration state of PD, spreading from peripheral parts of the olfactory system to the central part, regardless of degree of motor phenomenology [30]. However, we believe that it cannot be explained only by the generalized cognitive impairment and disease progression.

Although relatively small sample size, our data shows inter- esting result compared to previous studies of olfactory dysfunction shown in other neurodegenerative studies that smell identification score only reflects dementia conditions [28]. The olfactory system is very complex and involves many parts of the brain circuitry. Furthermore, evidence remains limited. So, fathoming the

underlying meaning of components in olfactory function remains a challenge. However, it is agreed that the olfactory threshold reflects the function of the peripheral olfactory system and odor discrimi- nation, and that odor identification is most related with higher- order odor quality perception, including semantic memory [28,31]. Therefore, different pathophysiological processes may un- derlie PD-CN, PD-MCI and MCI shown in the process of AD regarding the difference found in the olfactory function test and considering basic difference in pathophysiological process seen in PD and AD [21]. As there is no clear established theory yet, more data with prospective designs are needed, including studies that measure olfactory function related with disease progression and disease sub-classification, to clarify the nature of the disease.

Our study has some limitations. These include the relatively small sample size and the absence of follow-up examinations. The inclusion of sexual difference in a regression model of odor iden- tification may reflect the small sample size. Further studies with imaging analysis, more patients, and more follow-up data will contribute to a better understanding of the disease progression of PD-MCI.

Despite these limitations, the present study is notable as it is, to the best of our knowledge, the first report to objectively describe the characteristics of olfactory dysfunction in PD-MCI patients us- ing the detailed smell tests.

To summarize, olfactory dysfunction was associated more with PD-MCI patients than the PD-CN group. Each of the odor threshold, discrimination and identification and total score were more impaired in the PD-MCI group. Whether the cognitive impairment was multiple or single domain did not affect the olfactory function in this study. We believe that objective measurement of olfactory function from an initial diagnosis of PD can be used as a biomarker suggesting the degree of underlying pathology involving congnitive functions as well as olfactory system. There may be more extensive and complex neurodegenerative processes across the central ner- vous system in PD-MCI patients. Elucidating the characteristics of olfactory dysfunction in patients with PD-MCI will expand our knowledge of the pathophysiology of PD and help us to prepare for early intervention in the development of PD dementia.

Authors’ Role

1. Research project:
A. Conception: DY Kwon, MH Park

B. Organization: DY Kwon

C. Execution; HK Yoon, JH Choi, JW Park 2. Statistical Analysis:

A. Design: DY Kwon, JW Park
B. Execution: JW Park, JH Choi
C. Review and Critique; HK Yoon, MH Park

3. Manuscript Preparation:
A. Writing of the first draft: JW Park, DY Kwon B. Review and Critique: HK Yoon, MH Park

References

. [1] W.K.Seo,S.B.Koh,B.J.Kim,S.W.Yu,M.H.Park,K.W.Park,D.H.Lee,Prevalence of Parkinson’s disease in Korea, J. Clin. Neurosci. 14 (2007) 1155e1157.

. [2] M.E. Domellof, K.F. Lundin, M. Edstrom, L. Forsgren, Olfactory dysfunction and
dementia in newly diagnosed patients with Parkinson’s disease, Park. Relat.
Disord. 38 (2017) 41e47.

. [3] G. Santangelo, C. Vitale, M. Picillo, M. Moccia, S. Cuoco, K. Longo, D. Pezzella,
A. di Grazia, R. Erro, M.T. Pellecchia, M. Amboni, L. Trojano, P. Barone, Mild Cognitive Impairment in newly diagnosed Parkinson’s disease: a longitudinal prospective study, Park. Relat. Disord. 21 (2015) 1219e1226.

. [4] A.J. Yarnall, L. Rochester, D.J. Burn, Mild cognitive impairment in Parkinson’s disease, Age Ageing 42 (2013) 567e576.

. [5] I. Litvan, J.G. Goldman, A.I. Troster, B.A. Schmand, D. Weintraub, R.C. Petersen, B. Mollenhauer, C.H. Adler, K. Marder, C.H. Williams-Gray, D. Aarsland, J. Kulisevsky, M.C. Rodriguez-Oroz, D.J. Burn, R.A. Barker, M. Emre, Diagnostic criteria for mild cognitive impairment in Parkinson’s disease: movement Disorder Society Task Force guidelines, Mov. Disord. 27 (2012) 349e356.

. [6] L.M.Besser,I.Litvan,S.E.Monsell,C.Mock,S.Weintraub,X.H.Zhou,W.Kukull, Mild cognitive impairment in Parkinson’s disease versus Alzheimer’s disease, Park. Relat. Disord. 27 (2016) 54e60.

. [7] N.C.Palavra,S.L.Naismith,S.J.Lewis,MildcognitiveimpairmentinParkinson’s disease: a review of current concepts, Neurology Res. Int. 2013 (2013) 576091.

. [8] K.F. Pedersen, J.P. Larsen, O.B. Tysnes, G. Alves, Prognosis of mild cognitive impairment in early Parkinson disease: the Norwegian ParkWest study, JAMA
neurol. 70 (2013) 580e586.

. [9] J.G. Goldman, I. Litvan, Mild cognitive impairment in Parkinson’s disease,
Minerva Medica 102 (2011) 441e459.

. [10] S. Casjens, A. Eckert, D. Woitalla, G. Ellrichmann, M. Turewicz, C. Stephan,
M. Eisenacher, C. May, H.E. Meyer, T. Bruning, B. Pesch, Diagnostic value of the
impairment of olfaction in Parkinson’s disease, PLoS One 8 (2013) e64735.

. [11] I. Kawasaki, T. Baba, A. Takeda, E. Mori, Loss of awareness of hyposmia is associated with mild cognitive impairment in Parkinson’s disease, Park. Relat.
Disord. 22 (2016) 74e79.

. [12] J.H. Cho, Y.S. Jeong, Y.J. Lee, S.C. Hong, J.H. Yoon, J.K. Kim, The Korean version
of the Sniffin’ stick (KVSS) test and its validity in comparison with the cross- cultural smell identification test (CC-SIT), Auris Nasus Larynx 36 (2009) 280e286.

. [13] K.E.S.e.a Hong S.C, Y.S. Yoo, Development of KVSS test (Korean version of Sniffin’ sticks test), Korean J. Otolaryngol. 42 (1999) 855e860.

. [14] A.J. Hughes, S.E. Daniel, L. Kilford, A.J. Lees, Accuracy of clinical diagnosis of

[15] [16]

[17] [18]

[19]

[20] [21] [22]

[23]

[24] [25]

[26]

[27]

[28] [29]

[30] [31]

idiopathic Parkinson’s disease: a clinico-pathological study of 100 cases, J. Neurol. Neurosurg. Psychiatry 55 (1992) 181e184.
P. Martinez-Martin, A. Gil-Nagel, L.M. Gracia, J.B. Gomez, J. Martinez-Sarries, F. Bermejo, Unified Parkinson’s disease rating scale characteristics and structure. The cooperative multicentric group, Mov. Disord. 9 (1994) 76e83. J.Y. Szeto, L. Mowszowski, M. Gilat, C.C. Walton, S.L. Naismith, S.J. Lewis, Assessing the utility of the Movement Disorder Society Task Force Level 1 diagnostic criteria for mild cognitive impairment in Parkinson’s disease, Park. Relat. Disord. 21 (2015) 31e35.

J.M. Pinto, Olfaction, Proc. Am. Thorac. Soc. 8 (2011) 46e52.
C.C. Janvin, J.P. Larsen, D. Aarsland, K. Hugdahl, Subtypes of mild cognitive impairment in Parkinson’s disease: progression to dementia, Mov. Disord. 21 (2006) 1343e1349.
S.M. Hong, I.H. Park, K.M. Kim, J.M. Shin, H.M. Lee, Relationship between the Korean version of the Sniffin’ stick test and the T&T olfactometer in the Korean population, Clin. Exp. Otorhinolaryngol. 4 (2011) 184e187.
D.B. Miller, J.P. O’Callaghan, Biomarkers of Parkinson’s disease: present and future, Metabolism 64 (2015) S40eS46. K.Hoyles,J.C.Sharma,Olfactorylossasasupportingfeatureinthediagnosisof Parkinson’s disease: a pragmatic approach, J. Neurol. 260 (2013) 2951e2958. S. Vengalil, J.B. Agadi, K. Raghavendra, University of Pennsylvania smell identification test abnormalities in Parkinson’s disease, J. Assoc. Physicians India 64 (2016) 32e36.
B. Veyseller, B. Ozucer, A.B. Karaaltin, Y. Yildirim, N. Degirmenci, F. Aksoy, O. Ozturan, Connecticut (CCCRC) olfactory test: normative values in 426 healthy volunteers, Indian J. otolaryngology head neck Surg. Official Publ. Assoc. Otolaryngologists India 66 (2014) 31e34.
Y.-s.Y. Seok-chan Hong, Eun-Seo Kim, Sok-Chon Kim, Soo-Hong Park, Jin- Kuk Kim, Seong-Ho Kang, Development of KVSS test (Korean version of Sniffin’Sticks test), Korean J. Otolaryngol. 42 (1999) 855e860.
T. Baba, A. Kikuchi, K. Hirayama, Y. Nishio, Y. Hosokai, S. Kanno, T. Hasegawa, N. Sugeno, M. Konno, K. Suzuki, S. Takahashi, H. Fukuda, M. Aoki, Y. Itoyama, E. Mori, A. Takeda, Severe olfactory dysfunction is a prodromal symptom of dementia associated with Parkinson’s disease: a 3 year longitudinal study, Brain 135 (2012) 161e169.
M. Serrao, A. Ranavolo, C. Conte, C. Davassi, S. Mari, A. Fasano, G. Chini, G. Coppola, F. Draicchio, F. Pierelli, Effect of 24-h continuous rotigotine treatment on stationary and non-stationary locomotion in de novo patients with Parkinson disease in an open-label uncontrolled study, J. neurology 262 (2015) 2539e2547. J.Djordjevic,M.Jones-Gotman,K.DeSousa,H.Chertkow,Olfactioninpatients with mild cognitive impairment and Alzheimer’s disease, Neurobiol. Aging 29 (2008) 693e706.
J. Alves, A. Petrosyan, R. Magalhaes, Olfactory dysfunction in dementia, World J. Clin. Cases 2 (2014) 661e667.
W. Silverman, H.M. Wisniewski, M. Bobinski, J. Wegiel, Frequency of stages of Alzheimer-related lesions in different age categories, Neurobiol. Aging 18 (1997) 389e392, 377-379; discussion. H.Braak,K.DelTredici,U.Rub,R.A.deVos,E.N.JansenSteur,E.Braak,Staging of brain pathology related to sporadic Parkinson’s disease, Neurobiol. Aging 24 (2003) 197e211.
C. Huart, P. Rombaux, T. Hummel, Plasticity of the Human Olfactory System: the Olfactory Bulb, Molecules (Basel, Switzerland), vol. 18, 2013, pp. 11586e11600.

J.-W. Park et al. / Parkinsonism and Related Disorders 46 (2018) 69e73 73

Parkinsonism and Related Disorders 46 (2018) 74e78

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Short communication

Microstructural white matter abnormalities in patients with COL6A3 mutations (DYT27 dystonia)

Angela Jochim a, 1, Yong Li a, 1, Michael Zech a, b, Daniel Lam b, Nadine Gross a, Kathrin Koch c, Claus Zimmer c, Juliane Winkelmann a, b, d, e, Bernhard Haslinger a, *

a Department of Neurology, Klinikum Rechts der Isar, Technische Universita€t München, Ismaninger Strasse 22, 81675 München, Germany
b Institut für Neurogenomik, Helmholtz Zentrum München, Ingoldsta€dter Landstrasse 1, 85764 Neuherberg, Germany
c Department of Neuroradiology, Klinikum Rechts der Isar, Technische Universita€t München, Ismaninger Strasse 22, 81675 München, Germany d Munich Cluster for Systems Neurology, SyNergy, Feodor-Lynen-Strasse 17, 81377 München, Germany
e Institut für Humangenetik, Technische Universita€t München, Munich, Germany

articleinfo abstract

Article history:

Received 10 May 2017 Received in revised form 22 August 2017
Accepted 12 October 2017

Keywords:

Diffusion tensor imaging Fractional anisotropy Cerebello-thalamic tract COL6A3 associated dystonia DYT27 dystonia

Introduction: Recently, mutations in the collagen gene COL6A3 have been reported in patients with autosomal-recessive, isolated dystonia (DYT27). Zebrafish models of COL6A3 mutations showed deficits in axonal targeting mechanisms. Therefore, COL6A3 mutations have been considered to contribute to irregular sensorimotor circuit formation. To test this hypothesis, we examined structural abnormalities in cerebral fiber tracts of dystonia patients with COL6A3 mutations using diffusion tensor imaging. Methods: We performed a voxel-wise statistical analysis to compare fractional anisotropy within whole- brain white matter in four of the previously reported dystonia patients with COL6A3 mutations and 12 healthy controls. Region of interests-based probabilistic tractography was performed as a post-hoc- analysis.

Results: Dystonia patients with COL6A3 mutations showed significantly decreased fractional anisotropy bilaterally in midbrain, pons, cerebellar peduncles, thalamus, internal capsule and in frontal and parietal subcortical regions compared to healthy controls. Tractography revealed a decreased fractional anisot- ropy in patients with COL6A3-associated dystonia between bilateral dentate nucleus and thalamus. Conclusion: Diffusion tensor imaging demonstrates an altered white matter structure especially in various parts of the cerebello-thalamo-cortical network in dystonia patients with COL6A3 mutations. This suggests that COL6A3 mutations could contribute to abnormal circuit formation as potential basis of dystonia.

1. Introduction

To date, mutations in the gene encoding the a3-subunit of collagen VI, COL6A3, have been shown to cause autosomal- recessive isolated dystonia (DYT27) in five patients from three families [1]. The clinical phenotype ranges from mild focal cervical dystonia (CD) to severe generalized dystonia with orofacial, laryn- geal, cervical, upper limb and trunk involvement [2]. A confirma- tion of the pathogenicity of COL6A3 mutations in dystonia by other groups is still pending [3].

* Corresponding author. Neurologische Klinik und Poliklinik, Klinikum rechts der Isar, Technische Universita€t München, Ismaninger Str. 22, 81675 Munich, Germany.

E-mail address: bernhard.haslinger@tum.de (B. Haslinger). 1 These authors contributed equally.

https://doi.org/10.1016/j.parkreldis.2017.10.008

Col6a3 was expressed in neurons of adult mouse brains including striatum, cerebellum, midbrain and brainstem [1]. A zebrafish model with knockdown of the human COL6A3 exon 41- ortholog showed deficits in axonal targeting mechanisms, point- ing to a possible impact of COL6A3 mutations on sensorimotor circuit formation [1]. This might have an influence on the integrity of cerebral fiber tracts in individuals with COL6A3 mutations, which can be studied in vivo by diffusion tensor imaging (DTI).

DTI studies have contributed substantially to the present concept of hereditary dystonia as being caused by abnormal for- mation of circuits linking the sensorimotor cortex, thalamus, basal ganglia, brainstem and cerebellum [4]. The relevance of cerebello- thalamic pathways for the pathophysiology of hereditary dystonia seems consistent with the above mentioned Col6a3 expression in mouse brains [1]. To verify the link between mutations in COL6A3

and neuronal fiber tract changes, our DTI study aimed at examining in vivo possible alterations of white matter and fiber tract charac- teristics in four of the five so far identified dystonia patients with COL6A3 mutations.

2. Methods

2.1. Subjects

Four of our five identified patients with COL6A3 mutations participated in the study (Supplementary Table S1). One patient suffered from segmental dystonia with focal hand and cervical dystonia. Three patients had generalized dystonia, two with upper limb, trunk and cranio-cervical dystonia. Details about genetic findings and the heterogeneous clinical phenotype have been described elsewhere [1,2].

12 healthy controls (HCs) were matched in age and gender 3:1 to the patients (Supplementary Table S1). Three patients had been pre-treated with abobotulinumtoxinA (BTX) in cervical muscles. To minimize movement artefacts, MRI scans were performed when BTX was efficacious (Supplementary Table S1).

2.2. Image acquisition and image preprocessing

MRI scans were acquired on a Philips-Achieva 3T-scanner with an eight-channel head coil. Diffusion-weighted MRI (DWI) se- quences included 64 gradient directions for each subject (Supplement 2). To avoid head movement artefacts due to long MRI scan duration, image acquisition was distributed over two separate runs, each resulting in 32 DWIs and 1 non-diffusion-weighted image (b-value 1⁄4 0 s/mm2, average of six single b0 images). 3D- T1- and T2-FLAIR-scans were performed to exclude macroscopic anatomical pathologies.

Diffusion MR images were pre-processed with Explor- eDTI,v4.8.3 (Supplement 2). After raw data inspection, DWIs were corrected for head motion and eddy current-induced geometric distortions. DWIs were aligned to the first b0-images of the first run by applying an affine co-registration method (Supplement 2).

2.3. Tract-based spatial statistics (TBSS)

Further DWI processing including brain extraction, calculation of fractional anisotropy (FA) images, their registration to a standard Montreal Neurological Institute (MNI) space using nonlinear registration, creation of a mean FA skeleton and registration of each subject’s FA images to the skeleton as well as whole-brain voxel- wise statistical analyses using TBSS of the white matter was con- ducted using FSL (Supplement 2). A two-sample t-test was applied to compare FA contrast maps of patients and HCs. Group differences in the voxel-wise statistical analysis of skeletonized FA data were considered significant at a threshold of p < 0.05 family wise error (FWE)-corrected.

2.4. Region of interest (ROI)-based tractography

A ROI-based probabilistic fiber tractography analysis in pa- tients and HCs served as post hoc-analysis (ExploreDTI,v483, Supplement 2). All fiber tracts were reconstructed for each participant and considering the DTI literature in dystonia, a hypothesis-driven ROI-based analysis was focused on cerebello- thalamo-cortical and cerebello-pallidal tracts [4,5]. We extracted fiber tract segments between the following pairs of ROIs: Thala- mus_left (L)ePrecentral_L, Thalamus_right (R)ePrecentral_R, Thalamus_LePostcentral_L, Thalamus_RePostcentral_R (Supple- mental Fig. S1 and Supplement 2). Because of difficulties in

reconstructing fibers in the decussation area of the superior cerebellar peduncle, we made a combined analysis of bilateral tracts between dentate and thalamus (DentateeThalamus_L&R, BTDT) and between dentate and pallidum (Dentate- Pallidum_L&R).

Mean FA values were calculated for all extracted fiber segments (Supplement 2). A 2-sample-t test model was employed to reveal significant mean FA differences at group level with p < 0.0083 (Bonferroni-correction for multiple comparisons for 6 pairs of ROIs).

Furthermore, an along-tract analysis was applied to the fiber tract segments that showed significant differences in the ROI-based tractography (Supplement 2). It focuses on stepwise property changes of a fiber tract along the trajectory and therefore provides more details on connectivity changes than average values of the whole tract length. The selected fiber tract segments were divided into a predefined number of fragments (K), which was computed by the length of the fiber tracts segments divided by the voxel-size length of DWI images (2 mm). We labeled these tract fragments according to their position in the tract segment. A 2-sample-t test model estimated group differences for each fragment at p < 0.05/K (Bonferroni-correction for multiple comparisons).

One patient (F3-II-3) had undergone a surgery, most likely thalamotomy, at age 24. In order to estimate its impact, we repeated the tractography after exclusion of this patient and his matched HCs.

3. Results

3.1. Anatomical scans

There were no macroscopic anomalies that could affect the DTI results, except for a small gliosis in the patient who had undergone right hemispheric stereotactic surgery. No subject had to be excluded because of head motion artefacts.

3.2. FA differences between patients and HCs

The patients’ group revealed bilaterally decreased FA (p < 0.05, FWE-corrected) compared to matched HCs in superior, inferior and middle cerebellar peduncles, pons, midbrain, cerebral peduncles, thalamus, internal capsule and subcortical white matter close to gray matter in the frontal and parietal cortex (Fig. 1).

3.3. ROI-based tractography

Patients showed significant FA decreases within BTDT (p 1⁄4 0.0032), whereas no significant differences were found for other tract segments. An along-tract analysis for the BTDT (average tract length: 30 ± 1.24 mm, resulting in 15 (1⁄430mm/2 mm) tract positions (Fig. 2A), corrected threshold of p < 0.0033 (1⁄40,05/15)) showed the most profound FA reduction in patients in the middle four BTDT positions 6 to 9 (p 1⁄4 0.0036e0.0055, Fig. 2A and B), though without passing the Bonferroni-correction for multiple comparisons (p < 0.0033). These findings were consistent with regions of reduced FA in the TBSS analysis (Fig. 2C). Fig. 2D shows scatterplots of FA value distributions over the 15 tract positions for each subject.

Post-hoc tractography after exclusion of the post-surgical pa- tient and his matched HCs demonstrated again a significant FA reduction of BTDT (p 1⁄4 0.00043) and additionally now a significant FA decrease for four positions in the along-tract-analysis, Bonfer- roni-corrected for multiple comparisons (position (P)6: p 1⁄4 0.00219, P7: p 1⁄4 0.00085, P9: p 1⁄4 0.00005, P10: p 1⁄4 0.00149).

A. Jochim et al. / Parkinsonism and Related Disorders 46 (2018) 74e78 75

76 A. Jochim et al. / Parkinsonism and Related Disorders 46 (2018) 74e78

Fig. 1. FA decrease in patients compared to healthy controls.
Representative sagittal (A, D), coronal (B) and axial (C, E, F) sections visualize significant FA decreases in patients in comparison to healthy controls in the midbrain (A), cerebellum (B), internal capsule (C), pons (D), thalamus (E) and frontoparietal subcortical white matter (F). The significance of FA value differences between both groups is color-coded: Increasing significance of FA value differences is color-coded from red (significance threshold p < 0.05, FWE-corrected) to yellow (significance threshold p < 0.01, FWE-cor- rected) and overlaid onto a white matter skeleton of all subjects (green).
Abbreviations: FA: functional anisotropy, FWE: family wise error.

4. Discussion

We show that dystonia patients with COL6A36 mutations display changes of white matter architecture. These are distributed over various stations of the cerebello-thalamo-cortical network, i.e. the cerebellar peduncles, pons, midbrain, cerebral peduncles, as well as the thalamus, internal capsule and frontal and parietal subcortical white matter regions. Post-hoc ROI-based tractography further underlines the affection of cerebello-thalamic tracts.

FA is a correlate of restricted diffusivity, e.g. in fiber tracts, which can be decreased by conditions as degeneration, injure or impaired development, therefore it indicates changes of axonal density and integrity [4]. Tractography allows to characterize properties of fiber tracts between defined ROIs and therefore provides a more detailed information than TBSS.

The involvement of cerebello-thalamo-cortical tracts in patients with COL6A3 mutations is partly comparable to previous DTI find- ings in dystonia with TOR1A and THAP1 mutations. These showed an affection of cerebello-thalamic fiber tracts in the brainstem near the superior cerebellar peduncle in manifesting and non- manifesting gene mutation carriers [4].

Our study also showed differences in thalamo-cortical parts of the cerebello-thalamo-cortical pathway in patients with COL6A3 mutations. This is consistent with the afore mentioned study on manifesting dystonia with TOR1A or THAP1 mutations (in Ref. [4]), but differs from the “tandem lesion model” of recent DTI studies:

Additional to changes in cerebello-thalamic connections, micro- structural integrity of distal thalamo-cortical projections was pre- served in manifesting, but affected in non-manifesting TOR1A or THAP1 mutation carriers, proposing a protective effect of the distal lesion against phenotypic manifestation of dystonia [4,6]. An alteration of the recently described cerebello-pallidal connection [5] was not detected in our study.

Functionally, a microstructural alteration of cerebello-thalamo- cortical fiber tracts might lead to abnormal interactions of cere- bellar and basal ganglia networks via cortico-striato-pallido- thalamic and cerebello-thalamo-cortical pathways [6]. Reduced white matter connectivity of the cerebellar outflow tract was correlated with increased activation in the sensorimotor cortex in subjects with TOR1A or THAP1 mutations [4]. This points to an as- sociation of impaired cerebellar pathways with loss of cortical in- hibition, a main pathophysiological trait in dystonia [4]. The cerebellum is relevant for the prediction of sensory consequences of motor commands [7]. Abnormalities within cerebello-thalamo- cortical tracts may therefore contribute to impaired sensorimotor integration and maladaptive neuronal plasticity in dystonia [8]. Besides, our findings fit in well with the recent “system-level view” of multiple interconnections between cerebellum and cortex via basal ganglia [7] and in the emerging acknowledgement of a rele- vant role of the cerebellum in dystonia, either causal, contributory or compensatory [9]. While the interaction of cerebellum and basal ganglia was previously thought to take place mainly on the cortical

A. Jochim et al. / Parkinsonism and Related Disorders 46 (2018) 74e78 77

Fig. 2. Along-tract analysis showing stepwise FA changes in the bilateral fiber segment between thalamus and dentate.
A: The upper figure shows FA changes for patients and healthy controls along the 15 positions of the bilateral tract and the lower figure demonstrates the statistical comparison of the FA values along these 15 positions. The red boxes visualize the four parts of the tract with the most significant difference between both groups, in B projected onto the fiber tract between dentate nucleus and thalamus of a healthy control, in C projected onto the plot of the TBSS results.
D: Scatter plots of the along-tract analysis in the bilateral fiber segment between thalamus and dentate showing FA values of each single subject. The positions 1 to 15 reflect the 15 segments of the bilateral tract between thalamus and dentate. FA values of healthy subjects are indicated by triangles and FA values of patients by COL6A3 mutations with circles. The FA values of patient F3-II-3 who had undergone surgery (most likely thalamotomy) are specified by empty circles.
Abbreviations: FA: functional anisotropy, TBSS: tract-based spatial statistics.

level, studies on primates showed more direct pathways between cerebellum and basal ganglia (including dentate, thalamus and striatum), whose dysfunction might be relevant for the develop- ment of dystonia.

The clinical phenotype of COL6A3-associated dystonia includes focal manifestations like CD, writer’s cramp and spasmodic dysphonia. Some previous findings on these sporadic focal dysto- nias were comparable to our results. FA changes in basal ganglia, prefrontal and supplementary motor cortex were found in CD [10]. In spasmodic dysphonia, FA was decreased in the internal capsule associated with increased water diffusivity along the corticobulbar/-spinal tract [11]. In writer’s cramp, increased FA was found bilaterally in the posterior the internal capsule and thalamus [12].

From our first imaging study on COL6A3-associated dystonia, we cannot interpret our results as pathognomonic for generalized, especially COL6A3-associated dystonia. While hereditary general- ized types of dystonia with TOR1A, THAP1 and also COL6A3 muta- tions are associated with more widespread tract abnormalities, the findings in sporadic late-onset focal or segmental dystonia forms are more inconsistent. This might arise from the different etiolog- ical factors and the more distinct clinical phenotype. Environ- mental factors with secondary neuroplasticity could be more relevant in sporadic focal than in hereditary dystonia, the latter being assumed to partly result from primary abnormal neuroplasticity.

The main limitation of our study is the small number of currently identified patients with COL6A3 mutations. One of our patients had undergone a right hemispheric stereotaxy years ago. Effects on the white matter microstructure as a consequence have to be assumed. However, this is unlikely to have a relevant impact on our results. Otherwise, a right-hemispheric dominance of FA decreases should be expected. Furthermore, post hoc-tractography after exclusion of this patient confirmed the results for the cerebello-thalamic tract and even led to significance for several fiber tract segments in the along-tract analysis. The facts that the

reported cerebello-thalamo-cortical network abnormalities emerged in spite of the small number of patients and that they are in congruence with the previous findings in animal models, might even emphasize their pathogenetic relevance. This might be taken as an additional indirect confirmation of the relevance of COL6A3 mutations in dystonia, which has been raised in question, as it has not been confirmed yet by other groups [3].

Methodological limitations of DTI and tractography, like partial volume effects, low spatial resolution, and the possibility to encounter both false negative and false positive results have to be acknowledged.

In summary, the localization of our in vivo-DTI findings of ab- normalities of cerebral white matter architecture in the human brain is congruent with the detection of Col6a3 expression in pons, midbrain and cerebellar regions in the mouse brain [1]. Considering the role of collagen VI as an ECM component, which is probably relevant for the construction and maintenance of neural circuits, and the zebrafish findings of impaired axonal targeting mecha- nisms due to col6a3 suppression, the microstructural abnormalities of cerebello-thalamo-cortical pathways in patients with COL6A3 mutations suggest some causal alteration of white matter circuits in this newly identified hereditary dystonia.

Ethical approval

The study has been approved by the local ethics review board, and written informed consent according to the Declaration of Helsinki was obtained from all subjects.

Funding sources

There has been no funding of the study.

Conflicts of interest
Dr. Jochim has received travel grants by The International

78 A. Jochim et al. / Parkinsonism and Related Disorders 46 (2018) 74e78

Parkinson and Movement Disorder Society, Boston Scientific, Ipsen Pharma GmbH, Merz Pharmaceuticals GmbH, Pharm-Allergan GmbH and AbbVie Deutschland GmbH & Co. KG as well as speaker honoraria by Pharm-Allergan GmbH. Dr. Li has no financial disclosures. Dr. Zech has received travel expenses from Actelion Pharmaceuticals Ltd. And Pharm-Allergan GmbH, has received intramural funding from the Langmatz-Stiftung and research sup- port from the Else Kro€ner-Fresenius-Stitung. Dr. Lam has no financial disclosures. Dr. Grob has no financial disclosures. Dr. Koch has no financial disclosures. Prof. Dr. Zimmer has served on scien- tific advisory boards for Philips and Bayer Schering; serves as co- editor on the Advisory Board of Clinical Neuroradiology; has received speaker honoraria from Bayer-Schering and Philips and has received research support and investigator fees for clinical studies from Biogen Idec, Quintiles, MSD Sharp & Dome, Boeh- ringer Ingelheim, Inventive Health Clinical UK Ltd., Advance Cor, Brainsgate, Pfizer, Bayer-Schering, Novartis, Roche, Servier, Pen- umbra, WCT GmbH, Syngis, SSS International Clinical Research, PPD Germany GmbH, Worldwide Clinical Trials Ltd., Phenox, Covidien, Actelion, Medivation, Medtronic, Harrison Clinical Research, Concentric, Penumbra, Pharmtrace, Reverse Medical Corp., Premier Research Germany Ltd., Surpass Medical Ltd. and GlaxoSmithKline. Prof. Dr. Winkelmann has received speaker honoraria from UCB and receives research support from the Else Kro€ner-Fresenius-Stiftung. Prof. Dr. Haslinger receives research support from Ipsen Pharma GmbH and DFG, has received speaker honoraria from Pharm- Allergan GmbH and Ipsen Pharma GmbH and has received travel grants from Ipsen Pharma GmbH.

Acknowledgements

We are very grateful to Silke Zwirner (Technische Universita€t München) for her help in study organization as well as in language editing of this manuscript. Furthermore, we cordially thank all subjects who took part in this study as well as the Deutsche Dys- tonie Gesellschaft e.V. for their support.

Appendix A. Supplementary data

Supplementary data related to this article can be found at

https://doi.org/10.1016/j.parkreldis.2017.10.008. References

. [1] M. Zech, D.D. Lam, L. Francescatto, B. Schormair, A.V. Salminen, A. Jochim, T. Wieland, P. Lichtner, A. Peters, C. Gieger, H. Lochmuller, T.M. Strom, B. Haslinger, N. Katsanis, J. Winkelmann, Recessive mutations in the alpha3 (VI) collagen gene COL6A3 cause early-onset isolated dystonia, Am. J. Hum. Genet. 96 (6) (2015) 883e893.

. [2] A.Jochim,M.Zech,G.Gora-Stahlberg,J.Winkelmann,B.Haslinger,Theclinical phenotype of early-onset isolated dystonia caused by recessive COL6A3 mu- tations (DYT27), Mov. Disord. 31 (5) (2016) 747e750.

. [3] K. Lohmann, C. Klein, Update on the genetics of dystonia, Curr. Neurol. Neu- rosci. Rep. 17 (3) (2017) 26.

. [4] M. Niethammer, M. Carbon, M. Argyelan, D. Eidelberg, Hereditary dystonia as a neurodevelopmental circuit disorder: evidence from neuroimaging, Neu- robiol. Dis. 42 (2) (2011) 202e209.

. [5] D. Milardi, A. Arrigo, G. Anastasi, A. Cacciola, S. Marino, E. Mormina, A. Calamuneri, D. Bruschetta, G. Cutroneo, F. Trimarchi, A. Quartarone, Extensive direct subcortical cerebellum-basal ganglia connections in human brain as revealed by constrained spherical deconvolution tractography, Front. Neuroanat. 10 (2016) 29.

. [6] A. Vo, W. Sako, M. Niethammer, M. Carbon, S.B. Bressman, A.M. Ulug, D. Eidelberg, Thalamocortical connectivity correlates with phenotypic vari- ability in dystonia, Cereb. Cortex 25 (9) (2015) 3086e3094.

. [7] D. Caligiore, G. Pezzulo, G. Baldassarre, A.C. Bostan, P.L. Strick, K. Doya, R.C. Helmich, M. Dirkx, J. Houk, H. Jorntell, A. Lago-Rodriguez, J.M. Galea, R.C. Miall, T. Popa, A. Kishore, P.F. Verschure, R. Zucca, I. Herreros, Consensus paper: towards a systems-level view of cerebellar function: the interplay between cerebellum, Basal Ganglia, Cortex, Cerebellum 16 (1) (2017) 203e229.

. [8] A. Quartarone, M. Hallett, Emerging concepts in the physiological basis of dystonia, Mov. Disord. 28 (7) (2013) 958e967.

. [9] V.G. Shakkottai, A. Batla, K. Bhatia, W.T. Dauer, C. Dresel, M. Niethammer, D. Eidelberg, R.S. Raike, Y. Smith, H.A. Jinnah, E.J. Hess, S. Meunier, M. Hallett, R. Fremont, K. Khodakhah, M.S. LeDoux, T. Popa, C. Gallea, S. Lehericy, A.C. Bostan, P.L. Strick, Current opinions and areas of consensus on the role of the cerebellum in dystonia, Cerebellum 16 (2) (2017) 577e594.

. [10] G. Fabbrini, P. Pantano, P. Totaro, V. Calistri, C. Colosimo, M. Carmellini, G. Defazio, A. Berardelli, Diffusion tensor imaging in patients with primary cervical dystonia and in patients with blepharospasm, Eur. J. Neurol. 15 (2) (2008) 185e189.

. [11] K. Simonyan, F. Tovar-Moll, J. Ostuni, M. Hallett, V.F. Kalasinsky, M.R. Lewin- Smith, E.J. Rushing, A.O. Vortmeyer, C.L. Ludlow, Focal white matter changes in spasmodic dysphonia: a combined diffusion tensor imaging and neuro- pathological study, Brain 131 (Pt 2) (2008) 447e459.

. [12] C. Delmaire, M. Vidailhet, D. Wassermann, M. Descoteaux, R. Valabregue, F. Bourdain, C. Lenglet, S. Sangla, A. Terrier, R. Deriche, S. Lehericy, Diffusion abnormalities in the primary sensorimotor pathways in writer’s cramp, Arch. Neurol. 66 (4) (2009) 502e508.

Parkinsonism and Related Disorders 46 (2018) 79e83

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Short communication

Spastic paraplegia type 31: A novel REEP1 splice site donor variant and expansion of the phenotype variability

Masaki Kamada a, 1, Toshitaka Kawarai b, *, 1, Ryosuke Miyamoto b, Rie Kawakita c, Yuki Tojima d, Celeste Montecchiani e, f, Laura D’Onofrio e, Carlo Caltagirone g, h, Antonio Orlacchio e, f, Ryuji Kaji b

a Department of Neurological Intractable Disease Research, Kagawa University Faculty of Medicine, Kagawa, 761-0793, Japan
b Department of Clinical Neuroscience, Institute of Biomedical Sciences, Tokushima University Graduate School, Tokushima, 770-0042, Japan
c Department of Neurology, Kagawa University Hospital, Kagawa, 761-0793, Japan
d Faculty of Medicine, Tokushima University, Tokushima 770-0042, Japan
e Laboratorio di Neurogenetica, Centro Europeo di Ricerca sul Cervello (CERC) – Istituto di Ricovero e Cura a Carattere Scientifico (IRCCS) Santa Lucia, Rome, Italy
f Dipartimento di Scienze Chirurgiche e Biomediche, Universita di Perugia, Perugia, Italy
g Laboratorio di Neurologia Clinica e Comportamentale, IRCCS Santa Lucia, Rome, Italy
h Dipartimento di Medicina dei Sistemi, Universita di Roma “Tor Vergata”, Rome, Italy

articleinfo abstract

Article history:

Received 18 April 2017 Received in revised form 28 September 2017 Accepted 18 October 2017

Keywords:

Spastic paraplegia type 31 (SPG31) Receptor expression-enhancing protein 1 (REEP1)
Splice site donor variant Nonsense-mediated mRNA decay (NMD) Phenotype variability
Asymptomatic carrier

1. Introduction

Hereditary spastic paraplegia (HSP) is a clinically and genetically heterogeneous group of neurodegenerative disorders, mainly affecting corticospinal motor axons in a length-dependent manner. The main clinical features include progressive spasticity and

* Corresponding author: Department of Clinical Neuroscience, Institute of Biomedical Sciences, Tokushima University Graduate School, 3-18-15, Kuramoto- cho, Tokushima city 770-0042, Japan.

E-mail address: tkawarai@tokushima-u.ac.jp (T. Kawarai). 1 Contributed equally to this work.

https://doi.org/10.1016/j.parkreldis.2017.10.012

Mutations in REEP1 have been identified in three types of neurological disorders, autosomal dominant form of Hereditary Spastic Paraplegia type 31 (SPG31), autosomal dominant distal hereditary motor neuronopathy type VB (HMN5B), and autosomal recessive form of congenital axonal neuropathy and diaphragmatic palsy. Previous studies demonstrated different molecular pathogenesis in SPG31, including loss-of-function, gain-of-function and haploinsufficiency. A four-generation family from Japan, including 12 members, was investigated clinically and genetically. Seven affected members displayed pure spastic paraplegia. Impression of genetic anticipation was observed in the family, including ten- dency of earlier age-at-onset and increasing severity in subsequent generations. Genetic analysis revealed a heterozygous intronic variant, c.303þ2T > A, in REEP1, which segregated with disease, and was also identified in one unaffected member. The variant causes exon 4 skipping leading to frame shift and a truncated transcript identified by complementary DNA sequencing of reverse transcription poly- merase chain reaction products. Measurement of REEP1 transcripts in lymphocytes demonstrated a reduction through nonsense mediated mRNA decay (NMD). Our study demonstrated further evidence of allelic heterogeneity in SPG31, mutant REEP1 mRNA dosage effects through NMD and intra-familial phenotype variability.

weakness of the lower limbs. HSP can be clinically divided into two categories: pure form and complicated form. The former includes the main features only, and the latter includes additional and more extensive neurological or non-neurological features. More than 67 spastic paraplegia genes (SPGs) have been reported to date, which can be transmitted in an autosomal dominant (AD), autosomal recessive (AR), X-linked (XL), or mitochondrial manner [1].

Gene encoding Receptor Expression-Enhancing Protein 1 (REEP1) is heterozygously mutated in the autosomal dominant form of spastic paraplegia type 31 (SPG31; OMIM 610250) [2] and distal hereditary motor neuronopathy type VB (HMN5B; OMIM 614751) [3]. A homozygous REEP1 mutation was identified in single

80 M. Kamada et al. / Parkinsonism and Related Disorders 46 (2018) 79e83

patient with congenital axonal neuropathy and diaphragmatic palsy [4]. To date, over 40 mutations in REEP1 have been reported in SPG31, and most of them are frame shift mutations [5]. Considering genotype-phenotype correlations and presence of asymptomatic carrier, molecular pathogenesis of REEP1-related disorders is not due to simple haploinsufficiency. A novel splice site donor variant in REEP1, associated with intra-familial phenotypic variability, is here described.

2. Methods

Single Japanese family, including seven members affected with spastic paraplegia, was studied (Fig. 1A). All living members were clinically and genetically investigated. Information of the deceased affected member was obtained from interviews with the living family members. Molecular genetic study of the family was per- formed by whole exome sequencing (WES), PCR-direct sequencing,

bioinformatic analyses, and RNA transcript analysis in cultured T cells. Exome capture, high-throughput sequencing, filtering under genetic models and annotation were conducted as described in Supplementary Material. All coding exons and flanking intronic regions (greater than 20 base pairs from an exon) in the currently- known genes associated with spastic paraplegia were analysed. Genetic variations in the candidate genes were extracted by refer- ring to public databases, including the dbSNP Build 141 database, the 1000 Genomes Project, ESP5400 exomes, and ExAC. Details are available upon request. The intronic variant in REEP1, c.303þ2T > A, was validated by PCR-direct sequencing analysis as described in Supplementary Material. Genetic screening of the variant was extended in the control datasets, consisting of 109 healthy Japanese (female 56.0%, 61.8 ± 7.8 years, range 44e79) and 200 healthy Italian (female 52.5%, 46.2 ± 8.7 years, range 30e62) control subjects.

The intronic variant was bioinformatically evaluated in order to

Fig. 1. Pedigree chart, novel splice site donor mutation and aberrant splicing.
(Fig. 1A) Multi-generational family pedigree showing phenotype and genotype. Genomic DNAs are obtained from all living family members and one spouse (II-3). Genotypes are demonstrated in parenthesis as wild-type (T/T) and heterozygous (A/T) of REEP1 c.303þ2T > A variant. Solid symbols, affected individuals; circles, females; squares, males; slashes, deceased; bold line down the middle of the symbol with no shading asymptomatic carrier. Arrow plus P (Pb) indicates the proband.
(Fig. 1B) Sequencing electropherogram showing a heterozygous REEP1 c.303þ2T > A variant (red arrow) from intron 4.
(Fig. 1C) Agarose gel electrophoresis of the RT-PCR products revealed that the cDNA fragment of 434-bp corresponds to the predicted wild-type transcripts, whereas the cDNA small fragment of 313-bp corresponds to the transcripts lacking exon 4. Amplification of cDNA via RT-PCR using RNA extracted from; Cere 1⁄4 human cerebellum, FrCx 1⁄4 human frontal cortex, II-5 1⁄4 lymphocytes from the proband II-5, Cnt 1⁄4 lymphocytes from control individual.
(Fig. 1D) Sequence chromatograms from RT-PCR-amplified cDNA of REEP1 from the proband (II-5). Full length sequence containing non-mutated exon 3 and 4 junction is obtained from the PCR products with 434-bp (upper panel). Sequence analysis of the RT-PCR products with 313-bp demonstrates skipping of exon 4 and creating premature termination mutation in exon 5. Partial nucleotide sequence and the corresponding amino acid sequence are shown in lower panel. Amino acid residue position numbering is based on human longest isoform (NM_022912). Methionine encoded by the translation initiation site (start codon) is numbered as residue 1. Skipping of exon 4 creates a change in reading frame. Phenylalanine residue at position 62 is replaced with Lysine residue and premature termination codon is created at the position corresponding to codon 87 of REEP1 protein. Number in paracentesis is the position from Lysine-62 (counting starts with the Lysine as amino acid 1).
(Fig. 1E) qRT-PCR experiment. Comparison of expression level of REEP1 transcripts in cultured T cells from the proband (II-5 in Fig. 1a) demonstrates a significant increase with treatment of emetine, an NMD inhibitor. The asterisk (*) indicates a statistically significant difference (p < 0.05). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

understand whether it would influence splicing activity. Secondary structure and minimal free energy of wild and mutant RNA were predicted by the RNAFOLD program. Since the variant could affect mRNA splicing, both wild-type sequence and altered sequence were analysed in silico using the following five splice site prediction web interfaces: NetGene2, FruitFly, Spliceman, HSF Matrix and MaxENT. The previously reported exonic variant c.303G > A and intronic variant c.303þ1-7GTAATAT > AC were also analysed. (For a full description of the in silico analyses, refer to Supplementary Material).

REEP1 mRNA sequencing was conducted using cDNA synthe- sized from total RNA extracted from lymphocytes. Total RNA was reverse transcribed using random pentadecamer primers and reverse transcriptase (RT). Specific oligonucleotide primers were used for amplification of REEP1 transcripts and the RT-PCR products were subjected to sequence analysis. Measurement of REEP1 tran- scripts and nonsense mRNA mediated decay (NMD) analysis were also performed using the cultured T cells from the proband. Details of the analyses are described in Supplementary Material.

The study was performed according to a protocol reviewed and approved by the Institutional Review Boards (IRBs) of the Tokush- ima University Hospital, as well as by the Ethics Committee of the Istituto di Ricovero e Cura a Carattere Scientifico (IRCCS) Santa Lucia, Rome, Italy.

3. Results

The proband (II-5) developed his first clinical symptom of spasticity, including stumbling or tripping, from the fourth decade. No symptom related to sensory or autonomic nerve impairment was evident. The proband’s father (I-1) passed away in his 70s. He showed walking difficulty in his 50s, which slowly progressed. Within the same generation of the proband, there were three un- affected family members (II-1, II-2 and II-4). Neither signs of py- ramidal tract dysfunction nor peripheral neuropathy were shown in the member (II-2). The third generation was found to be affected, and severity seems to be independent of disease duration. The fourth generation was found to have a much earlier age of onset. Clinical features of all affected individuals in the Japanese HSP family are summarized in Table 1.

Whole exome sequencing demonstrated no coding variants in the currently-known genes associated with spastic paraplegia, but revealed a novel intronic variant in REEP1. The variant is located at the second position at 50-splie-site of intron 4, c.303þ2T > A (Fig. 1B), which is highly conserved across species (Fig. S3 in Sup- plementary Material). The intronic variant was found in all affected members and one unaffected member (II-2). The variant was ab- sent in the control datasets, and is not registered in the public databases investigated.

A stem-loop structure composed of wild sequence was pre- dicted to be altered by the intronic variant (Fig. S1). Minimum free energy was drastically changed, 6.40 kcal/mol in the wild to 4.40 kcal/mol in the variant c.303þ2T > A. Detrimental effects on splicing by the variant was predicted in all five analyses. The stem-loop structure and minimum free energy were also drastically altered by the variants, c.303G > A and c.303þ1-7GTAATAT > AC. (For a full description of the in silico analyses, refer to Supple- mentary Material).

In analysis of REEP1 transcripts, two PCR products were obtained from the proband and sequence analysis of the shorter products demonstrated REEP1 transcripts lacking exon 4, indicating an aberrant splicing by the intronic variant. No aberrant transcripts were detected in cerebellum, front al cortex and lymphocytes from healthy individuals (Fig. 1C and D). Significant changes of expres- sion of RREP1 transcripts in cultured T cells were observed with the treatment of emetine, which increased the expression level by approximately 1.7-fold in the proband (Fig. 1E). Increased aberrant REEP1 transcript by the treatment was evident in agarose gel electrophoresis (Fig. S2.).

4. Discussion

We report a novel REEP1 splice site donor variant, c.303þ2T > A, in a Japanese HSP family. The intronic variant cosegregated with the disease phenotype in the family and genotyping of our control datasets did not detect it. Moreover, the variant is not registered in any public database as rare normal variant or mutation in REEP1. These indicate that the variant is associated with the disease. Alteration of splicing by the variant was biologically confirmed and exon 4 was skipped out. Mutant REEP1 transcripts harbouring

Table 1

Clinical features of affected members. Individual ID

Age at examination (years) Age at onset (years)a Disease duration (years) Disability stageb

SPRS scorec
Lower limb hyperreflexia
Lower limb spasticity
Lower limb pyramidal weakness Babinski sign
Upper limb hyperreflexia
Upper limb spasticity
Sphincter disturbances
Scoliosis
Pescavus
Sensory deficits
Cognitive impairment

I-1

deceased

50sd

>10d 4de

unknown
unknown
unknown
unknown
unknown
unknown
unknown
unknown
-d unknowneþeeee unknowneeeeee -deeeeee

M. Kamada et al. / Parkinsonism and Related Disorders 46 (2018) 79e83 81

II-5 III-1

67 46

III-3

III-4

IV-1

IV-2

Early 40s >20
2
12

þþþ þ
e
þ

Early 30s >15
1
12

þþþ þ
e
þ

15
Early 10s
>3
1
3
þþþ
þ
e
þ
e
e
e ee

e
e
e eee

e e e

43
Middle 30s >10
1
ND
þþþ
þ
e
þ
e
e
e

41
Middle 30s >5
1
5
þþþ
þ
e
þ
e
e
e

13
Early 10s >1
1
7
þþþ
þ
e
þ
e
e
e

þ and , indicate the presence and absence of a feature, respectively.

. a Age at onset was calculated approximately as the time when difficulty in walking first appeared in the affected individuals.

. b Disability stages: 1, no mobility problems or slight stiffness of the legs; 2, moderate gait stiffness; 3, problems running, but able to walk alone; 4, problems walking; 5,

wheelchair user.
c SPRS, Spastic Paraplegia Rating Scale [12].

. d Data was obtained from interviews of the living family members.

. e At 70s.

82 M. Kamada et al. / Parkinsonism and Related Disorders 46 (2018) 79e83

premature termination codon were subjected to NMD pathway and partially degraded as suggested by a block of translation using emetine. Mutant REEP1 transcripts escaping from the NMD pathway would presumably be translated into protein carrying a premature stop codon (p.Phe62Lysfs*25), leading to production of truncated REEP1 proteind61 amino acids from native protein and additional 24 amino acids at the C-terminus, lacking the functional domain TB2_DP1_HVA22 (Fig. S3). Evaluation of REEP1 protein was not conducted due to technical and material limitations.

Three REEP1 variants have been reported in patients with spastic paraplegia, which have a potential effect on the splicing of exon 4, c.303þ1-7GTAATAT > AC [5], c.303þ2T > A (this study), and c.303G > A [6]. Skipping of exon 4 was biologically demonstrated in the former two variants. The exonic variant c.303G > A has been reported in a sporadic case with pure form of spastic paraplegia [6], which is a synonymous single nucleotide polymorphism (SNP). Although amino acid encoded by the codon (AAG/A) remains un- changed (p.Lys101Lys), splicing might be impaired by the exonic variant. Bioinformatic analyses for the exonic variant by this study provided equivocal results. Biological investigation is required to confirm the aberrant splicing by the variant.

Two variants, c.303þ1-7GTAATAT > AC [4] and c.303þ2T > A (this study), correspond to distinct phenotypes. Homozygous skipping of exon 4 was identified in a case with severe lower neuron involvement, similar to spinal muscular atrophy with res- piratory distress type 1 (SMARD1; OMIM 253300), while hetero- zygous skipping was observed in asymptomatic carriers [4] and in a family including six patients with pure form of spastic paraplegia and one asymptomatic carrier (this study). An animal model of REEP1-HSP has been created, in which exon 2 of Reep1 was knocked out, leading to a frame shift in the translation of exon 3. Only ho- mozygous knockout mice Reep1 / showed behavioural abnor- mality reflecting spastic clonus, and axonal deficits in the corticospinal tract were supported by ultrastructural studies [7]. Other Reep1 knockout mice also demonstrated a similar phenotype, thus, only homozygous mice exhibited spasticity and motor dysfunction [8]. The genotype-phenotype correlations in SPG31 patients and the results obtained from animal model suggested that molecular pathogenesis of SPG31/REEP1 is not due to simple haploinsufficiency. Most of the reported variants in REEP1 belong to a frame shift mutation, which leads to premature termination [6]. Mutant REEP1 proteins containing N-terminal fragment might be produced, which could act in a dominant negative manner on the native REEP1 protein or exert detrimental effects on interaction between atlastin-1 and spastin [9]. If these biological processes occur, mutant REEP1 mRNA dosage effect through NMD would be crucial for developing SPG31 or REEP1-associated diseases.

Intra-familial phenotypic variation was shown in the family, including age-at-onset and severity. Decrease of age at onset and increased severity in successive generations was observed, indi- cating a possible genetic anticipation. Moreover, an elderly asymptomatic carrier was identified, suggesting the involvement of additional genetic or environmental modifiers for disease devel- opment. Co-inheritance of modifier gene(s) or genetic factor(s) may influence the phenotype. Previous genetic studies in SPG4/SPAST and DYT1/TOR1A showed that specific intragenic or intrergenic alleles, or haplotype in cis or trans has an effect on disease severity or penetrance of the disease-causing mutation [10,11]. We reviewed the data of WES, but neither novel nor non-synonymous variant was found in REEP1 and nearby genes upstream and downstream, including SNORD94 and MRPL35, however, potential effect on phenotype by intervening sequence variant(s) at REEP1 locus cannot be excluded. The presence of intergenic modifier(s) accounting for developing the disease or phenotype cannot be excluded. Cross-sectional reviewing of the pooled genetic and

clinical data would reveal potential genetic or non-genetic modi- fier(s) in REEP1-associated diseases.

In conclusion, the evidence of further allelic heterogeneity in SPG31/REEP1 and intra-familial phenotypic variation was demon- strated. We also confirmed alteration of splicing leading to decreased dosage of REEP1 transcripts via NMD. Further accumu- lation of genotype-phenotype correlations, evaluation of dosage of REEP1 transcripts, functional analysis of mutant REEP1 proteins, and proteomic analysis are required for a better understanding of pathogenesis in SPG31/REEP1.

Funding

This work was supported by Health and Labour Science Research Grants for Research on rare and intractable Diseases: Clinical Research for Establishment of Evidence-Based Guideline for Hereditary Dystonias and Huntington Disease, and Grants-in-Aid from the Research Committee of CNS Degenerative Diseases from the Ministry of Health, Labour, and Welfare of Japan, and the Italian Ministero della Salute (Grant no. GR09.109 to AO).

Full financial disclosures of all authors

Dr. Masaki Kamada: no conflict to report.
Dr. Toshitaka Kawarai: no conflict to report. Dr. Ryosuke Miyamoto: no conflict to report. Ms. Yuki Tojima: no conflict to report.
Dr. Celeste Montecchiani: no conflict to report. Dr. Laura D’Onofrio: no conflict to report.
Dr. Carlo Caltagirone: no conflict to report.
Dr. Antonio Orlacchio: no conflict to report. Dr. Ryuji Kaji: no conflict to report.

Ethical standards

The authors hereby declare that the research documented in the submitted manuscript has been carried out in accordance with ethical standards laid down in the 1964 declaration of Helsinki and approved by the Institutional Review Boards (IRBs) of the Tokush- ima and Kagawa Universities, as well as by the Ethics Committee of the Istituto di Ricovero e Cura a Carattere Scientifico (IRCCS) Santa Lucia, Rome, Italy.

Acknowledgements

We are grateful to healthy donors, patients and their family members for their participation in this study. The authors thank Ms. Akemi Takahashi for her technical support. We thank Michela Renna, MA, for the language advice, as well as the Support Center for Advanced Medical Sciences, Tokushima University School of Medicine, for the use of their facilities to prepare the manuscript. We are extremely grateful to the Genetic Bank of the Laboratorio di Neurogenetica, CERC-IRCCS Santa Lucia, Rome, Italy for the service provided. Ms. Yuki Tojima carried out the research under the aca- demic supports by Medical Scholars Research Program of the Tokushima University Graduate School.

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.012.

References

[1] T. Lo Giudice, F. Lombardi, F.M. Santorelli, T. Kawarai, A. Orlacchio, Hereditary

spastic paraplegia: clinical-genetic characteristics and evolving molecular

mechanisms, Exp. Neurol. 261 (2014) 518e539.

. [2] S. Züchner, G. Wang, K.N. Tran-Viet, M.A. Nance, P.C. Gaskell, J.M. Vance,
A.E. Ashley-Koch, M.A. Pericak-Vance, Mutations in the novel mitochondrial protein REEP1 cause hereditary spastic paraplegia type 31, Am. J. Hum. Genet. 79 (2) (2006) 365e369.

. [3] C. Beetz, T.R. Pieber, N. Hertel, M. Schabhuttl, C. Fischer, S. Trajanoski, E. Graf, S. Keiner, I. Kurth, T. Wieland, R.E. Varga, V. Timmerman, M.M. Reilly, T.M. Strom, M. Auer-Grumbach, Exome sequencing identifies a REEP1 muta- tion involved in distal hereditary motor neuropathy type V, Am. J. Hum. Genet. 91 (1) (2012) 139e145.

. [4] G. Schottmann, D. Seelow, F. Seifert, S. Morales-Gonzalez, E. Gill, K. von Au, A. von Moers, W. Stenzel, M. Schuelke, Recessive REEP1 mutation is associated with congenital axonal neuropathy and diaphragmatic palsy, Neurol. Genet. 1 (4) (2015) e32.

. [5] J.K. Fink, Hereditary spastic paraplegia: clinico-pathologic features and emerging molecular mechanisms, Acta Neuropathol. 126 (3) (2013) 307e328.

. [6] C. Beetz, R. Schüle, T. Deconinck, K.N. Tran-Viet, H. Zhu, B.P. Kremer, S.G. Frints, W.A. van Zelst-Stams, P. Byrne, S. Otto, A.O. Nygren, J. Baets, K. Smets, B. Ceulemans, B. Dan, N. Nagan, J. Kassubek, S. Klimpe, T. Klopstock, H. Stolze, H.J. Smeets, C.T. Schrander-Stumpel, M. Hutchinson, B.P. van de Warrenburg, C. Braastad, T. Deufel, M. Pericak-Vance, L. Schols, P. de Jonghe, S. Züchner, REEP1 mutation spectrum and genotype/phenotype correlation in

hereditary spastic paraplegia type 31, Brain 131 (Pt 4) (2008) 1078e1086. [7] C. Beetz, N. Koch, M. Khundadze, G. Zimmer, S. Nietzsche, N. Hertel, A.K. Huebner, R. Mumtaz, M. Schweizer, E. Dirren, K.N. Karle, A. Irintchev, V. Alvarez, C. Redies, M. Westermann, I. Kurth, T. Deufel, M.M. Kessels, B. Qualmann, C.A. Hubner, A spastic paraplegia mouse model reveals REEP1-

dependent ER shaping, J. Clin. Invest. 123 (10) (2013) 4273e4282.
[8] B. Renvoise, B. Malone, M. Falgairolle, J. Munasinghe, J. Stadler, C. Sibilla, S.H. Park, C. Blackstone, Reep1 null mice reveal a converging role for heredi- tary spastic paraplegia proteins in lipid droplet regulation, Hum. Mol. Genet.

25 (23) (2016) 5111e5125.
[9] S.H. Park, P.P. Zhu, R.L. Parker, C. Blackstone, Hereditary spastic paraplegia

proteins REEP1, spastin, and atlastin-1 coordinate microtubule interactions

with the tubular ER network, J. Clin. Invest. 120 (4) (2010) 1097e1110.
[10] C.A. Hewamadduma, J. Kirby, C. Kershaw, J. Martindale, A. Dalton, C.J. McDermott, P.J. Shaw, HSP60 is a rare cause of hereditary spastic para- paresis, but may act as a genetic modifier, Neurology 70 (19) (2008)

1717e1718.
[11] N.J. Risch, S.B. Bressman, G. Senthil, L.J. Ozelius, Intragenic Cis and Trans

modification of genetic susceptibility in DYT1 torsion dystonia, Am. J. Hum.

Genet. 80 (6) (2007) 1188e1193.
[12] R. Schüle, T. Holland-Letz, S. Klimpe, J. Kassubek, T. Klopstock, V. Mall, S. Otto,

B. Winner, L. Schӧls, The Spastic Paraplegia Rating Scale (SPRS): a reliable and valid measure of disease severity, Neurology 67 (3) (2006) 430e434.

M. Kamada et al. / Parkinsonism and Related Disorders 46 (2018) 79e83 83

Parkinsonism and Related Disorders 46 (2018) 84e86

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Correspondence
TGM6 gene mutations in undiagnosed cerebellar ataxia patients

Keywords:

Spinocerebellar ataxias SCA35
Mutations

Autosomal dominant spinocerebellar ataxias (SCAs) are the most common types of hereditary cerebellar ataxia. Mutations in TGM6 have been identified as the cause of SCA35 in several Chinese SCA families. The main clinical manifestations include a slowly pro- gressive course, trunk/limb ataxia, and hand tremors [1e3]. How- ever, no TGM6 mutations in other populations or in individuals with sporadic cerebellar ataxia were reported [4].

In this study, we performed comprehensive mutation screening of TGM6 by Sanger sequencing in a cohort of undiagnosed cere- bellar ataxia patients, including 75 probands of autosomal domi- nant SCA families and 102 patients with sporadic cerebellar ataxia. All patients were followed in the Department of Neurology in the First Affiliated Hospital of Zhengzhou University. Meanwhile, 200 healthy Chinese individuals were analyzed as ethnically matched controls. The study was approved by the Ethics Commit- tee of First Affiliated Hospital of Zhengzhou University. All subjects in our research were Chinese and provided written informed consent.

We identified two mutations of TGM6 in these patients, including a novel splice-site mutation in a SCA family (family1) and a de novo missense mutation in a sporadic cerebellar ataxia pa- tient (family2). In family 1 (Fig. 1A), the proband (III-4) presented with slowly progressive gait unsteadiness, dysarthria, and hand tremor with onset of symptoms at age 54. Her sister (III-3) had been experiencing similar symptoms since age 59. Their deceased father had suffered from progressive gait disturbance and dysar- thria beginning at age 50 and had died following a stroke at age 52. Other members of the family (III-1, III-2, III-5, II-1, and II-4) were without neurological symptoms. MRI of the brain in III-3 and III-4 demonstrated diffuse cerebellar atrophy; MRI was normal in III-5 (Fig. 1B). In family 2 (Fig. 1A), the proband was a 62-year-old female patient (II-1) who presented with progressive gait unstead- iness, dysarthria, hand tremor, and memory impairment that had

http://dx.doi.org/10.1016/j.parkreldis.2017.07.001

first appeared at age 60. MRI of the brain demonstrated mild cere- bellar atrophy (Figure not shown). However, other family members (I-1, I-2 andII-2) were all healthy, without any neuropsychological dysfunction.

A splice-site mutation in TGM6, c.7þ1G > T, was identified in the proband (III4) of family 1 (Fig. 1C). Subsequently, Sanger sequencing in other family members (II-1, II-4, III-1, III-2, III-3, and III-5) confirmed that the other patient (III-3) also harbored the mutation, and that no clinically unaffected family members carried the muta- tion except for a 44-year-old pre-symptomatic carrier (III-5). The mutation was absent in 200 ethnically matched controls. In family 2, mutation screening for TGM6 showed a missense mutation c.1478C > T, p.P493L in the proband (II-1) (Fig. 1C), while other fam- ily members (I-1, I-2 andII-2) in the family and 200 healthy controls were without the mutation.

Transglutaminase 6 (TGM6, NM_198994) encodes transglutami- nase 6 protein (TG6), which is a member of the transglutaminase family of Ca2þ dependent enzymes that catalyze the cross-linking of proteins and the conjugation of polyamines to proteins [5]. Pre- vious functional studies and bioinformatics analysis have revealed that TGM6 is associated with ataxia and that mutation in TGM6 may decrease TG6 stability and transglutaminase activity [1e3]. In this study, the mutation c.7þ1G > T affected the initial nucleotide of the splice donor site of intron 1 (IVSþ1G > T). Several splice-site mutation bioinformatics tools, including HSF (Human Splicing Finder), SSS (Splice site score calculation), and Max EntScan, revealed decreased splicing ability of the splice-site mutation, which might influence the normal splicing of mRNA and further damage the physiological functions of TG6. The mutation c.1478C > T, p.P493L affected conserved amino acids (Fig. 1D). Using I-Mutant, the mutation would decrease the stability of TG6. Using AGGRESCAN, the wild type 493P was located at a hotspot area (ag- gregation-prone segment) from the 496th to the 503th site, and the mutant type 493L extended this area to involve the 494th site to the 503th site, which might also change the normal protein folding.

In summary, we identified a novel splice-site mutation of TGM6 in a SCA family, and a de novo missense mutation of TGM6 in a spo- radic cerebellar ataxia patient. Our study illustrates the importance of TGM6 mutations in Chinese cerebellar ataxia patients, not only in SCA families, but also in individuals with sporadic cerebellar ataxia. Genetic testing of SCA35 should be considered for undiagnosed cerebellar ataxia patients in clinical practice.

Correspondence / Parkinsonism and Related Disorders 46 (2018) 84e86 85

Fig. 1. The pedigree charts, genetic findings and brain MRI in the family1 and faimly2. A: Pedigree chart: asterisk, members sequenced. B: T2-weighted brain MRI of III-3, III-4 and III-5 in family1. C: Sequencing chromatograms. D: Conservation analysis of the mutation site, 493P.

Disclosure

The authors report no conflicts of interest relevant to the manuscript.

Potential conflict of interest None.

Acknowledgements

Natural Science Foundation of China (to Dr Changhe Shi) and grants 81530037 and 81471158 from the National Natural Science Founda- tion of China (to Dr Yuming Xu).

References

[1] J.L. Wang, X. Yang, K. Xia, Z.M. Hu, L. Weng, X. Jin, et al., TGM6 identified as a novel causative gene of spinocerebellar ataxias using exome sequencing, Brain a J. Neurology 133 (2010) 3510e3518.

[2] M.Li,S.Y.Pang,Y.Song,M.H.Kung,S.L.Ho,P.C.Sham,Wholeexomesequencing identifies a novel mutation in the transglutaminase 6 gene for spinocerebellar ataxia in a Chinese family, Clin. Genet. 83 (2013) 269e273.

[3] Y.C. Guo, J.J. Lin, Y.C. Liao, P.C. Tsai, Y.C. Lee, B.W. Soong, Spinocerebellar ataxia

The work was supported by grant U1404311 from the National

Correspondence / Parkinsonism and Related Disorders 46 (2018) 84e86

[4] [5]

35: novel mutations in TGM6 with clinical and genetic characterization, Neurology 83 (2014) 1554e1561. B.L.Fogel,J.Y.Lee,J.Lane,A.Wahnich,S.Chan,A.Huang,etal.,Mutationsinrare ataxia genes are uncommon causes of sporadic cerebellar ataxia, Mov. Disord. official J. Mov. Disord. Soc. 27 (2012) 442e446.

P. Grenard, M.K. Bates, D. Aeschlimann, Evolution of transglutaminase genes: identification of a transglutaminase gene cluster on human chromosome 15q15. Structure of the gene encoding transglutaminase X and a novel gene family member, transglutaminase Z, J. Biol. Chem. 276 (2001) 33066e33078.

* Corresponding author. Department of Neurology, The First Affiliated Hospital of Zhengzhou University, Zhengzhou University, 1 Jian-she east road, Zhengzhou 450000, Henan, China.

** Corresponding author. E-mail addresses: shichanghe@gmail.com (C.-h. Shi), xuyuming@zzu.edu.cn (Y.-m. Xu).

1 March 2017

Zhi-hua Yang, Meng-meng Shi, Yu-tao Liu, Yan-lin Wang, Hai-yang Luo, Zhi-lei Wang, Chang-he Shi**, Yu-ming Xu* Department of Neurology, The First Affiliated Hospital of Zhengzhou University, Zhengzhou University, Zhengzhou, 450000, Henan, China

Parkinsonism and Related Disorders 46 (2018) 87e89

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Correspondence

A novel compound heterozygous TH mutation in a Japanese case of dopa- responsive dystonia with mild clinical course

Dopa-responsive dystonia (DRD) is due to decreased dopamine synthesis involving four genes: GTP cyclohydrolase 1 (GCH1), sepiapterin reductase (SPR), 6-pyruvoyl-tetrahydropterin synthase (PTPS) and tyrosine hydroxylase (TH). GCH1 deficiency is inherited in an autosomal dominant mode and the other three in a recessive mode [1]. TH deficiency directly leads to constant impairment of catecholaminergic neurotransmission, disturbance of prenatal brain development and postnatal growth failure. The severe condi- tions caused by TH deficiency are not always cured by levodopa [2]. Here, we describe a patient carrying a novel compound heterozy- gous TH mutation with mild clinical course. This study was approved by the Ethical Committee of Tokushima University.

The patient is a 3 year and 10 month old girl, who was delivered at full term after an uncomplicated pregnancy. She is the first child of non-consanguineous parents [Fig. 1A]. Initial development appeared to be normal; however, miosis, occasional upward eye de- viation and unusual foot posture were observed at the age of 3 months. Neurological examinations revealed little spontaneous movement, paroxysmal irritability, poor eye contact and bilateral ptosis. Videograms recorded by the parents showed paroxysmal dystonic movements of the foot and oculogyric crises (Supplemen- tary video). Cerebrospinal fluid analysis at the age of 5 months showed low homovanillic acid (HVA) levels of 71.4 nmoL/L (normal range, 231e840 nmoL/L), and normal 5-hydroxyindolacetic acid (5- HIAA) levels of 286.6 nmoL/L (normal range, 83e341 nmoL/L). Segawa syndrome or dopa-responsive dystonia was suspected, and a trial of levodopa (1.0 mg/kg/day) was initiated at the age of 8 months, leading to the disappearance of ptosis, oculogyric crises and dystonic movements. Her global development quotient (mental age/chronological age X 100) gradually improved from the first to the last assessment: 52 at age 1 year and 6 months, and 64 at age 3 years. No motor or autonomic signs are currently evident, including hypotonia, hypokinesia, dystonia, parkinsonism, and temperature instability.

Supplementary video related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.019.

Segment 1, Dystonic posture is seen in the left foot; Segment 2, Oculogyric crisis, ptosis and miosis are observed at the age of 3 months.; Segment 3, Ptosis and miosis disappeared after adminis- tration of tropicamide eye drops.; Segment 4, Titubation is seen but she can walk alone at the age of three years. The patient’s mother asks her about her current age and she answers to the ques- tion using fingers.

Sequence analyses of TH gene in the proband showed com- pound heterozygosity for two mutations, c.698G > A and c.1141-

https://doi.org/10.1016/j.parkreldis.2017.10.019

1G > A. The former nucleotide substitution results in Arginine-to- Histidine amino acid change at codon 233 (p.Arg233His), where the Arginine residue is highly conserved in evolution. The latter is located at the splice acceptor site at the 30 end of the intron 10 terminate, which is highly conserved across species, and the intronic variant is predicted to alter the secondary structure of TH mRNA [Fig. 1BeF].

Patients with homozygous p.Arg233His have been reported to show severe phenotype of progressive infantile encephalopathy. A residual activity of 14% has been demonstrated in an assay of steady-state kinetic properties of the recombinant mutant enzyme [3]. Our case carries a compound heterozygous mutation, c.698G > A (p.Arg233His) and c.1141-1G > A. Biological effects by the splice site acceptor mutation c.1141-1G > A remain undeter- mined due to technical and biomaterial limitations. Bioinformatic analysis demonstrated several cryptic donor and acceptor sites in introns 10 and 11 (supplementary material). Although aberrant splicing is elicited by the splice site acceptor mutation, normal splicing may be allowed to proceed to some extent [4].

Both wild and aberrant TH transcripts are synthesized: the former is further translated to wild TH protein and the latter may be subjected to nonsense-mediated mRNA decay. Mutant TH pro- tein harbouring 233His residue and decreased wild protein might be attributed to residual functional activity. Taking into account the mild clinical course, the residual activity may be beyond critical threshold. The patient’s HVA concentration was approximately 30.9% of the lower limit of the reference range, which supports this hypothesis.

The patient’s mother has the heterozygous p.Arg233His variant and shows restless legs syndrome (RLS). It has been postulated that dopaminergic deficiency is an underlying pathophysiological con- dition of RLS. Biochemical properties of tyrosine hydroxylase in the substantia nigra have been investigated [5]. We conducted further genetic analyses, including haplotyping at the TH locus, and sequenced the genes currently known for RLS, but could not demonstrate any supporting evidence for intragenic cis or trans modification, or multifactorial genetic model (supplementary material).

To date, a total of 55 and more nucleotide variants have been re- ported in TH-DRD, including 3 in promoter region, 44 in missense, 5 in frameshift, and 3 in splice site (including this case). p.Arg233His is the most frequent variant, and c.1141-1G > A is novel. Our study demonstrates a further evidence of allelic heterogeneity in TH-DRD and contributes to the understanding of genotype-phenotype correlations.

88 Correspondence / Parkinsonism and Related Disorders 46 (2018) 87e89

Fig. 1. Pedigree chart, sequence chromatograms, and results by bioinformatics analyses. (A) Pedigree chart. Solid symbols filled with black, affected individuals with DRD; Solid symbols filled with grey, affected with restless legs syndrome; Open symbols, unaffected individuals; slashes, deceased; arrow, patient (IV-1). TH genotype is described in pa- rentheses; Wt, wild type; MV, missense variant c.698G > A; IV, intronic variant c.1141-1G > A.
(B) Sequence fluorescent chromatograms of TH exon 6 sequence obtained from father, mother and patient. Nucleotide variant c.698G > A (p.Arg233His) is indicated by a red arrow. (C) The p.233R (red colour) is completely conserved amongst TH homologues. Amino acid numbering above the alignment is based on the human protein.

(D) Sequence fluorescent chromatograms of TH exon 11 and 30 adjacent sequence. Mutation c.1141-1G > A is indicated by a red arrow. Small letter indicates intronic sequence and capital letter inside rectangle represents exonic sequence.
(E) Schematic representation and sequence of boundary between TH intron 10 and exon 11 in various species. The first nucleotide at 30-splice-site, guanine, in human as well as other species is shown in red colour, which is substituted by adenine in the patient.

(F) Secondary structure of the wild and mutant RNA predicted by the RNAFOLD program. The secondary structure of the wild-type RNA sequence contains a stem-loop structure; a c.1141-1G residue is located within the stem (dG 1⁄4 4.70 KcaL/mol). The mutation, c.1141-1G > A, created two smaller stem-loop structures with (dG 1⁄4 3.50 KcaL/mol), with a c.1141-1A residue located at the outer loop. Sequence of exon 11 is shown in bold. Sequence information and method of bioinformatic analyses are available in supplementary material. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Disclosure

The authors report no conflicts of interest relevant to the manuscript.

Relevant conflicts of interest/financial disclosures Nothing to report.

Acknowledgements

We are grateful to healthy donors, patients and their family

members for their participation in this study. The authors thank Ms. Akemi Takahashi for her technical support. We thank Michela Renna, MA, for the language advice, as well as the Support Center for Advanced Medical Sciences, Tokushima University School of Medicine, for the use of their facilities to prepare the manuscript. We are extremely grateful to the Genetic Bank of the Laboratorio di Neurogenetica, CERC-IRCCS Santa Lucia, Rome, Italy for the service provided. This work was supported by Health and Labour Science Research Grants for Research on rare and intractable Diseases: Clin- ical Research for Establishment of Evidence-Based Guideline for Hereditary Dystonias and Huntington Disease, and Grants-in-Aid from the Research Committee of CNS Degenerative Diseases from

the Ministry of Health, Labour and Welfare, Japan, and the Italian Ministero della Salute (Grant no. GR09.109 to AO).

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.019.

References

. [1] S. Wijemanne, J. Jankovic, Dopa-responsive dystoniaeclinical and genetic het- erogeneity, Nat. Rev. Neurol. 11 (7) (2015) 414e424.

. [2] G.F. Hoffmann, B. Assmann, C. Brautigam, C. Dionisi-Vici, M. Haussler, J.B. de Klerk, M. Naumann, G.C. Steenbergen-Spanjers, H.M. Strassburg, R.A. Wevers, Tyrosine hydroxylase deficiency causes progressive encephalopathy and dopa-nonresponsive dystonia, Ann. Neurol. 54 (Suppl 6) (2003) S56eS65.

. [3] A. Fossbakk, R. Kleppe, P.M. Knappskog, A. Martinez, J. Haavik, Functional studies of tyrosine hydroxylase missense variants reveal distinct patterns of molecular defects in Dopa-responsive dystonia, Hum. Mutat. 35 (7) (2014) 880e890.

. [4] R.J. Janssen, R.A. Wevers, M. Haussler, J.A. Luyten, G.C. Steenbergen-Spanjers, G.F. Hoffmann, T. Nagatsu, L.P. Van den Heuvel, A branch site mutation leading to aberrant splicing of the human tyrosine hydroxylase gene in a child with a severe extrapyramidal movement disorder, Ann. Hum. Genet. 64 (Pt 5) (2000) 375e382.

. [5] J.R. Connor, X.S. Wang, R.P. Allen, J.L. Beard, J.A. Wiesinger, B.T. Felt, C.J. Earley, Altered dopaminergic profile in the putamen and substantia nigra in restless leg syndrome, Brain 132 (Pt 9) (2009) 2403e2412.
Kozue Kuwabara1

Department of Pediatrics, Ehime Prefectural Central Hospital, Kasuga- cho 83, Matsuyama City, Ehime, 790-0024, Japan

Toshitaka Kawarai*,1 Department of Clinical Neuroscience, Institute of Biomedical Sciences, Tokushima University Graduate School, Kuramoto-cho 3-18-15, Tokushima City, 770-0042, Japan

Yasushi Ishida

Department of Pediatrics, Ehime Prefectural Central Hospital, Kasuga- cho 83, Matsuyama City, Ehime, 790-0024, Japan

Ryosuke Miyamoto

Department of Clinical Neuroscience, Institute of Biomedical Sciences, Tokushima University Graduate School, Kuramoto-cho 3-18-15, Tokushima City, 770-0042, Japan

Ryosuke Oki

Department of Clinical Neuroscience, Institute of Biomedical Sciences, Tokushima University Graduate School, Kuramoto-cho 3-18-15, Tokushima City, 770-0042, Japan

Antonio Orlacchio

Laboratorio di Neurogenetica, Centro Europeo di Ricerca sul Cervello (CERC) – Istituto di Ricovero e Cura a Carattere Scientifico (IRCCS) Santa Lucia, Via del Fosso di Fiorano 64, Rome 00143, Italy

Dipartimento di Scienze Chirurgiche e Biomediche, Universita di Perugia, Piazza Lucio Severi 1, Perugia 06132, Italy

Yoshiko Nomura

Yoshiko Nomura Neurological Clinic for Children, Yushima 1-2-13, Bunkyo-ku, Tokyo, 113-0034, Japan

Mitsumasa Fukuda

Department of Pediatrics, Ehime University Graduate School of Medicine, 454 Shitsukawa, Toon, Ehime, 791-0295, Japan

Eiichi Ishii

Department of Pediatrics, Ehime University Graduate School of Medicine, 454 Shitsukawa, Toon, Ehime, 791-0295, Japan

Haruo Shintaku

Department of Pediatrics, Osaka City University Graduate School of Medicine, Asahimachi 1-5-7, Abeno-ku, Osaka City, 545-8586, Japan

Ryuji Kaji

Department of Clinical Neuroscience, Institute of Biomedical Sciences, Tokushima University Graduate School, Kuramoto-cho 3-18-15, Tokushima City, 770-0042, Japan

* Corresponding author. Department of Clinical Neuroscience, Institute of Biomedical Sciences, Tokushima University Graduate School, 770-0042, Japan. E-mail address: tkawarai@tokushima-u.ac.jp (T. Kawarai).

25 July 2017

Correspondence / Parkinsonism and Related Disorders 46 (2018) 87e89 89

1 These authors contributed equally.

Parkinsonism and Related Disorders 46 (2018) 90e91

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Correspondence

Anterior cingulate cortex as an element of a possible novel motor circuit of the basal ganglia

Keywords:

Parkinsonism Cerebrovascular disease Basal ganglia
Anterior cingulate cortex PET

Dear Sir,

Brain regions can be functionally parceled into several circuits centered on the basal ganglia (BG); the anterior cingulate cortex (ACC) is a key element of the limbic circuit of BG [1]. There has been little evidence supporting the role of ACC on BG motor circuit until now [2,3].

Here, we report on a patient with an ACC infarction who showed Parkinsonism after recovering from motor weakness.

A 45-year-old female was admitted due to sudden weakness of her left limbs. Her medical history was unremarkable. Neurological examination (NE) showed dysarthria and mild hemiparesis. Initial neuropsychological evaluation showed mild frontal-executive dysfunction.

Brain magnetic resonance imaging (MRI) showed an acute cere- bral infarction involving ACC including the cingulate motor area (CMA) and medial superior frontal cortex (MSFC) (Fig. 1A and B). MR angiography was normal. 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) showed hypometabolism in the frontal cortex, caudate nucleus (CN), anterior-dorsal portion of the putamen (PU), thalamus (TH), and substantia nigra (SN) on the right side (Fig. 1C). 18F-fluorinated N-3-fluoropropyl-2-beta- carboxymethoxy-3-beta-(4-iodophenyl) nortropane (18F-FP-CIT) PET showed decreased uptake of 18F-FP-CIT in the posterior portion of right PU (Fig. 1D). Cardiac workups were unremarkable. Autoan- tibodies, such as anti-b2-glycoprotein 1, anti-cardiolipin, anti-DNA and anti-polymyositis/scleroderma antibodies, were detected. She was diagnosed as having antiphospholipd syndrome.

The patient’s hemiparesis rapidly improved within 5 days. Follow-up NE showed normal muscle power. However, the move- ment of her left limbs was slightly slow and the amplitude gradu- ally decreased on repetition (Supplemental video 1). Rigidity and postural instability were absent. She was treated with hepariniza- tion and rivaroxaban.

Supplementary video related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.018.

https://doi.org/10.1016/j.parkreldis.2017.10.018

During follow-up, she had no neurological problems, except persistent bradykinesia on NE (Supplemental video 2). Follow-up neuroimaging was performed 22 months after onset. 18F-FPCIT PET showed normal findings. 18F-FDG PET demonstrated a slightly recovery in metabolism, except in TH (Fig. 1E). Voxel-based subtrac- tion analysis of 18F-FDG PET showed metabolic recovery in ACC, MSFC, and PU (Fig. 1F). We received the informed consent for the report and supplemental video from the patient.

Supplementary video related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.018.

In the initial phase, the patient suffered from temporary motor weakness and cognitive impairment, which could be attributed to a lesion effect of ACC and MSFC. Hemiparesis and dysarthria can be understood by dysfunction of CMAs, which are anatomically related to the spinal cord, via the corticospinal tract (CST) [2]. As hemiparesis was mild and short-lasting, the lesion effect of CMA could be milder than that of the primary motor cortex. Although she did not complain any subjective clumsiness of left hand, NE persistently demonstrated slowness with decremental response, which was more compatible with bradykinesia as an isolated form of Parkinsonism, rather than pyramidal weakness.

Parkinsonism is currently understood as being associated with an impaired motor circuit. In the current patient, although anatom- ical lesions were to ACA region, functional studies showed exten- sive hypometabolism in CN, PU, TH and SN, and dopaminergic deficiency in the striatum. Since these areas play an important role in the motor function of BG, it is possible to assume a new mo- tor circuit of BG that encompassing both ACC and PU.

Involvement of SN was demonstrated in two ways. First, impaired SN metabolism was incompletely recovered on the follow-up 18F-FDG PET. Previous studies reported delayed SN dam- age via trans-synaptic degeneration in cases with striatal infarc- tions [4]. Striatal dysfunction secondary to an ACA infarction may result in SN hypometabolism. Second, involvement of SN pars com- pacta (SNpc) was assumed by decreased uptake of 18F-FP-CIT in the posterior PU. However, given that most striatal output is directed toward SN pars reticulata, the involvement of SNpc secondary to the ipsilateral striatum is challenging to explain. In the current case, the direct projection from striatal striosomes to SNpc can be considered. Interestingly, afferents from limbic cortices, including ACC, preferentially innervates striosomes [5]. Therefore, involve- ment of SNpc could be attributed to the ACC-striatum (strio- some)-SNpc connection.

Finally, the unchanged hypometabolism of TH might underlie her persistent Parkinsonism, since TH is a final output of the BG

motor circuit.
In conclusion, we suggest a novel BG motor circuit that origi-

nates from ACC. Apart from its direct connection with CST, however, the role of CMA remains uncertain. Further investigations including functional and pathological studies on the contribution of CMA or striosomes in patients with ACC lesions might enable the detailed characteristics of this novel pathway to be elucidated.

Acknowledgments

None.

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.018.

References

. [1] G.E. Alexander, M.D. Crutcher, Functional architecture of basal ganglia circuits: neural substrates of parallel processing, Trends. Neurosci. 13 (1990) 266e271.

. [2] C. Amiez, M. Petrides, Neuroimaging evidence of the anatomo-functional orga-
nization of the human cingulate motor areas, Cereb. Cortex 24 (2014) 563e578.

. [3] T. Paus, Primate anterior cingulate cortex: where motor control, drive and
cognition interface, Nat. Rev. Neurosci. 2 (2001) 417e424.

. [4] B. Winter, P. Brunecker, J.B. Fiebach, G.J. Jungehulsing, G. Kronenberg,

M. Endres, Striatal infarction elicits secondary extrafocal MRI changes in ipsilat-

eral substantia nigra, PLoS One 10 (2015) e0136483.
[5] J.R. Crittenden, A.M. Graybiel, Basal Ganglia disorders associated with imbal-

ances in the striatal striosome and matrix compartments, Front. Neuroanat. 5 (2011) 59.

Chaewon Shin1, Young Nam Kwon1, Dokyung Lee Department of Neurology, Kyung Hee University Hospital, Kyung Hee University College of Medicine, Seoul, Republic of Korea

Il Ki Hong

Department of Nuclear Medicine, Kyung Hee University Hospital, Kyung Hee University College of Medicine, Seoul, Republic of Korea

Hyug-Gi Kim, Kyung Mi Lee

Department of Radiology, Kyung Hee University Hospital, Kyung Hee University College of Medicine, Seoul, Republic of Korea

Tae-Beom Ahn* Department of Neurology, Kyung Hee University Hospital, Kyung Hee University College of Medicine, Seoul, Republic of Korea

* Corresponding author. Department of Neurology, Kyung Hee University Hospital, 23 Kyungheedae-ro, Dongdaemun-gu, Seoul 130-872, Republic of Korea. E-mail address: ricash@hanmail.net (T.-B. Ahn).

25 July 2017

Correspondence / Parkinsonism and Related Disorders 46 (2018) 90e91 91

Fig. 1. Neuroimages of the patient. A. Brain magnetic resonance imaging (MRI) showing an acute infarction in the medial superior frontal cortex (MSFC) and anterior cingulate cortex (ACC) on fluid attenuated inversion recovery (FLAIR) and diffusion-weighted images. Circles represent rostral (small yellow circle) and caudal (large green circle) cingulate motor area (CMA) regions. B. T1-weighted MRI demonstrating low signal intensities (arrow) in the ACC, including CMA. There is no lesion in the supplementary motor area (SMA) in the coronal view. va 1⁄4 vertical line passing through the anterior commissure, vp 1⁄4 vertical line passing through the posterior commissure C. Positron emission tomography (PET) using 18F-fluorodeoxyglucose (18F-FDG) showing hypometabolism in the frontal cortex, caudate nucleus, anterior putamen, thalamus, and substantia nigra in the right hemisphere. D. PET using 18F-fluorinated N-3-fluoropropyl-2-beta-carboxymethoxy-3-beta-(4-iodophenyl) nortropane (18FP-CIT) demonstrating decreased uptake in the right posterior puta- men. E. A follow-up 18F-FDG PET showing improved metabolism in the putamen. Hypometabolism in the right thalamus remained unchanged. F. Voxel-based subtraction analysis of two 18F-FDG-PET images (C and E) showing the metabolic recovery in the ACC, MSFC, and anterior-dorsal putamen. Metabolic activity of other motor-related areas is not decreased. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

These two authors contributed equally to this study.

Parkinsonism and Related Disorders 46 (2018) 92e94

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Correspondence
When shaking during standing points to hereditary spastic paraplegias

Keywords:

Shaking
Tremor
Standing
Spastic paraplegias Paraparesis Orthostatic

Shaking on standing is a highly disabling syndrome caused by different disorders [1]. Differential diagnosis heavily relies on elec- trophysiology. High-frequency tremor (>13 Hz) of the legs when standing is the hallmark of classic orthostatic tremor (OT). Alterna- tively, other mimics should be considered [1]. Hereditary spastic paraplegias (HSP) are a heterogeneous group of inherited neuro- logical disorders with the cardinal feature of a corticospinal motor neurons dysfunction, classified as either pure or complex based on the absence or presence, respectively, of associated signs [2]. Here- in, we describe two cases of HSP-associated shaking on standing.

1. Case 1

A 74-year-old woman visited us for a 15-year history of leg shaking on standing. Shaking on standing progressively increased over time, leading to a few falls and limitation of her activities. Holding onto something allowed her to be more stable. Shaking would completely disappear during walking, sitting or lying down. She had no benefit from several medications (propranolol, primidone, gabapentin, levodopa) and only marginal benefit from clonazepam (1 mg/day) and diazepam (2 mg/day). Eventually, she was referred to us as a candidate for surgical options.

She reported a positive family history of similar problems (Fig. 1A). On examination, she showed shaking only on standing (Video 1). Her phenotype was consistent with OT, further confirmed by EMG recording showing an 18Hz legs tremor, highly coherent between homologous leg muscles (Fig. 1B). Brain and spine MRI were normal. Prompted by the family history, the mild pyramidal signs and recent data reported in the literature [3], we tested the patient for mutations in the REEP1 gene (SPG31). A pre- viously reported mutation (c.50G > A) was detected [4]. Finally, the patient underwent spinal cord stimulation without any benefit despite multiple intensive programming sessions.

Abbreviations: EMG, electromyography; HSP, hereditary spastic paraparesis; MRI, magnetic resonance imaging; pseudo-OT, pseudo- orthostatic tremor; OT, orthostatic tremor.

https://doi.org/10.1016/j.parkreldis.2017.10.017

Supplementary video related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.017.

2. Case 2

A 30-year-old woman visited us for a slowly progressive gait disturbance further worsened by the recent onset of shaking on standing. Born full-term from non-consanguineous parents, she had delayed acquisition of language with moderate mental retarda- tion. Family history was negative for neurological disorders. She developed progressive gait difficulties and falls in her late teens as well as mood disturbances.

On examination, she had a severely spastic and mildly ataxic gait, only possible with bilateral support, with hypertone only in the legs (Video 2 and 3). Although no parkinsonian signs were detected, a treatment with levodopa was attempted (up to 600 mg/die), which resulted in a subjective and objective gait improvement (Video 3). However, over time she started reporting shaking on standing that appeared after walking few steps or standing upright for few minutes (Video 3). Shaking was extin- guished by leaning on objects. An acute levodopa challenge decreased both the latency of tremor appearance and its severity. EMG showed a low frequency tremor at about 9 Hz, which was not coherent between homologous leg muscles (Fig. 1C). Needle EMG and nerve conduction studies showed axonal sensitive-motor pol- yneuropathy and brain MRI thinning of the corpus callosum (Fig. 1D). Genetic testing revealed no spatacsin (SPG11) mutations, but compound heterozygous mutations in spastizin (SPG15) (c.1902_1903del and c.2254 > T).

Supplementary video related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.10.017.

3. Discussion

We describe two HSP cases that e although with different de- gree of disability and complexity of the phenotype e shared a rare and specific sign, namely shaking on standing.

REEP1 mutations (SPG31) are the third most common cause of autosomal dominant HSP, accounting for up to 8% of pure forms [4], but our case presented exclusively with electrophysiologically- confirmed OT. This is the first independent replication that OT might be part of phenotypical spectrum of REEP1 mutations [3]. In contrast with other cases, we showed that spinal stimulation was unsuccess- ful, likely mirroring the genetic heterogeneity of ‘primary’ OT.

SPG15 represents the second most common cause of autosomal

Correspondence / Parkinsonism and Related Disorders 46 (2018) 92e94 93

Fig. 1. A: Pedigree of case 1 showing an autosomal dominant inheritance pattern. Subject I-1 (father of the index case) presented shaking on standing since the age of 60. Subject III- 1 reports shaking on standing since the age of 40; he lives in another country and could not be assessed.
B: Electrophysiology of case 1. Surface EMG recorded while standing from right quadriceps (R QC) and left quadriceps (L QC) demonstrating short, synchronous, rhythmic bursts at a frequency of about 19 Hz (left panel) and highly coherent between homologous muscles (right panel).

C: Electrophysiology of case 2. Surface EMG recorded while standing from R QC, L QC, right tibialis anterior (R TA) and left tibialis anterior (L TA), demonstrating rhythmic bursts with a frequency of about 9 Hz.
D: Brain MRI of case 2. T1-weighted sagittal MRI images (left) show enlarged ventricles, cerebellar atrophy and thinning of the corpus callosum. Axial fluid attenuation inversion recovery images (right) show increased signal particularly anterior and posterior to the lateral ventricles.

recessive HSP, accounting for up to 14.8% of complex forms [2]. Similarly to other published cases [2], our patient presented motor symptoms since the second decade preceded by cognitive impair- ment and behavioral disturbances. Our patient is the first SPG15 case presenting with shaking on standing at slow frequency. Several types of ‘slow-OT’ have been described and our case’s pre- sentation might be underlined by two different mechanisms: 1. The so-called ‘pseudo-OT’, known to be caused by parkinsonian tremor and responding to levodopa [1]; 2. A pyramidal clonus, although in these cases frequency is usually lower and there should be no response to levodopa [5]. In fact, although our patient’s response to levodopa is in keeping with previously reported cases [2], we cannot exclude that tremor may have fatigued after standing mul- tiple times.

Shaking on standing is a heterogeneous syndrome that may be under-reported in patients with HSP. Our cases also emphasize that electrophysiology is mandatory to correctly identify the under- lying dysfunction and possibly tailor correct management.

Conflict of interest None.

Informed consent

Both patients provided written informed consent for video publication.

Authors’ contributions

drafted the manuscript. Roberto Erro executed the study and revised the manuscript for intellectual content. Alfonso Fasano conceived the study and revised the manuscript for intellectual content. Renato P. Munhoz revised the manuscript for intellectual content.

Study funding

No targeted funding reported.

Full financial disclosures

MP received salary from the University of Salerno, Italy.

RE received consultancies from Zambon and honoraria for speaking from Teva.

RPM reports no financial disclosures.

AF received speaker and/or consulting honoraria from Apple, Abbvie, Boston Scientific, Chiesi pharmaceuticals, Ipsen, Med- tronic, TEVA Canada, UCB pharma and research grants from Neure- caOnlus, AFaR, the Division of Neurology – University of Toronto, McLaughlin Centre, Weston Foundation and the Michael J. Fox Foundation. He is in the editorial board of BioMed Research Inter- national and PlosOne.

Acknowledgments

We are grateful to Drs. Robert Chen and Ariel Levy for providing the images of the electrophysiological study of case 1.

References

Marina Picillo conceived, organized and executed the study and

[1] R. Erro, K. Bhatia, C. Cordivari, Shaking on standing: a critical review, Mov.

94 Correspondence / Parkinsonism and Related Disorders 46 (2018) 92e94

Disord. Clin. Pract. 1 (2014) 173e179.

. [2] V. Pensato, B. Castellotti, C. Gellera, D. Pareyson, C. Ciano, L. Nanetti, E. Salsano,
G. Piscosquito, E. Sarto, M. Eoli, I. Moroni, P. Soliveri, E. Lamperti, L. Chiapparini, D. Di Bella, F. Taroni, C. Mariotti, Overlapping phenotypes in complex spastic paraplegias SPG11, SPG15, SPG35 and SPG48, Brain 137 (2014) 1907e1920.

. [3] R. Erro, C. Cordivari, K.P. Bhatia, SPG31 presenting with orthostatic tremor, Eur. J. Neurol. 21 (4) (2014) 34e35.

. [4] C. Beetz, R. Schule, T. Deconinck, K.N. Tran-Viet, H. Zhu, B.P. Kremer, S.G. Frints, W.A. van Zelst-Stams, P. Byrne, S. Otto, A.O. Nygren, J. Baets, K. Smets, B. Ceulemans, B. Dan, N. Nagan, J. Kassubek, S. Klimpe, T. Klopstock, H. Stolze, H.J. Smeets, C.T. Schrander-Stumpel, M. Hutchinson, B.P. van de Warrenburg, C. Braastad, T. Deufel, M. Pericak-Vance, L. Scho€ls, P. de Jonghe, S. Züchner, REEP1 mutation spectrum and genotype/phenotype correlation in hereditary spastic paraplegia type 31, Brain 131 (2008) 1078e1086.

. [5] R. Iansek, The effects of reflex path length on clonus frequency in spastic mus- cles, J. Neurol. Neurosurg. Psychiatry 47 (10) (1984) 1122e1124.
Marina Picillo, Roberto Erro
Centre for Neurodegenerative Diseases (CEMAND), Department of

Medicine and Surgery, Neuroscience Section, University of Salerno, Salerno, Italy

Renato P. Munhoz, Alfonso Fasano* Morton and Gloria Shulman Movement Disorders Clinic and the Edmond J. Safra Program in Parkinson’s Disease, Toronto Western Hospital and Division of Neurology, University of Toronto, Toronto, Ontario, Canada

Krembil Research Institute, Toronto, Ontario, Canada

* Corresponding author. Division of Neurology, University of Toronto, Movement Disorders Centre, Toronto Western Hospital, 399 Bathurst St, 7 Mc412, Toronto, ON, M5T 2S8, Canada. E-mail address: alfonso.fasano@gmail.com (A. Fasano).

3 August 2017

Parkinsonism and Related Disorders 46 (2018) 95e97

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Correspondence

Severe camptocormia due to myositis of paraspinal muscles as an early manifestation of Parkinson’s disease

This 49-year old male was presented to our clinic in August 2016 with progressive camptocormia, which had begun to develop in May 2016 (see video). Initially, camptocormia would arise after a walking distance of about 200 m. On presentation to our clinic, the patient was unable to maintain an upright posture for more than a minute in a standing position or walking. However, he was able to sit upright for extended periods of time and was asymptom- atic in the supine position. He reported no further symptoms and his past medical history was unremarkable besides a herniated vertebral disc (L5/S1, right side) in 2010. On inquiry, the patient re- ported suffering from hyposmia from the age of 25 and stated that his wife had been complaining about his pronounced movements during sleep, indicating a REM-sleep behavior disorder (RBD). He denied a family history of PD and other movement disorders. The neurological examination revealed a slight desynchronization in finger and toe tapping on the left side in addition to the camptocor- mia and was unremarkable otherwise. There were no signs of mus- cle weakness, atrophy, rigor or tremor in the extremities and notably no indications for dystonia of the abdominal muscles. The patient scored seven points on the MDS-UPDRS Part III. Initial lab- oratory testing showed a slightly elevated creatine kinase (CK) (200 U/l, normal: < 190 U/l) and lactate dehydrogenase (LDH) (249 U/l, normal: < 225 U/l) indicating ongoing muscle loss. A neuropsycho- logical examination revealed minor impairments in the episodic memory and executive tasks compatible with mild cognitive impairment (MCI). Myography of paraspinal muscles identified iso- lated myopathic potentials. Analysis of cerebrospinal fluid (CSF) revealed a blood-brain barrier disturbance (protein 943 mg/l) and negative oligoclonal bands. Extensive laboratory testing delivered negative results for myasthenia gravis, antinuclear (ANA), anti- neutrophil cytoplasmic (ANCA), antineuronal and autoimmune en- cephalitis antibodies (VGKC, GABA, NMDA, AMPA1/2). MRI of the lumbar region showed signs of fatty transformation of the psoas and paraspinal muscles (see Fig. 1a). Consequently, we performed a biopsy of paraspinal muscles. A histological analysis revealed an unspecific myositis with necrosis of individual muscle fibers, upre- gulation of MHC-I and focal, endomysially located, inflammatory infiltrates, dominated by CD8-cells (see Fig. 1b). There were no signs of poly- or dermatomyositis, inclusion body myositis, mito- chondrial myopathy, lipid or glycogen storage diseases. Alpha- Synuclein-staining remained negative. MRI of the brain showed no abnormalities. However, transcranial ultrasonography revealed a moderately hyperechogenic substantia nigra on the right side, indicating an early stage of PD and corresponding to the symptoms on the left half of the body. Furthermore, functional imaging studies of the brain (see Fig. 1c) revealed a severe reduction of striatal

https://doi.org/10.1016/j.parkreldis.2017.11.008

dopamine transporter availability ([123I]FP-CIT-SPECT) and a mod- erate to severe hypometabolism in temporoparietal to lateral occip- ital and, to a lesser extent, frontal cortical areas (right > left)([18F] FDG-PET).

Supplementary video related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.11.008.

Because an L-Dopa-challenge-test had not lead to a significant improvement of symptoms we instead began the immunosuppres- sive treatment with Prednisolon 1mg/kg. The patient initially responded well to treatment, reporting a walking distance of up to 400m without camptocormia after two weeks (see video). Consequently, we initiated a long-term immunosuppressive treat- ment with azathioprin in February 2017.

Supplementary video related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.11.008.

In summary, the aforementioned findings are compatible with PD with MCI (alternatively, MCI due to incipient dementia with Lewy bodies (LBD)). The present case illustrates that paraspinal myositis leading to camptocormia may also arise at a young age and antedate motor symptoms of PD, contrasting the usual symp- tom onset during an advanced stage of the disease [1]. We were not able to find any other reports about camptocormia as the pre- senting symptom in PD or LBD.

Heterogenic findings in electromyography and unreliable re- sponses to dopaminergic medication and corticosteroids suggest diverse etiologies of camptocormia in PD [2e5]. Among others, Djaldetti et al. discuss a primary myopathy of the paraspinal mus- cles of hitherto unknown origin as well as a strategic damage to the somatotopically structured striatum and putamen [3]. In the pre- sent case, the unusually young age of the patient, the mild motor symptoms, the histological findings including CD8 dominated inflammation of muscle fibers and the positive response to cortico- steroids point towards an autoimmune-mediated disorder. Howev- er, the causal connection of PD and autoimmune-mediated myositis remains unclear. Future studies should focus on the prevalence of (paraspinal) myositis antedating PD to establish a temporal connec- tion, which may help us to understand their complex relationship.

Authors’ contributions

Each author contributed significantly to the submitted work: Drs. Fritsch, Klebe, Weiller, Rijntjes, Lambeck, Schro€ter and Whit- taker treated the case-patient at the Freiburg center and drafted the manuscript. Dr. Meyer performed the SPECT and PET studies,

96 Correspondence / Parkinsonism and Related Disorders 46 (2018) 95e97

Fig. 1. a: Biopsy of paraspinal muscle showing upregulation of MHC I and focal, CD8-dominated inflammatory infiltrates. Arrow: Atrophy of a muscle fiber. Fig. 1b: MRI-imaging shows streaky signal increase in T1-and T2-weighted sequences along the quadratus lumborum muscle and corresponding thinning of paravertebral muscles in the stir-/T2tirm- sequence with signs of fatty degeneration. No substantial change over a period of two years (2014e2016). Fig. 1c [18F]FDG-PET (AeB) and [123I]FP-CIT-SPECT (C) imaging studies of cerebral glucose metabolism and striatal dopamine transporter availability, respectively. A, transaxial [18F]FDG-PET images; B, three-dimensional stereotactic surface projections of [18F]FDG uptake (left) and its statistical deviation (Z score; right) from age-matched healthy control subjects; C, transaxial SPECT images of the patient (left) and an age-matched healthy control subject (right; shown a parametric images of the distribution volume ratio [DVR], as a quantitative measure of dopamine transporter availability).

Dr. Doostkam analyzed the biopsy and Dr. Guggenberger evaluated the MRI.

Funding sources and conflict of interest

The authors declare that there are no conflicts of interest rele- vant to this work and we have read the Journal’s position on issues involved in ethical publication and affirm that this work is consis- tent with those guidelines.

Financial disclosure

Dres. Rijntjes, Whittaker, Schro€ter, Guggenberger, Doostkam, Fritsch and Lambeck declare that there are no additional

disclosures to report. Dr. Klebe received travel grants from Abbvie and Bayer and speaker fees from UCB pharmaceuticals, Dr. Meyer received financial support for research projects by GE and Piramal, Dr. Weiller received financial support for research projects by the DFG Cluster of excellence.

Acknowledgements

The authors thank their patient for participating in this study.

References

[1] P. Srivanitchapoom, M. Hallett, Camptocormia in Parkinson’s disease: defini- tion, epidemiology, pathogenesis and treatment modalities, J. Neurol.

Neurosurg. Psychiatry [Internet] 87 (1) (2016) 75e85. Available from: http://

jnnp.bmj.com/content/87/1/75?etoc.

. [2] S. Spuler, H. Krug, C. Klein, I.C. Medialdea, W. Jakob, G. Ebersbach, et al., Myop-
athy causing camptocormia in idiopathic Parkinson’s disease: a multidisci-
plinary approach, Mov. Disord. 25 (5) (2010) 552e599.

. [3] E. Melamed, R. Djaldetti, Camptocormia in Parkinson’s disease, J. Neurol. 253
(SUPPL. 7) (2006) 14e16.

. [4] S. Wunderlich, I. Csoti, K. Reiners, T. Günthner-Lengsfeld, C. Schneider,
G. Becker, et al., Camptocormia in Parkinson’s disease mimicked by focal
myositis of the paraspinal muscles, Mov. Disord. 17 (3) (2002) 598e600.

. [5] S. Nakane, M. Yoshioka, N. Oda, T. Tani, K. Chida, M. Suzuki, et al., The charac- teristics of camptocormia in patients with Parkinson’s disease: a large cross- sectional multicenter study in Japan, J. Neurol. Sci. [Internet]. Elsevier B.V. 358 (1e2) (2015) 299e303. Available from: https://doi.org/10.1016/j.jns.2015.
09.015.
Konrad Whittaker*, Nils Schroeter, Michel Rijntjes Department of Neurology and Neuroscience, Medical Center – University of Freiburg, Breisacher Straße 64, D-79106 Freiburg, Germany
Konstanze Guggenberger
Department of Neuroradiology, Medical Center – University of Freiburg, Breisacher Straße 64, D-79106 Freiburg, Germany
Johann Lambeck, Brita Fritsch, Cornelius Weiller

Department of Neurology and Neuroscience, Medical Center – University of Freiburg, Breisacher Straße 64, D-79106 Freiburg, Germany

Philipp T. Meyer

Department of Nuclear Medicine, Medical Center – University of Freiburg, Breisacher Straße 64, D-79106 Freiburg, Germany

Soroush Doostkam

Department of Neuropathology, Medical Center – University of Freiburg, Breisacher Straße 64, D-79106 Freiburg, Germany

Stephan Klebe

Department of Neurology and Neuroscience, Medical Center – University of Freiburg, Breisacher Straße 64, D-79106 Freiburg, Germany

* Corresponding author. E-mail address: konrad.whittaker@uniklinik-freiburg.de (K. Whittaker).

3 August 2017

Correspondence / Parkinsonism and Related Disorders 46 (2018) 95e97 97

Parkinsonism and Related Disorders 46 (2018) 98e99

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Correspondence

Celebrating thirty years of deep brain stimulation in movement disorders patients: A successful marriage between neurologists and neurosurgeons

In February 1987, a 23-year-old woman who suffered since childhood from generalized dystonia secondary to an encephalop- athy underwent deep brain stimulation (DBS) surgery of the right ventralis intermedius (Vim) thalamic nucleus with a moderate response during the first two years [1]. This intervention occurred in Grenoble, France, and was performed by the neurosurgeon Alim Louis Benabid and the neurologist Pierre Pollak [2]. A few months later, a 46-year-old patient with tremor-dominant Parkinson’s dis- ease (PD) underwent the same surgery [3]. This patient is still alive and benefiting from DBS for PD, thus achieving almost 30 years of life with continuous brain stimulation.

The year 1987 started a new era in the treatment of movement disorders. The idea of using electrodes to stimulate the brain was not new. The stimulation as a tool for the exploration of brain areas prior to ablation was proposed after the introduction of human ste- reotactic surgery in the early 1950s [4]. Benabid and co-workers had the merit to introduce the Vim-DBS as a treatment for tremor using continuous high frequency stimulation. Their intuition that frequency above 100 Hz applied to the Vim could mimic the effects of a lesion but with the advantage to be completely reversible was a major step towards the application of DBS as a therapeutic tool.

The patient 1 (acquired dystonia), after an initial benefit, pro- gressively lost the DBS effect. At 6-year follow-up the pulse gener- ator drained out and it was not replaced due to the lack of DBS efficacy. Patient 2 (Parkinson’s disease) continued to benefit from Vim-DBS for 10 years. Although Vim-DBS has proven efficacy for parkinsonian tremor, it does not significantly improve bradykine- sia, rigidity, or dyskinesia. Indeed, despite a very good tremor con- trol from Vim-DBS, the patient had a progressive worsening of the other motor signs. In 1997, this patient was operated on with bilat- eral subthalamic nuclei DBS (STN-DBS) and the Vim-DBS was turned off. Indeed, advances in understanding the basal ganglia pathophysiology moved the interest to targeting the STN for treat- ing PD. It was in 1993 that a PD patient underwent the first STN DBS in the world, again in Grenoble [5].

The popularity of DBS technique spread quickly among the neurological and neurosurgical community. Several centers in France and neighboring countries began to refer patients to Greno- ble, requesting the “new brain stimulation therapy able to improve refractory tremors” (Fig. 1). From 1991 on, after the publication in

https://doi.org/10.1016/j.parkreldis.2017.11.007

The Lancet reporting on the effects of DBS in 32 tremor patients [2], other centers around the world also began the DBS adventure. In the following years, DBS stimulation was applied to other neurolog- ical diseases (Tourette’s syndrome and epilepsy) and psychiatric diseases (depression, obsessiveecompulsive disorder and post- traumatic stress disorder). To date, about 150,000 patients world- wide have been treated with DBS for disabling movement disorders.

Technological developments have also greatly contributed to the success of DBS. During the first surgeries, ventriculography was the imaging technique applied to set the surgical coordinates. Currently, high-resolution imaging and segmentation of the brain structures play a critical role in DBS surgery accuracy as well as pro- vide detailed functional datasets which can be used to investigate the connectivity properties of DBS targets. New techniques of stim- ulation, such as the use of directional electrodes to control the steering of current through the brain and leads multiple contacts, are also available. Even more interesting seems to be the closed- loop – “on demand”- stimulation, where the stimulation is auto- matically adjusted according to the variations in the neural signals over the time. Taken all together, these technological advances could tailor DBS according to each patient’s needs (“precision DBS”).

Although DBS significantly impacts the outcome of some dis- eases, the dissemination of this information is still suboptimal. The expected number of patient candidates for DBS is lower compared to the actual number being referred for surgery. Global economics is another important factor to point out; the exponential growth of DBS around the world has not been homogeneous in these 30 years as the therapy and participating research centers have been restricted to a limited number of countries.

History has shown that DBS has revolutionized the treatment of movement disorders starting three decades ago. Evolution in un- derstanding brain pathology and biomarkers of disease progression will contribute to the research of new brain targets as well as to the development of new techniques. The “DBS revolution” has been possible because of the productive and uninterrupted teamwork between neurologists and neurosurgeons. This fertile marriage, supported by a multidisciplinary cooperation, will still be the key of success in the upcoming years.

Fig. 1. Letters from several European countries referring patients to deep brain stim- ulation surgery after the initial case in 1987.

Study funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Disclosures

Dr. R.G. Cury reports no disclosures.
Dr. V. Fraix reports no disclosures.
Dr. E. Moro has received honoraria from Medtronic for lecturing

and scientific board services. She has also received research grant support from Merz.

Acknowledgements

The authors thank all past and present staff members of the Gre- noble’s team who have been involved in the DBS program over

these 30 years: C. Ardoin, A-L Benabid, A. Benazzouz, A. Bichon, A. Castrioto, S. Chabardes, E. Chevrier, B. Debu, A. Kistner, A. Koud- sie, P. Krack, P. Limousin, E. Lhomme e, P. Pelissier, J-E Perret, B. Pial- lat, P. Pollak, E. Schmitt, and E. Seigneuret.

References

. [1] L.Vercueil,P.Pollak,V.Fraix,E.Caputo,E.Moro,A.Benazzouz,J.Xie,A.Koudsie, A.L. Benabid, Deep brain stimulation in the treatment of severe dystonia, J. Neu- rol. 248 (2001) 695e700.

. [2] A.L. Benabid, P. Pollak, C. Gervason, D. Hoffmann, D.M. Gao, M. Hommel, J.E. Perret, J. de Rougemont, Long-term suppression of tremor by chronic stim- ulation of the ventral intermediate thalamic nucleus, Lancet 337 (1991) 403e406.

. [3] A.L. Benabid, P. Pollak, A. Louveau, S. Henry, J. de Rougemont, Combined (tha- lamotomy and stimulation) stereotactic surgery of the VIM thalamic nucleus for bilateral Parkinson disease, Appl. Neurophysiol. 50 (1987) 344e346.

. [4] E. Faggiani, A. Benazzouz, Deep brain stimulation of the subthalamic nucleus in Parkinson’s disease: from history to the interaction with the monoaminergic systems, Prog. Neurobiol. 151 (2017) 139e156.

. [5] P. Pollak, A.L. Benabid, C. Gross, D.M. Gao, A. Laurent, A. Benazzouz, D. Hoffmann, M. Gentil, J. Perret, Effects of the stimulation of the subthalamic nucleus in Parkinson disease, Rev. Neurol. 149 (1993) 175e176.
Rubens Gisbert Cury

Service de Neurologie, Centre Hospitalier Universitaire de Grenoble, Universit e Grenoble Alpes, INSERM U1216, Grenoble, France

Department of Neurology, School of Medicine, University of Sa~o Paulo, Sa~o Paulo, Brazil

Valerie Fraix, Elena Moro* Service de Neurologie, Centre Hospitalier Universitaire de Grenoble, Universit e Grenoble Alpes, INSERM U1216, Grenoble, France

* Corresponding author. Movement Disorders Unit, Department of Psychiatry and Neurology, Centre Hospitalier Universitaire de Grenoble, Universite Grenoble Alpes, BP217 38043 Grenoble CEDEX 09, France. E-mail address: elenamfmoro@gmail.com (E. Moro).

20 October 2017

Correspondence / Parkinsonism and Related Disorders 46 (2018) 98e99 99

Parkinsonism and Related Disorders 46 (2018) 100e101

Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: http://www.elsevier.com/locate/parkreldis

Correspondence
Speech induced cervical dystonia: An unusual task specific dystonia

Keywords:

Task specific dystonia Speech induced Cervical dystonia Dystonia

Dystonia is defined as “sustained or intermittent muscle con- tractions causing abnormal, often repetitive, movements, postures, or both” [1]. Task specific dystonia (TSD) encompasses a group of focal dystonia which affect an isolated body part and are triggered by a specific task. These usually present as involuntary muscle con- tractions which interfere with repetitive skillful motor tasks. Frequently reported TSDs include writer’s cramp, musician’s dysto- nia, sports-related dystonia, and occupational dystonia [2]. This report describes an unusual case of a task specific cervical dystonia, wherein the trigger for onset of dystonia was speech. Involvement of cervical musculature in TSD is uncommon [3,4] and only a single case of speech induced cervical dystonia has been previously re- ported [5].

An eighteen-year-old man presented at our movement disor- ders clinic with a seven-year history of progressive speech distur- bances and a three-year history of progressive abnormal neck movements associated with speech. The speech disturbances were insidious in onset, and characterized by stammering and oc- casional slurring of speech. The frequency and severity of stam- mering increased during periods of stress and anxiety. Speech characteristics such as volume and comprehension were normal. He did not have any difficulty in jaw opening or closing and had no difficulty swallowing. Abnormal movements of the neck were insidious in onset and characterized by involuntary extension of the neck with jerky movements of the neck to either side. These movements were always triggered by initiation of speech. Neck movements did not occur with laughing, whispering, or sustained phonation, and were never present without speech. He was unable to suppress the neck movements and did not describe the presence of an urge associated with the neck movements, and no geste antagoniste was reported. He denied any other abnormal move- ments and his past medical history was unremarkable. There was with no history of exposure to dopamine blocking drugs. The pa- tient was born out of a consanguineous parentage and reported normal birth and developmental history. There was history of “stammering” in 3 family members e paternal aunt, paternal uncle and a paternal cousin. However, none of them reported similar involuntary movements associated with speech and they were not available for examination.

https://doi.org/10.1016/j.parkreldis.2017.11.006

On examination (Video 1), there was no deviation or abnormal contraction of neck muscles observed when the patient was at rest. Range of neck movements and strength of neck muscles were normal. At the onset of speech, the patient developed mild torticollis to the right side, followed by neck extension associated with jerky movements to either side, which lasted for 3-4 seconds following which the head would return to the neutral position. However, the movement recurred if he continued with the attempt to speak. During sustained phonation, there was no voice tremor or presence of concomitant neck movements. The rest of his neurolog- ical examination was unremarkable. Routine blood investigations, thyroid function, serum copper and serum ceruloplasmin were normal. Slit lamp examination for a KaysereFleischer ring was negative and a computed tomography scan of the brain was normal. Video of the patient was taken after informed written consent.

Supplementary video related to this article can be found at https://doi.org/10.1016/j.parkreldis.2017.11.006.

Three cases of task specific cranio-cervical dystonia have been reported, of which only one is a speech-induced cervical dystonia. In the other two cases, persistent cervical dystonia resulted from (a) chronic repetitive right lateral neck flexion secondary to occu- pational requirement of cradling the phone between head and shoulder for extended periods of time, (b) several decades of writing with a pen held in the mouth in a bilateral traumatic arm amputee [4]. The report by Bouchard et al., described findings akin to the present case [5].

Our case presented with involuntary head extension which was triggered by speech. A diagnosis of TSD was supported by the spec- ificity of this movement to articulation and non-occurrence with other forms of phonation or movement of other parts of the body. The diagnosis of a complex motor tic was excluded due to the absence of a premonitory urge and the absence of suppressibil- ity. A psychogenic movement disorder was less likely owing to the absence of distractibility, entrainability and the consistency of movements.

Speech induced dystonia is a rare form of TSD, which is usually reported with involvement of the lip, jaw, tongue, lower face or blepharospasm. The pathophysiology of TSD is unclear and dysfunction of brain plasticity has been suggested. Abnormal cortical inhibition during laryngeal activation may result in a dys- tonic overflow to surrounding musculature [5].

An early onset of cranial dystonia in children and young adults may suggest future development of a combined dystonia syndrome or a neurodegenerative disease [6]. Regular follow up and moni- toring is crucial in patients presenting with similar symptoms.

In conclusion, TSD is a form of focal dystonia triggered by highly

skilled, repetitive tasks which usually involve the hands or the face. Although speech induced cervical dystonia is extremely unusual, it should be considered in patients with a similar phenomenology.

Financial disclosure/conflict of interest
None of the authors have any financial disclosure to make or

have any conflict of interest. Source of funding

Nil.

References

. [1] A.Albanese,K.Bhatia,S.B.Bressman,M.R.Delong,S.Fahn,V.S.Fung,M.Hallett, J. Jankovic, H.A. Jinnah, C. Klein, A.E. Lang, J.W. Mink, J.K. Teller, Phenomenology and classification of dystonia: a consensus update, Mov. Disord. 28 (7) (2013) 863e873.

. [2] C.M. Stahl, S.J. Frucht, Focal task specific dystonia: a review and update, J. Neu- rol. 264 (7) (2017) 1536e1541.

. [3] A. Schramm, M. Naumann, K. Reiners, J. Classen, Task-specific craniocervical dystonia, Mov. Disord. 23 (7) (2008) 1041e1043.

. [4] E. Hogg, M. Tagliati, Overuse cervical dystonia: a case report and literature re- view, Tremor Other Hyperkinet Mov. (N Y) 6 (2016) 413.

[5] M. Bouchard, S. Furtado, Speech-induced cervical dystonia, Can. J. Neurol. Sci. 38 (6) (2011) 929e930.

[6] V.S. Fung, H.A. Jinnah, K. Bhatia, M. Vidailhet, Assessment of patients with iso- lated or combined dystonia: an update on dystonia syndromes, Mov. Disord. 28 (7) (2013) 889e898.

Shweta Prasad, Abhishek Lenka

Department of Clinical Neurosciences, National Institute of Mental Health & Neurosciences (NIMHANS), Hosur Road, Bangalore 560029, Karnataka, India

Department of Neurology, National Institute of Mental Health & Neurosciences (NIMHANS), Hosur Road, Bangalore 560029, Karnataka, India

Pramod Kumar Pal* Department of Neurology, National Institute of Mental Health & Neurosciences (NIMHANS), Hosur Road, Bangalore 560029, Karnataka, India

* Corresponding author. E-mail address: pal.pramod@rediffmail.com (P.K. Pal).

25 August 2017

Correspondence / Parkinsonism and Related Disorders 46 (2018) 100e101 101