Corticobasal Syndrome: Clinical, Neuropsychological, Imaging

CORTICOBASAL SYNDROME: CLINICAL, NEUROPSYCHOLOGICAL, IMAGING,

GENETIC AND PATHOLOGICAL FEATURES

Mario Masellis

A thesis submitted in conformity with the requirements for the degree of Doctorate of

Philosophy in the Graduate Department of Institute of Medical Sciences, University of Toronto

395 Wellington Street 395, rue Wellington Ottawa ON K1A 0N4 Ottawa ON K1A 0N4 Canada Canada Your file Votre référence ISBN: 978-0-494-97827-6

Our file Notre référence ISBN: 978-0-494-97827-6

NOTICE: AVIS: The author has granted a non- L'auteur a accordé une licence non exclusive exclusive license allowing Library and permettant à la Bibliothèque et Archives Archives Canada to reproduce, Canada de reproduire, publier, archiver, publish, archive, preserve, conserve, sauvegarder, conserver, transmettre au public communicate to the public by par télécommunication ou par l'Internet, prêter, telecommunication or on the Internet, distribuer et vendre des thèses partout dans le loan, distrbute and sell theses monde, à des fins commerciales ou autres, sur worldwide, for commercial or non- support microforme, papier, électronique et/ou commercial purposes, in microform, autres formats. paper, electronic and/or any other formats.

The author retains copyright L'auteur conserve la propriété du droit d'auteur ownership and moral rights in this et des droits moraux qui protege cette thèse. Ni thesis. Neither the thesis nor la thèse ni des extraits substantiels de celle-ci substantial extracts from it may be ne doivent être imprimés ou autrement printed or otherwise reproduced reproduits sans son autorisation. without the author's permission.

In compliance with the Canadian Conformément à la loi canadienne sur la Privacy Act some supporting forms protection de la vie privée, quelques may have been removed from this formulaires secondaires ont été enlevés de thesis. cette thèse.

While these forms may be included Bien que ces formulaires aient inclus dans in the document page count, their la pagination, il n'y aura aucun contenu removal does not represent any loss manquant. of content from the thesis. Corticobasal Syndrome: Clinical, Neuropsychological, Imaging, Genetic and Pathological

Features

Doctorate of Philosophy 2012

Mario Masellis

Graduate Department of Institute of Medical Sciences, University of Toronto

ABSTRACT

Corticobasal Syndrome (CBS) is a rare movement and cognitive disorder. There is significant heterogeneity observed in it clinical presentation, neuroimaging, pathology and genetics.

Understanding this heterogeneity is a priority and may help to shed light on underlying pathogenic mechanisms. We first demonstrated that truncating mutations in the progranulin gene

(PGRN) can cause familial CBS associated with frontotemporal lobar degeneration (FTLD)- ubiquitin pathology. This study identified a mutation in PGRN (Intervening Sequence 7+1 guanine > adenine [IVS7+1G>A]) that segregated with CBS in a family. The mutation was predicted to result in a shortened messenger RNA (mRNA) sequence and the absence of the mutant PGRN allele was confirmed in the reverse transcriptase-polymerase chain reaction (RT-

PCR) product, which supported the model of haploinsufficiency for PGRN-linked disease. In a second familial study, clinical, radiological, genetic, and pathological studies were performed to contrast clinical features of the affected members. Sequencing PGRN revealed a novel, heterozygous cytosine-adenine dinucleotide deletion in exon 11 (g.2988_2989delCA,

P439_R440fsX6). The proband`s clinical diagnosis was frontotemporal dementia and parkinsonism (FTDP). The proband‟s brother with the same mutation presented initially as a progressive non-fluent aphasia (PNFA), and later evolved into a CBS. Pathological analysis

ii revealed Frontotemporal Lobar Degeneration-Ubiquitin (FTLD-U)/ TAR DNA-binding protein

43 (TDP43) positive pathology. The next studies shift away from pathogenic mechanisms to focus on brain-behavioural correlations and phenotypic heterogeneity in a prospective sample of

31 CBS cases. We provide the first direct correlative analysis between the severity of ideomotor apraxia, a common sign in CBS, and cerebral SPECT perfusion imaging. Reductions in perfusion within the left inferior parietal lobule (t=5.7, p=0.03, Family-Wise Error [FWE] corrected), including the left angular gyrus (t=5.7, p=0.02, FWE corrected), were associated with more severe ideomotor apraxia. We stratified the sample into CBS presenting with early motor features (CBS-M; n=9) or early dementia (CBS-D; n=22), which identified that CBS-M were more likely to have cortical sensory loss than CBS-D (p=0.005). In contrast, the presence of aphasia was found to be more common and severe in CBS-D compared to CBS-M (p=0.02).

CBS-M patients had significantly reduced perfusion in the right supplementary and premotor areas compared to CBS-D (p<0.05).

iii

ACKNOWLEDGEMENTS

I would like to first and foremost thank my supervisor, Dr. Sandra Black, for her support and mentorship over the many years that I have known her. I first met Dr. Black (a.k.a. Sandy) back in 2001 when I began my residency program in psychiatry. I attended many of her cognitive neurology clinics as an intern and it was this initial exposure to the field of neurology and neurodegenerative disease that made me decide to transfer into the neurology residency program.

The commitment and passion that she displayed towards treating patients and their families afflicted with these devastating diseases was truly an inspiration for me. She taught me that every patient has something unique to offer not only in terms of developing my clinical skills, but also importantly in terms of asking novel questions about the diseases and their heterogeneous presentations that could be assessed using the scientific method and valuable data gathered from clinical and neuroimaging studies. In 2004, I had an opportunity to do my fourth year project course in her lab, which further stimulated me to pursue a career in research and enroll in the

Ph.D. program following completion of my residency. Needless to say, my experiences in her lab have been outstanding. As a result of this training, I have learned a new research method, applied neuroimaging, which I now can add to my repertoire of techniques to use in my clinical and genetic studies. Sandy‟s enthusiasm for research is incredible and her passion to understand and to investigate novel therapies to treat these devastating disorders has also stimulated me to pursue a career as a clinician-scientist. I would also like to thank Sandy for her support of my research ideas, which have led to several peer-reviewed funding projects during and beyond my training. Thank you Sandy for your ongoing support and I look forward to collaborating with you on many interesting projects to come!

I would also like to thank Dr. James Kennedy (a.k.a. Jim) for his mentorship over many years and also for his contribution to and participation on my program advisory committee. Jim stimulated my initial interest in scientific research in the mid-1990s when I completed a M.Sc. in his laboratory. My thesis was on pharmacogenetics and I am pleased to say that I continue this very interesting line of research using the combination of my genetic, clinical and neuroimaging training. I would also like to thank Dr. Robert Chen for his commitment and contributions that he has made as a member of my program advisory committee. I appreciate the efforts of Dr.

Antonio Strafella for reading my thesis under tight time lines and also for being present at my final program advisory committee meeting to participate as an additional examiner.

I would like to thank my friends and colleagues: Dr. Brad MacIntosh, Dr. Kie Honjo, Dr. Galit

Kleiner-Fisman, Dr. Ekaterina Rogaeva, Dr. Anthony E. Lang, Dr. Eric Roy, Isabelle Guimont,

Philip Francis, and Gregory Szilagyi. Their technical and thought-stimulating advice and suggestions have helped to bring me to this point today. I would also like to thank Kayla

Sherborn for her impeccable organizational skills in helping to assemble components of this thesis.

I would like to thank my in-laws for their support over the last few years and also for their assistance in making home life more manageable. I would like to thank my parents for providing me with the opportunities to pursue higher level education and for providing the right environment for me to succeed. I greatly appreciate their continued support and inspiration and I

v am indebted to them for their patience especially over the last few years during the preparation of this thesis.

Last but not least, I would like to thank my beautiful wife, Paola Masellis, for her incredible patience and support that she has provided since the first day that I met her and over the course of the last few years during the completion of this thesis. We have been through a lot together, especially in recent years, and I am indebted to her kindness, love, and caring attitude. With the completion of this thesis, I look forward to many good times ahead and many more years of positive and happy life experiences together with her. Thanks for all that you do!

TABLE OF CONTENTS

TITLE PAGE i ABSTRACT ii ACKNOWLEDGEMENTS iv TABLE OF CONTENTS vii LIST OF TABLES xi LIST OF FIGURES xiii ABBREVIATIONS xvii CONTRIBUTIONS xxv 1.0: GENERAL INTRODUCTION 1 1.1 Corticobasal Degeneration: Historical Perspective 2 1.2 Epidemiology of CBS 4 1.3 Illustrative case examples 4 1.4 Symptoms and signs of corticobasal syndrome 10 1.4.1 Clinical motor and sensory features 11 1.4.2 Clinical cognitive features 14 1.4.3 Apraxia 19 1.5 Neuroimaging in CBS 21 1.5.1 Structural neuroimaging studies 22 1.5.2 Functional neuroimaging studies: PET and SPECT 25 1.6 Pathological heterogeneity in CBS 31 1.7 Genetics of CBS and CBD 37 1.8 Synopsis and overall research objective 42 1.8.1 Specific objectives 43 1.9 Description of chapters 45 1.9.1 Chapter 2: Novel splicing mutation in the progranulin gene causing 45 familial corticobasal syndrome 1.9.2 Chapter 3: Intra-Familial Clinical Heterogeneity due to FTLD-U with 46 TDP43 Proteinopathy Caused by a Novel Deletion in Progranulin Gene (PGRN) 1.9.3 Chapter 4: Ideomotor Apraxia in Corticobasal Syndrome: Brain 46 Perfusion and Neuropsychological Correlates 1.9.4 Chapter 5: Clinical, neuropsychological, MRI and SPECT 47 characterization of a prospective sample of patients with corticobasal syndrome 2.0: NOVEL SPLICING MUTATION IN THE PROGRANULIN GENE CAUSING 48 FAMILIAL CORTICOBASAL SYNDROME 2.1 Summary 49 2.2 Introduction 50 2.3 Methods 53 2.3.1 Subjects 53 2.3.2 Neuropathology 54

vii

2.3.3 Genetic Analysis 54 2.4 Results 55 2.4.1 Clinical features and autopsy results 55 2.4.1.1 Case #4150 (Proband) 57 2.4.1.2 Case #4993 (sister of proband) 61 2.4.1.3 Neuropathology (Case #4993) 61 2.4.2 Genetic analysis 63 2.5 Discussion 64 2.6 Acknowledgements 67 2.7 Addendum 68 3.0: INTRA-FAMILIAL CLINICAL HETEROGENEITY DUE TO FTLD-U WITH 69 TDP43 PROTEINOPATHY CAUSED BY A NOVEL DELETION IN PROGANULIN GENE (PGRN) 3.1 Abstract 70 3.2 Introduction 71 3.3 Materials and methods 72 3.3.1 Subjects 72 3.3.2 Genetic analysis 73 3.3.3 Neuropathological analysis 74 3.4 Results 74 3.4.1 Clinical, neuropsychological, and radiographic features 74 3.4.2 Neuropathology (III:2) 81 3.4.3 Family history 82 3.4.4 Genetic analysis 83 3.5 Discussion 86 3.6 Acknowledgements 89 4.0 IDEOMOTOR APRAXIA IN CORTICOBASAL SYNDROME: BRAIN 91 PERFUSION AND NEUROPSYCHOLOGICAL CORRELATES 4.1 Abstract 92 4.2 Introduction 93 4.3 Materials and methods 96 4.3.1 Subjects 96 4.3.2 Description of neuropsychological measures 97 4.3.3 Brain SPECT acquisition and processing 99 4.3.3.1 Regional perfusion ratios 99 4.3.4 Data analysis 100 4.3.4.1 Demographic, clinical and neuropsychological measures 100 4.3.4.2 Statistical Parametric Mapping (SPM) SPECT analysis 100 4.3.4.3 Region of interest (ROI) SPECT analysis 101 4.3.4.3.1 Comparison of CBS cases to controls 101 4.3.4.3.2 Confirmation of voxel-wise correlation with WAB praxis using 102 ROI method 4.3.4.4 Brain MRI acquisition and processing 102 4.3.4.4.1 Brain extraction and automated tissue segmentation 103 4.3.4.4.2 Post-hoc MRI analysis 103

viii

4.4 Results 104 4.4.1 CBS vs controls 104 4.4.1.1 Demographic data 104 4.4.1.2 Clinical features 105 4.4.1.3 SPM and ROI SPECT analysis 105 4.4.1.4 CBS sample with praxis scores available 107 4.4.1.5 Comparison of apraxic to borderline/non-apraxic CBS patients: 109 Neuropsychological and SPECT analysis 4.4.1.6 Perfusion versus ideomotor apraxia 113 4.4.1.7 Post-hoc atrophy analysis 117 4.5 Discussion 117 4.6 Acknowledgements 127 5.0 CLINICAL, NEUROPSYCHOLOGICAL, MRI, AND SPECT 132 CHARACTERIZATION OF A PROSPECTIVE SAMPLE OF PATIENTS WITH CORTICOBASAL SYNDROME 5.1 Abstract 133 5.2 Introduction 134 5.3 Methods 137 5.3.1 Subjects 137 5.3.2 Neuropsychological, neuropsychiatric, and functional measures 139 5.3.3 Brain MRI 140 5.3.4 Brain SPECT 141 5.3.5 Regional perfusion ratios 142 5.3.6 Pathological analysis 142 5.3.7 Data analysis 143 5.3.7.1 Demographic, clinical and neuropsychological measures 143 5.3.7.2 Region of interest (ROI) SPECT analysis 143 5.3.7.3 Statistical Parametric Mapping SPECT analysis 144 5.4 Results 145 5.4.1 CBS cases versus controls 145 5.4.1.1 Neuropsychological, behavioural and functional assessment 145 5.4.1.2 MRI features 146 5.4.2 Early dementia vs. early motor presentations 149 5.4.2.1 Demographic and clinical characteristics 149 5.4.2.2 Neuropsychological, behavioural and functional evaluation 150 5.4.2.3 MRI features 154 5.4.2.4 SPM and ROI SPECT 155 5.4.3 Description of pathological series and relation to MRI findings 156 5.5 Discussion 158 5.5.1 CBS presenting with early dementia vs. early motor features 158 5.5.2 Pathology 164 5.5.3 MRI investigation 166 5.5.4 Limitations 168 5.5.5 Conclusions 169 6.0 SUMMARY AND GENERAL DISCUSSION 171

6.1 Representative sample 172 6.1.1 Demographic features 172 6.1.2 Clinical and neuropsychological features 173 6.1.3 Neuropsychiatric features 176 6.2 Apraxia in CBS 178 6.3 Comment on the neuroimaging methods 180 6.4 Can CBS serve as a model of etiology for common sporadic disorders? 185 7.0 CONCLUSIONS AND FUTURE DIRECTIONS 188 8.0 REFERENCES 192

LIST OF TABLES

Chapter 2

Table 1 Scores on neuropsychological and functional measures for Page 58 case #4150 compared to standardized scores calculated based on normal population matched for age and years of education.

Chapter 3

Table 1 Raw scores on neuropsychological and functional measures Page 76 for proband (III:1) and proband‟s brother (III:2).

Chapter 4

Table 1 Demographics of patients with corticobasal syndrome and Page 104 control group.

Table 2 Clinical characteristics of CBS sample. Page 105

Table 3 Demographic features of CBS presenting with apraxia vs. Page 109 those without significant apraxia.

Table 4 Mean scores on neuropsychological, neuropsychiatric and Page 111 functional measures in CBS presenting with apraxia vs. those without significant apraxia.

Table 5 Areas of hypoperfusion on SPECT in the CBS group that Page 114 correlate with WAB praxis scores in the regression analyses.

Supplementary Areas of hypoperfusion on SPECT in all CBS patients, CBS Page 130 Table 1 with left side of body most affected, and CBS with right side of body most affected relative to controls.

Chapter 5

Table 1 Case summaries of clinical, pathological, and MRI features of Page 147 CBS patients.

Table 2 MRI atrophy patterns in CBS cases stratified according to Page 148

body side most affected by motor symptoms.

Table 3 Demographic features of CBS groups presenting with early Page 149 dementia versus early motor features.

Table 4 Mean scores on neuropsychological measures in CBS patients Page 151 presenting with early dementia vs. early motor symptoms.

Table 5 Mean scores on behavioural and functional measures in the Page 154 CBS group.

Table 6 MRI atrophy patterns in CBS cases stratified by the presence Page 154 or absence of aphasia as determined by the WAB.

Table 7 Areas of relative hypoperfusion on SPECT in CBS patients Page 156 presenting with early dementia versus those presenting with early motor features.

Table 8 Areas of relative hypoperfusion on SPECT in CBS patients Page 156 presenting with early motor versus those presenting with early dementia.

Supplementary Mean scores on behavioural and functional measures in the Page 170 Table 1 CBS group.

xii

LIST OF FIGURES

Chapter 1

Figure 1 (A) Loss of neurons in the outer layers of the right parietal Page 2 cortex and disorganized cortical architecture in the deeper layers; (B) Swollen, pale neurons with eccentric nuclei in the left superior parietal region.

Figure 2 (A) Brain SPECT showing right parieto-occipital Page 7 hypoperfusion, and (B) T1-weighted MRI showing right > left biparietal atrophy and also a lesser degree of frontal atrophy.

Figure 3 (A) Brain SPECT showing left > right bifrontal hypoperfusion, Page 10 and (B) T1-weighted MRI showing superior left superior frontal > parietal atrophy.

Figure 4 Macroscopic brain specimen showing left frontal > temporal Page 32 atrophy of Pick‟s disease.

Figure 5 Microscopic Lewy body pathology showing Lewy bodies, Page 33 cytoplasmic stippling, neuropil grains and Lewy neurites immunostained by antibodies to alpha-synuclein.

Figure 6 Microscopic pathology of CBD stained with Gallyas Page 35 demonstrating (A) oligodendroglial coils, (B) neuronal pre- tangles in the precentral region, (C) ballooned neurons, and (D) astrocytic plaques in the basal ganglia.

Figure 7 Microscopic agyrophilic grain disease pathology showing (A) Page 36 branched astrocytes in the amygdale, and (B) agyrophilic grains and coiled bodies in the prosubiculum.

Figure 8 Microscopic Alzheimer‟s pathology showing (A) astrocytic Page 36 plaques in frontal regions, and (B) neurofibrillary tangles in the CA1 region of the hippocampus.

Figure 9 Schematic representation of the MAPT genomic region and 3- Page 39 repeat and 4-repeat Tau transcripts.

Figure 10 (A) H1 and H2 linkage disequilibrium blocks showing a 900 Page 41 kb region of inversion, and (B) sub-structure of the MAPT gene and associated H1 and H2 haplotypes.

xiii

Chapter 2

Figure 1 (A) The pedigree structure of the Canadian family showing the Page 56 inheritance of the disease (with age-at-onset). (B) Genomic DNA (gDNA) and RT-PCR (cDNA) sequence fluorescent chromatograms around the PGRN mutation (IVS7+1G>A) observed in the patients and the sequence around common synonymous variation rs25646; (C) An agarose gel photograph of the PGRN product from RT-PCR, using RNA obtained from white blood cells of the affected family member (#4150) and normal control.

Figure 2 Corresponding (A) T1-weighted Magnetic Resonance Imaging Page 59 (MRI) and (B) Technetium 99m-ethyl cysteinate dimer (99mTc- ECD) Single Photon Emission Computed Tomography (SPECT) scans of the brain of Case #4150.

Chapter 3

Figure 1 T1-weighted brain MRI and corresponding 99mTc-ECD brain Page 79 SPECT images of proband‟s brother (III:2) in radiographic axial orientation- Session 1

Figure 2 T1-weighted brain MRI and corresponding 99mTc-ECD brain Page 79 SPECT images of proband‟s brother (III:2) in radiographic axial orientation - Session 2

Figure 3 T1-weighted brain MRI and corresponding 99mTc-HMPAO Page 79 (800MBq) brain SPECT images of proband (III:1) in standard radiographic axial orientation.

Figure 4 Micrographs demonstrating a large number of TDP43 Page 82 inclusions found in the fascia dentata, substantia nigra, and CA1 region.

Figure 5 Detection of PGRN mutation P439_R440fsX6. A) Pedigree Page 84 showing family history of neurodegenerative condition. B) Electropherogram showing start of deletion marked with an arrow.

Figure 6 Amplification from genomic DNA (gDNA; lane 1) using Page 85 primers specific for the mutant allele demonstrate the mutant

xiv

fragment of 153 bp as expected. Amplification from cDNA (lane 2) shows an absence of the expected product supportive of non-sense mediated decay.

Chapter 4

Figure 1 Statistical parametric maps (SPM) of bilateral frontal, parietal Page 106 and temporal surface regions of the brain showing decreased perfusion in (A) all CBS cases compared to controls and (B) CBS cases with predominant symptoms on their left side (CBS-L) compared to controls overlaid on brain MRI template.

Figure 2 Frequency of different aphasia categories on the Western Page 113 Aphasia Battery (WAB) distributed according to the CBS group with apraxia versus those with borderline/no apraxia.

Figure 3 Statistical parametric map of surface regions of the brain Page 115 showing decreased perfusion in the left inferior parietal region, including the angular gyrus, that correlate with WAB praxis scores in the regression analyses.

Supplementary Mean proportion of different MRI tissue classes underlying the Page 128 Figure 1A FWE-corrected SPM mask.

Supplementary Mean proportion of different MRI tissue classes underlying the Page 129 Figure 1B FDR-corrected SPM mask.

Chapter 5

Figure 1 Normalized (z-) scores of neuropsychological measures in Page 145 patients with CBS compared to control group.

Figure 2 Frequency of (A) extrapyramidal and (B) cortical features of Page 150 CBS patients presenting with early dementia vs. early motor symptoms.

Figure 3 Frequency of CBS patients with early dementia vs. early motor Page 153 presentation stratified according to category on the Western Aphasia Battery (WAB).

Figure 4 Statistical parametric maps overlaid on multi-slice brain MRI Page 155 template showing decreased perfusion in left fusiform gyrus (uncorrected p<0.001) in CBS cases presenting with early dementia versus early motor features.

xvi

ABBREVIATIONS

β-CIT 2-β-carbomethoxy-3-β-(4-iodophenyl)-tropane

C Degrees Celsius

μg Micrograms

18F-dopa 18F-6-fluorodopa

3R Three repeat

4R Four repeat

99mTc-ECD Technetium-99m ethyl cysteinate dimer

A Adenine

ABA-2 Apraxia Battery for Adults-2

ACTB Beta Actin

AD Alzheimer‟s disease

ADL Activities of Daily Living

AGD Agyrophilic Grain Disease

AIR Automated Image Registration package

ANCOVA Analysis of covariance

ANOVA Analysis of variance

ANT Anterior

AOO Age of onset

APOE Apolipoprotein E

APX Apraxia

AQ Aphasia quotient

AT Anterior temporal

BAs Brodmann Areas

BNT Boston naming test

xvii

BOLD Blood Oxygen Level Dependent bp Base pair bvFTD Behavioral variant of frontotemporal dementia

C Cytosine

CAA Cerebral Amyloid Angiopathy

CBD Corticobasal degeneration

CBS Corticobasal syndrome

CBS-D Corticobasal syndrome presenting with early dementia

CBS-L Corticobasal syndrome cases with left-sided symptoms

CBS-M Corticobasal syndrome presenting with early motor features

CBS-R Corticobasal syndrome cases with right-sided symptoms cDNA Complimentary deoxyribonucleic acid

CDR Clinical Dementia Rating

CHMP2B Chromatin-modifying protein 2B

CJD Creutzfeldt-Jakob disease cm Centimeters

Cog Cognitive Neurology Clinic

CSDD Cornell Scale for Depression in Dementia

CSF Cerebrospinal fluid

CT Computerized Tomography

CVLT California Verbal Learning Test

D Aspartic acid

D2 Dopamine D2 receptor

DAD Disability Assessment for Dementia

DAT Dopamine transporter

xviii

DEM Dementia onset

D-KEFS Delis-Kaplan Executive Function System

DLB Dementia with Lewy Bodies

DNA Deoxyribonucleic acid

DRS Dementia Rating Scale

DSM-IV Diagnostic and Statistical Manual - IV

DTI Diffusion tensor imaging

EEG Electroencephalograph

F Female

FAS F-, A-, S-word phonemic fluency

FBI Frontal Behavioural Inventory

FDG Fluorodeoxyglucose

FDR False discovery rate

FLAIR Fluid Attenuated Inversion Recovery fMRI Functional Magnetic Resonance Imaging

FOV Field of View

Fr Frontal

FTD Frontotemporal dementia

FTLD Frontotemporal Lobar Degeneration

FTDP Frontotemporal dementia and parkinsonism

FTDP-17 Frontotemporal dementia with parkinsonism linked to chromosome 17

FTLD-U Frontotemporal Lobar Degeneration-Ubiquitin

FWE Family-Wise Error

FWHM Full width at half maximum

xix g Gram

G Guanine gDNA Genomic deoxyribonucleic acid

Gen Generalized

GLM General linear model

GRN Granulin

GWAS Genome wide association studies

HMPAO Hexamethylpropyleneamine Oxime iADL Instrumental Activities of Daily Living

IBZM 123I-iodobenzamide

IF Inferior frontal

IP Inferior parietal

IMA Ideomotor apraxia

IMP N-isopropyl-p[123I]iodoamphetamine

IVS7+1G>A Intervening Sequence 7+1 guanine > adenine

L Left

L-dopa Levodopa

LD Linkage disequilibrium

LFB Luxol fast blue

LKA Limb-kinetic apraxia

M Male

MANCOVA Multivariate analysis of covariance

MAPT Microtubule-Associated Protein Tau mCi Millicurrie

Mb Megabases

xx mBq Megabecquerel

MD Movement Disorders Clinic

MDRS Mattis Dementia Rating Scale

Min Minutes miRNA Micro ribonucleic acid mL Milliliter mm Millimeter

MMSE Mini Mental Status Examination

MND Motor neuron disease

MNI Montreal Neurological Institute

Motor Motor onset

MR Magnetic Resonance mRNA Messenger ribonucleic acid

MRI Magnetic Resonance Imaging ms Millisecond

MSA Multiple system atrophy n Sample size

N/T Not testable nAPX Those without significant apraxia

NART-R National Adult Reading Test-Revised

NCO Normal cut-off

NEX Number of excitations

NPI Neuropsychiatric Inventory

O Occipital

OMIM On-line Mendelian Inheritance in Man

xxi p Probability value

P Parietal

P301S Proline301Serine

PCR Polymerase Chain Reaction

PD Parkinson‟s disease

PET Positron Emission Tomography

PGRN Progranulin

PNFA Progressive non-fluent aphasia

POST Posterior

PPA Primary Progressive Aphasia

PSEN1 Presenilin 1

PSP Progressive Supranuclear Palsy

PT Posterior temporal

Q-Q Quantile-Quantile

R Right rCBF Regional cerebral blood flow

RNA Ribonucleic acid

ROI Regions of interest

RT-PCR Reverse transcriptase-polymerase chain reaction

SD Standard deviation

Sec Second

SEM Standard Error of Mean

SF Superior frontal

SNCA Alpha-synuclein

SNPs Single Nucleotide Polymorphisms

xxii

SP Superior parietal

SPECT Single-Photon Emission Computed Tomography

SPGR Spoiled gradient

SPM Statistical Parametric Mapping

SPM5 Statistical Parametric Mapping version 5

SPSS Statistical Package for the Social Sciences

SS Scaled Score

SYM Symmetrical

T Thymine

T2/PD T2/Proton density

TDP43 TAR DNA-binding protein 43

Te Temporal

TE Echo time

TMEM106B Transmembrane protein 106B

TMT-A Trail Making Test A

TMT-B Trail Making Test B

TOLA Test of Oral and Limb Apraxia

TR Repetition time

TRODAT [2-[[2-[[[3-(4-chlorophenyl)-8-methyl-8-azabicyclo[3.2.1]oct-2- yl]methyl](2mercaptoethyl)amino]ethyl]amino]ethanethiolato(3-)- N2,N2‟,S2,S2‟]oxo-[1R-(exo-exo)]- [99mTc] technetium) ([99mTc]TRODAT-1)

VBM Voxel-based morphometry

WAB Western Aphasia Battery

WCST Wisconsin Card Sort Test

WMH White Matter Hyperintensities

xxiii

WMS-R Wechsler Memory Scale-Revised

ZS Z-score

xxiv

CONTRIBUTIONS

Chapter 2.0 Novel splicing mutation in the progranulin gene causing familial corticobasal syndrome

Mario Masellis,* Parastoo Momeni,* Wendy Meschino, Reid Heffner Jr., Christine Sato, Yan

Liang, Peter St George-Hyslop, John Hardy, Juan Bilbao, Sandra Black, and Ekaterina Rogaeva

As published in: Brain (2006); 129: 3115-3123

Mario Masellis extracted the clinical information on all family members, interpreted and integrated the clinical, neuropsychological, neuroimaging, genetic and pathological data, and was responsible for writing the manuscript. Sequencing and genotyping was performed by Parastoo

Momeni and Ekaterina Rogaeva. Pathological analysis was done by Reid Heffner Jr. and Juan

Bilbao.

Chapter 3.0 Intra-Familial Clinical Heterogeneity due to FTLD-U with TDP43

Proteinopathy Caused by a Novel Deletion in Progranulin Gene (PGRN)

Tomasz Gabryelewicz*, Mario Masellis*, Mariusz Berdynski*, Juan M. Bilbao, Ekaterina

Rogaeva, Peter St. George-Hyslop, Anna Barczak, Krzysztof Czyzewski, Maria Barcikowska,

Zbigniew Wszolek, Sandra E. Black and Cezary Zekanowski As published in: J Alzheimers Dis

(2010); 22: 1123-1133.

xxv

Gabryelewicz. Sequencing of the proband and genotyping of the controls was performed by

Mariusz Berdynski and Cezary Zekanowski. Ekaterina Rogaeva performed the genotyping in the brother of the proband. Pathological analysis was done by Juan Bilbao.

Chapter 4.0 Ideomotor Apraxia in Corticobasal Syndrome: Brain Perfusion and

Neuropsychological Correlates

Mario Masellis, Philip L. Francis, Kie Honjo, Bradley J. MacIntosh, Isabelle Guimont, Gregory

M. Szilagyi, Wendy R. Galpern, Galit Kleiner-Fisman, James L. Kennedy, Robert Chen, Eric A.

Roy, Curtis B. Caldwell, Anthony E. Lang, Sandra E. Black. As submitted to: Cortex

Mario Masellis clinically assessed several of the patients included in this study, extracted the clinical information, designed the study, performed the data analysis and wrote the manuscript.

Philip Francis assisted with the SPM analysis of SPECT data and Kie Honjo assisted with the

MRI segmentation procedure. Bradley J. MacIntosh assisted with the atrophy correction procedure. Wendy R. Galpern, Galit Kleiner-Fisman and Anthony E. Lang assessed and collected clinical data on patients ascertained from a movement disorders clinic.

Chapter 5.0 Clinical, neuropsychological, MRI and SPECT characterization of a prospective sample of patients with corticobasal syndrome

Mario Masellis, Philip L. Francis, Isabelle Guimont, Wendy Galpern, Juan Bilbao, Kie Honjo,

Fuqiang Gao1, Gregory Szilagyi, Farrell Leibovitch, James L. Kennedy, Galit Kleiner-Fisman,

Lisa Ehrlich, Robert Chen, Anthony E. Lang, Sandra E. Black

xxvi

Mario Masellis clinically assessed several of the patients included in this study, extracted the clinical information, designed the study, performed the data analysis and wrote the manuscript.

Philip Francis assisted with the SPM analysis of SPECT data and Kie Honjo assisted with an independent visual read of the MRI data. Wendy R. Galpern, Galit Kleiner-Fisman and Anthony

E. Lang assessed and collected clinical data on patients ascertained from a movement disorders clinic. Juan Bilbao performed the neuropathological analysis.

xxvii

1.0 General Introduction

1.1 Corticobasal Degeneration: Historical Perspective

Rebeiz and colleagues described the first case of corticobasal degeneration (CBD) in 1967 and subsequently characterized three cases from the clinical and pathological perspective in 1968 in their seminal paper „Corticodentatonigral Degeneration with Neuronal Achromasia‟ in which they coined the term based on pathological changes noted in the brain (figure 1) [Rebeiz et al.

1967;Rebeiz et al. 1968].

A B

Figure 1. (A) Loss of neurons in the outer layers of the right parietal cortex and disorganized cortical architecture in the deeper layers, and (B) swollen, pale neurons with eccentric nuclei in the left superior parietal region. Adapted from Rebeiz et al. [Rebeiz et al. 1968]

Since then, many terms have been used to describe this enigmatic disorder of interest to cognitive and movement disorder neurologists worldwide. These include: cortical degeneration with swollen chromatolytic neurons, corticobasal ganglionic, cortical basal ganglionic, and the most common designation, corticobasal degeneration (CBD) [Mahapatra et al. 2004].

Patients suffering from CBD pathology or from the typical clinical syndrome have an insidious onset and gradual decline in function due to a combination of cortical and subcortical/ extrapyramidal clinical features not attributable to other etiologies such as stroke or tumour

[Boeve et al. 2003]. The cortical features may include focal or asymmetric ideomotor apraxia, alien limb phenomenon, cortical sensory loss, visual or sensory hemi-neglect, constructional apraxia, focal or asymmetric myoclonus, and apraxia of speech/nonfluent aphasia. The extrapyramidal features may consist of appendicular rigidity lacking prominent and sustained L- dopa response, and appendicular dystonia. Supportive criteria include cognitive dysfunction with relative preservation of learning and memory on psychometric testing, asymmetric atrophy on computed tomography or magnetic resonance imaging, typically maximal in frontoparietal cortical regions, and asymmetric hypoperfusion or hypometabolism on single-photon emission computed tomography (SPECT) and positron emission tomography (PET), respectively, typically maximal in frontoparietal cortex ± basal ganglia ± thalamus. The clinical syndrome produced by CBD pathology is most often markedly asymmetrical with either left or right hemisphere signs noted in the early stages of the disease although symmetrical cases at onset have been infrequently described [Hassan et al. 2010].

1.2 Epidemiology of CBS

CBS and its most commonly associated underlying pathology, CBD, are extremely rare syndromic/ pathologic entities and, as a result, it is difficult to estimate their true incidence and prevalence. Corticobasal syndrome typically presents in the sixth to eighth decade of life and has a mean age of onset of approximately 63 (standard deviation 7.7) years [Wenning et al. 1998]. It is estimated that CBS accounts for four to six percent of all cases of parkinsonism and, based on the incidence of Parkinson‟s disease, it is speculated that the incidence of CBS lies somewhere between 0.62 to 0.92 per 100,000 per year [Mahapatra et al. 2004;Togasaki and Tanner

2000;Wenning et al. 1998]. Based on the average duration of survival of approximately 7.9 years, prevalence can be estimated at about 4.9 to 7.3 per 100,000 [Mahapatra et al.

2004;Togasaki and Tanner 2000;Wenning et al. 1998]. Despite its rarity, CBS is an extremely interesting syndrome particularly pertaining to the enormous amount of heterogeneity that can be seen on multiple levels including clinical, neuropsychological, neuroimaging, genetic and pathological features. A selection of some of the more common symptoms and signs of CBD will now be illustrated through a review of two representative cases whose clinical and research data were included in the thesis experiments.

1.3 Illustrative case examples

Case 1: CBD with early motor presentation

This 65 year old right-handed woman with no relevant past medical history presented at age 62 with the insidious onset and progressive decline in the use of her left arm. Her presenting complaint was that she could not knit because her “left hand was somewhat awkward.” It would

4 not do what she “wanted it to do.” Shortly thereafter, she noted difficulty using her left hand to cut steak and onions with a knife and fork, de-bone chicken, button up her jacket and fold laundry. She also endorsed troubles with going down stairs. She also noted that she was becoming more “impatient.” She saw a neurologist early on in the disease course at age 62 and was noted on exam to have difficulties with fine motor coordination of her left hand and to a lesser degree her left lower extremity. There was also mild pseudoathetosis of the left fingers.

Otherwise, her neurological exam, including “higher mental functions”, was intact. She was diagnosed with “left upper extremity apraxia” and referred on to a movement disorders clinic where a provisional diagnosis of corticobasal syndrome was made. This was based on history and the emergence of left-sided rigidity and overflow dystonic posturing of the left arm while walking – slight abduction at the shoulder, extended at the elbow and wrist with a clenched fist, in addition to an action tremor, but none at rest on the left. Initial brain MRI and SPECT scans were reported as normal. An EEG demonstrated “non-specific bitemporal slow waves.” A trial of levodopa was initiated for several months with no response; she eventually discontinued it. Her motor symptoms continued to slowly worsen.

Over time, she noted that her left hand and arm “has a mind of its own.” It moved “against” her will and she used her right arm to keep her left in check. She also lost the ability to write with even her right hand. Her husband also endorsed that she was not seeing things as easily in her left visual world. Her medications at this time included amantadine 100 mg tid. Although there were no cognitive issues endorsed by the patient or caregiver, cognitive screening revealed an

MMSE of 21/28 (total score reduced to 28 given that apraxia interfered with tasks involving drawing and writing) with points lost predominantly on attention and delayed recall. A cognitive

5 screening battery revealed difficulties with tasks involving sustained attention, working memory, executive functions and praxis. Delayed verbal recall was impaired, but benefited from cueing.

Neurological exam revealed left greater than right-sided rigidity and paratonia. She had a classic alien-limb phenomenon involving the left upper extremity. Proprioception was reduced on the left and there was bilateral agraphesthesia. There was a left-sided grasp reflex. About two months after this initial visit, she continued to decline with worsening left-sided dystonia and apraxia creating an essentially useless left arm. A repeat brain SPECT revealed decreased perfusion in the right parietal and lateral occipital region (Figure 2). Brain MRI revealed generalized atrophy most prominent in the right posterior region (Figure 2). Neuropsychological testing revealed a preservation of frontal lobe executive function. About seven months later (age

66), she was having increasing difficulties with ambulation requiring a cane and wheelchair for distances. MMSE was 25/28. Shortly thereafter, she lost the ability to ambulate and became wheel-chair bound. She had moderate dysarthria. Rigidity was present in all four limbs although it remained worse on the left. Ideomotor apraxia was also becoming worse in the right hand.

There was also evidence for a mild orofacial and oculomotor apraxia. About eight months later

(age 67) she continued to decline with slower speech, increasing word-finding difficulties, and occasional semantic paraphasia. There were no complaints of memory loss. Her MMSE was

18/28. She developed a classic alien limb phenomenon of her right arm, with constant involuntary grabbing of the left arm and touching of faces. Her rigidity was severe with superimposed spasticity and hyperreflexia. Over the next nine months (age 68), her speech became severely dysarthric and eventually progressed to mutism. Her swallowing also became impaired and she developed recurrent pneumonia presumably on the basis of aspiration. She developed severe, generalized rigidity and it was uncomfortable to move her. Approximately,

6 one year later (age 69) she passed away from respiratory complications related to her neurodegenerative condition. Pathological diagnosis was CBD.

Figure 2. (A) Brain SPECT showing right parieto-occipital hypoperfusion, and (B) T1-weighted MRI showing right > left biparietal atrophy and also a lesser degree of frontal atrophy.

Case 2: CBD with early dementia presentation

This 61 year old woman presented with the insidious onset and gradual decline in expressive language and speech production. Her past medical history was significant for hypertension, diabetes mellitus, and hypercholesterolemia. Her initial cognitive symptoms occurred at age 59 with word-finding difficulties and difficulties putting together sentences. As an example, she occasionally left out verbs and prepositions when speaking, for example, “the dog - the backyard.” There were no complaints of memory loss initially. Her initial MMSE was 28/30. She lost one point on delayed recall and on figure copying. She also had difficulties with serial 7s.

Her language was described as non-fluent. Naming, reading, writing, comprehension and repetition were intact. On this initial assessment, her neurological exam was otherwise normal. A

CT head revealed generalized cerebral atrophy and a brain SPECT revealed mild hypoperfusion in the posterior parietotemporal regions as well as bifrontally. Her initial diagnosis was primary progressive non-fluent aphasia (PNFA).

Approximately one year later (age 62), her problems with fluency progressed. Occasionally, she would substitute in the incorrect word while speaking. She would also repeat words that someone else had said representing echolalia. Her word pronunciation declined and her speech became more strained. Despite these difficulties, she was able to sing along to songs. Even though the patient denied any short-term memory difficulties, her family noted that there was some forgetfulness as she would often not recall things on a grocery list. Apathy and depressive symptoms were present with the patient becoming more withdrawn from interaction with others and also less interested in doing things that she enjoyed. She was frequently tearful. She began to have postural instability with episodes of spontaneously falling backwards. Her MMSE had declined to 18/30 with five points lost on orientation, four points lost on attention/calculation, and one point lost on each of delayed recall, three-step command, and figure-copying. A cognitive screening battery revealed prominent deficits on tasks involving sustained attention, working memory, executive functions and ideomotor praxis. Language assessment revealed strained, effortful speech with paraphasias and decreased spontaneous output. Naming for low frequency words was impaired as was repetition. Comprehension remained relatively spared.

Short-term memory was also impaired, but benefited from cueing. Visuospatial function was relatively preserved. Neurological exam revealed slow, hypometric saccades horizontally and

8 difficulty with eliciting downward saccades. She had a positive grasp reflex bilaterally as well as a snout/pout response. Tone was increased moreso on the right. Strength was within normal limits as were reflexes and plantar responses. There was some evidence for mild bradykinesia on rapid alternating movements. Gait revealed some slowing with decreased arm swing on the right.

On pull-test, there was an absence of the postural reflex; she fell straight backwards. The diagnostic impression at this time was primary progressive non-fluent aphasia evolving into CBS with some features of Progressive Supranuclear Palsy (PSP). Re-assessment three months later revealed continual decline in terms of memory, language, falls, apathy and loss of instrumental activities of daily living. Her exam revealed ongoing troubles with saccadic eye movements and increasing rigidity and bradykinesia on the right greater than left side. A repeat SPECT scan revealed moderate to severe hypoperfusion of the left frontal lobe extending to the left temporal lobe, caudate, and less so to the thalamus (Figure 3). MRI revealed left greater than right-sided atrophy involving the frontal, temporal and parietal regions (Figure 3). Neuropsychological testing revealed deficits across all cognitive domains. Her WAB category was anomic aphasia.

Evaluation approximately eight months later (age 63) revealed worsening expressive language function with preserved comprehension; she could say only one to two words at a time. She continued to have frequent falls. She progressed to dependence on all activities of daily living.

She could only walk if assisted. Her MMSE score was 8/28. She was able to name 5/6 objects, and followed some commands. She had no extraocular movements to command or pursuit, but they were present on vestibular oculoreflex testing. There was increased axial tone with nuchal hyperextension. There was marked rigidity of the right arm and leg with significantly less rigidity on the left. There was a severe ideomotor apraxia on the right greater than left. Re-

9 assessment nine months later (age 64) revealed worsening aphasia; she was now only able to say single word sentences and had difficulties comprehending even simple instructions. Gait had worsened and she required a two-person assist to transfer, and was wheel-chair bound. On exam, the MMSE score was 2/28. The physical exam was unchanged except for worsening rigidity and postural instability. She died at age 65 due to respiratory complications related to the neurodegeneration. Pathological diagnosis was CBD.

Figure 3. (A) Brain SPECT showing left > right bifrontal hypoperfusion, and (B) T1-weighted MRI showing superior left superior frontal > parietal atrophy.

1.4 Symptoms and signs of corticobasal syndrome

The two cases described in the preceding section provide an illustrative account of several common symptoms and signs associated with CBS and also demonstrate the evolution of the

10 clinical syndrome over time. Several classical papers detailing the frequency of clinical signs in

CBS will now be reviewed.

1.4.1 Clinical motor and sensory features

A large prospective study from a movement disorders clinic identified that 64% (23/36) of patients presenting with CBS had “clumsiness of one hand or arm with loss of manual dexterity” as the most common initial complaint [Rinne et al. 1994]. A disturbance of gait due to leg stiffness, jerking, clumsiness, imbalance or combinations thereof, was the next most frequent presenting complaint (28%; 10/26) [Rinne et al. 1994]. Rare initial presentations included prominent sensory symptoms, isolated speech disorder with dysarthria, or a prominent behavioural syndrome [Rinne et al. 1994]. Another early clinical study of 15 patients demonstrated that postural-action tremor, apraxia, limb dystonia or cortical sensory loss were the most frequent initial presenting symptoms [Riley et al. 1990]. Wenning and colleagues

[Wenning et al. 1998] found a similar distribution of the most common clinical signs mentioned above. A retrospective chart review of 147 CBS cases from multiple centers found that rigidity

(92%), apraxia (82%), bradykinesia (80%) and gait disorder (80%) were the most common signs observed in their sample [Kompoliti et al. 1998].

The disorder progressed over time to involve the ipsilateral limb, typically the leg, and then eventually involved the contralateral side usually starting with the arm. With progression, other cortical and extrapyramidal features of the syndrome emerge although many of these signs can also be present early on. Of the extrapyramidal features, asymmetric rigidity and akinesia/

11 bradykinesia that typically do not respond to levodopa were common findings and eventually occurred in all patients [Kompoliti et al. 1998;Riley et al. 1990;Rinne et al. 1994]. Limb dystonia was also a common finding and usually involved the most affected limb with adduction at the shoulder, flexion posturing at the elbow and clawing of the fingers around the adducted thumb into the palm, often with skin breakdown – the so-called “clenched fist” [Rinne et al.

1994;Vanek and Jankovic 2001]. Extension of one or more fingers has also been observed

[Rinne et al. 1994;Vanek and Jankovic 2001]. Limb dystonia has been associated with pain in prior studies [Rinne et al. 1994;Vanek and Jankovic 2001] and may respond to local botulinum toxin injectons into the affected muscles [Cordivari et al. 2001].

Cortical features that involve the limbs also typically present asymmetrically, but will also progress to bilateral involvement over time. Apraxia is the most common cortical feature and will invariably occur in all patients at some point during the course of the disease [Leiguarda et al. 1994;Rinne et al. 1994;Stamenova et al. 2009]. Apraxia will be discussed in more detail in a subsequent section. Cortical sensory loss, manifest as agraphesthesia, astereognosis, sensory extinction, hemi-neglect, and/or loss of two-point discrimination and/or proprioception, presented asymmetrically in several studies [Kompoliti et al. 1998;Mahapatra et al. 2004;Riley and Lang 2000;Rinne et al. 1994]. Irregular jerking (myoclonus) in CBS was focal involving one limb, typically occurred in distal regions, and was elicited with action and/ or evoked by a stimulus [Riley et al. 1990;Thompson et al. 1994]. The myoclonus was cortical in origin as determined by electrophysiological studies demonstrating evidence of enhanced cortical excitability via cutaneous or mixed nerve stimulation [Thompson et al. 1994]. Alien limb phenomenon is a particularly interesting cortical feature whereby the affected limb acts out on its

12 own, sometimes without the patient being aware of its movement and behavior [Riley et al.

1990]. As in our first case, her left hand had a “mind of its own” and it moved “against her will” representing what is now considered to be the true form of alien limb phenomenon [Boeve et al.

2003]. “Levitation” of a limb, originally described by Denny-Brown et al. [Denny-Brown et al.

1952], was thought to originate from lesions in the parietal lobe and should be distinguished from the true alien limb phenomenon. Although both phenomena occur in CBS, levitation is more common than alien limb phenomenon [Riley et al. 1990] and previous studies that have grouped these signs together have likely artificially inflated the frequency of “alien limb” in this syndrome [Kompoliti et al. 1998]. Similar to case 1, asymmetric pyramidal findings including superimposed spasticity, hyperreflexia, and positive Babinski signs have also been observed

[Kompoliti et al. 1998;Mahapatra et al. 2004;Riley et al. 1990;Rinne et al. 1994;Wenning et al.

1998]. Frontal release signs were often present and can be more pronounced on the most affected side of the body [Kompoliti et al. 1998;Mahapatra et al. 2004;Riley et al. 1990;Rinne et al.

1994;Wenning et al. 1998].

Eye movement abnormalities similar to that described in case 2, were observed in 72% (26/36) of CBS cases and were considered supranuclear in nature manifesting as oculomotor apraxia, saccadic (jerky) pursuit movements, and/or restriction in the range of saccadic and pursuit movement vertically more than horizontally [Rinne et al. 1994]. In four cases, a frank limitation of vertical downgaze was noted reminiscent of that observed in progressive supranuclear palsy

(PSP) [Rinne et al. 1994]. Similar findings were noted in other studies [Kompoliti et al.

1998;Mahapatra et al. 2004;Riley and Lang 2000]. The constellation of clinical exam features is variable across individual patients. That is, not all patients manifest every sign that has been

13 associated with CBS. In addition, the body side most affected and the timing in which the different clinical signs present is also variable across patients although over time the signs are present bilaterally.

1.4.2. Clinical cognitive features

The majority of the studies reviewed in the preceding section were conducted in specialist movement disorder clinics. While several of the early studies acknowledged that a few of their cases presented with an early dementia syndrome [Rinne et al. 1994;Wenning et al. 1998], a general conclusion drawn was that early dementia was not a common initial presentation of the

CBS. This viewpoint changed when Grimes et al. [Grimes et al. 1999b] reported that dementia was the most common initial presentation in a case series of patients selected based on having a pathological diagnosis of CBD. In a retrospective review of clinical features of 13 patients with a post-mortem diagnosis of CBD, only four patients had a diagnosis of CBS in life, while six patients had a primary diagnosis of Alzheimer‟s disease and three were diagnosed with atypical dementia (two with frontotemporal dementia or Pick‟s disease and one with dementia and

Parkinsonism) [Grimes et al. 1999b]. In longitudinal follow-up, three of the four cases who presented initially with the classic perceptuomotor disorder went on to develop clinical evidence for dementia [Grimes et al. 1999b]. In addition, 11 of 13 cases with dementia during the disease course developed a motor disorder initially, concurrently or at a later time point and the majority of these patients would have retrospectively met criteria for CBS underscoring the importance of longitudinal follow-up [Grimes et al. 1999b]. The heterogeneity observed in the initial clinical presentation and evolution over time of patients with CBD pathology likely results from the

14 differences in the distribution and severity of the underlying histopathological lesions [Lang

2003]. A subsequent study also identified two patients who presented with a frontotemporal dementia syndrome in life who subsequently were found to have a pathological diagnosis of

CBD demonstrating the overlap of these disorders [Mathuranath et al. 2000].

Several studies have attempted to clarify the nature of the underlying cognitive deficits associated with CBS and CBD pathology. One of the earliest cognitive studies compared the neuropsychological profile of 15 patients with a clinical diagnosis of CBS to that of 19 matched normal controls, as well as to that of patients with PSP (n=15) or Alzheimer‟s disease (AD; n=15) [Pillon et al. 1995]. CBS patients demonstrated a moderate degree of dementia based on global measures of cognition used, such as the Mattis Dementia Rating Scale and Raven‟s

Progressive Matrices. They also demonstrated prominent troubles with executive dysfunction similar to that seen in PSP, but more severe than that observed in Alzheimer‟s disease and this was thought to be due to abnormal function of the frontal-subcortical circuit including damage to the basal ganglia and connections with prefrontal cortical regions [Pillon et al. 1995]. Although mild learning deficits on verbal episodic memory tasks were found in CBS and PSP, the deficits significantly benefited from semantic cueing in contrast to that observed in AD cases, in which both cued recognition and recall were impaired [Pillon et al. 1995]. This finding is also consistent with impaired frontal-subcortical retrieval processes in CBS and PSP compared to prominent hippocampal involvement of encoding and retrieval processes in AD. Similar to patients with PSP, CBS patients demonstrated deficits in dynamic motor execution including difficulties with control and inhibition as well as temporal organization and bimanual coordination [Pillon et al. 1995]. These motor execution deficits were not observed in patients

15 with AD. In contrast, asymmetric ideomotor apraxia was noted mainly in CBS patients reflecting involvement of premotor and parietal regions and was not commonly observed in the PSP or AD groups [Pillon et al. 1995].

Using the Delis-Kaplan Executive Function System, Huey et al. [Huey et al. 2009a] compared

51 patients with behavioural variant FTD and 50 patients with CBS on various standardized measures of executive function and identified MRI correlates within each of the groups. Both groups were more impaired on executive functions compared to their performance on an episodic memory task – the Wechsler Memory Scale-third edition [Huey et al. 2009a]. A between group comparison revealed that FTD patients were significantly more impaired on most executive functions than the CBS group, except for those tasks weighted towards motor and/ or visuospatial abilities, including the Trail Making Test and the two timed measures of the Tower

Test [Huey et al. 2009a]. Within the CBS group, atrophy on MRI in the dorsal frontal, parietal, and temporal-parietal cortical regions in addition to the thalamus was correlated with performance on executive tasks [Huey et al. 2009a]. This study confirms in the largest CBS sample to date ascertained from a single site that executive dysfunction is a prominent feature associated with CBS implicating significant frontal lobe dysfunction in this disorder. Graham et al. [Graham et al. 2003b] reviewed the literature on cognitive dysfunction in CBS and summarized that deficits on frontal lobe tasks such as the Wisconsin Card Sort Test, trail making and initial letter and category fluency were invariably affected across most patients with CBS.

Several studies of CBS have revealed that language impairment is a common cognitive feature of this disorder. Frattali et al. [Frattali et al. 2000] studied 15 patients with a clinical diagnosis of

CBS and found that eight (53%) of these patients had a classifiable aphasia based on a standardized language assessment using the Western Aphasia Battery. Six patients were categorized as having an anomic aphasia, one patient had a Broca‟s aphasia, while one had a transcortical motor aphasia. An additional patient demonstrated an apraxia of speech [Frattali et al. 2000]. MRI scans were assessed visually and the patients with language dysfunction were found to have more frontal, temporal and parietal cortical atrophy as well as subcortical white matter and callosal changes [Frattali et al. 2000]. Another study followed 35 patients with CBS longitudinally, 15 with a motor onset and 20 with a cognitive onset, and observed that 13 patients

(37%) in the cognitive onset group presented initially with a disorder of progressive aphasia

[Kertesz et al. 2000b]. Over longitudinal follow up, all but one patient in the motor onset group, that is, 97% of the sample demonstrated a disorder of language [Kertesz et al. 2000b]. Formal assessment of language using the Western Aphasia Battery was conducted in 21 CBS patients and this demonstrated that patients with cognitive onset had significantly lower scores than the motor-onset group [Kertesz et al. 2000b]. This indicated the presence of more severe forms of aphasia in the cognitive onset group. Graham et al. [Graham et al. 2003a] also performed a detailed assessment of language in a series of ten unselected patients with CBS and demonstrated that eight patients (80%) had language impairment characterized by deficits in phonologic processing and in spelling (orthographic processing). Only two of their patients demonstrated a clinically evident non-fluent aphasia [Graham et al. 2003a]. These important early studies of language function in CBS among others were reviewed and this has lead some authors to conclude that presentation with a progressive apraxia of speech and/ or progressive non-fluent

17 aphasia is strongly associated with the later development of a CBS and may also be predictive of

CBD pathology [Josephs and Duffy 2008].

There have been very few studies examining visuospatial functioning in CBS. Tang-Wai et al.

[Tang-Wai et al. 2003] reported two cases of patients with pathologically proven CBD, who presented initially with a progressive focal visuospatial syndrome and then evolved into a full- blown CBS. A clinical study of 88 patients with atypical parkinsonian syndromes using the

Visual Object and Space Perception battery, including 20 patients with multiple system atrophy,

43 with PSP, and 25 with CBS, demonstrated that only the CBS group had evidence for significant visuospatial dysfunction that was independent of their performance on other cognitive tasks [Bak et al. 2006]. They hypothesized that the observed visuospatial deficit reflects dysfunction of the dorsal visual stream due to involvement of the parietal lobes by the pathologies that can produce CBS [Bak et al. 2006].

The studies discussed in the preceding paragraphs described cognitive and neuropsychological features of patients clinically diagnosed with CBS and only a small proportion of these patients had pathologically confirmed CBD. We will now review the findings of a longitudinal clinical and neuropsychological study of 15 patients with pathologically proven CBD [Murray et al.

2007]. Similar to prior studies, only six patients (40%) had a clinical diagnosis of CBS in life whereas other primary or differential diagnoses included progressive non-fluent aphasia, behavioural variant FTD, Alzheimer‟s disease, atypical dementia, atypical PSP, and dementia with Lewy bodies [Murray et al. 2007]. Using a comprehensive neuropsychological battery, a

18 specific cognitive profile of CBD was identified that included deficits in the performance of gestural, language, visuospatial, executive, and social functioning with relative sparing of episodic memory, even at the late stages of the disease. These neuropsychological deficits correlated with burden of CBD Tau-related pathology in the frontal and parietal regions as well as the basal ganglia with minimal involvement of the temporal lobes and hippocampi [Murray et al. 2007].

1.4.3 Apraxia

In general terms, apraxia is “characterized by loss of the ability to execute or carry out skilled movements and gestures, despite having the desire and the physical ability to perform them”

(http://www.ninds.nih.gov/disorders/apraxia/apraxia.htm). Apraxia is the main clinical feature that distinguishes CBS from other parkinsonian disorders and it is observed in 100% of CBS cases during longitudinal follow-up [Leiguarda et al. 1994;Rinne et al. 1994;Stamenova et al.

2009]. Different types of apraxia have been reported in CBS including subtypes of limb apraxia, such as limb-kinetic apraxia, ideomotor apraxia (IMA), and ideational/conceptual apraxia.

Orofacial apraxia and apraxia of speech have also been observed. A full description of the types of apraxia identified in CBS have been extensively reviewed elsewhere [Gross and Grossman

2008;Josephs and Duffy 2008;Leiguarda and Marsden 2000;Stamenova et al. 2009;Zadikoff and

Lang 2005]. Limb apraxia is the most common type observed in CBS and the remainder of this section will focus on this topic.

Several models of limb apraxia have been described in the literature based on original case studies and series, and the left parietal lobe has been implicated in most [Geschwind

1975;Goldenberg 2009;Heilman and Rothi 1993;Liepmann 1920;Roy 1996]. Although it is beyond the scope of this thesis to describe these models in detail, highlights of several models will be briefly reviewed. One of the earliest models postulated the existence of visual mental images of the intended movement stored in posterior brain regions such as the parieto-occipitotemporal junction in the dominant left hemisphere that are then transferred forward to central sensorimotor regions for the task to be carried out [Liepmann 1920;Goldenberg 2009].

Geschwind alternatively proposed that comprehension of verbal commands to carry out a motor task is achieved in Wernicke‟s area and then is carried forward to the sensorimotor cortex via the arcuate fasciculus passing under the parietal lobes [Geschwind 1975;Goldenberg 2009].

Therefore, damage in the parietal region on the left can result in apraxia through disruption of this circuit [Geschwind 1975]. A more recent neuroanatomical theory of praxis based on the original Liepmann model suggests that „praxicons‟ or „movement formulae‟ are stored in the left inferior parietal lobule, which then are transformed into „innervatory patterns‟ or „motor schema‟ in the premotor and supplementary motor areas, before being decoded by the primary motor cortex to perform motor tasks both ipsilaterally and contralaterally [Heilman 1979;Heilman and

Rothi 1993;Ochipa and Gonzalez Rothi 2000]. An information-processing model of apraxia proposes the existence of three systems: the sensory-perceptual, conceptual, and production system [Roy 1996]. Depending on where damage occurs across these systems, specific praxis deficits will be observed [Roy 1996]. The various pathologies that can produce the CBS localize to the frontoparietal cortex and its subcortical connections and this is thought to be the reason that limb apraxia is so commonly observed in CBS.

As previously mentioned, ideomotor apraxia, limb-kinetic apraxia and, less often, conceptual/ideational apraxia have been the main subtypes of limb apraxia studied in CBS in that order [Gross and Grossman 2008;Stamenova et al. 2009;Zadikoff and Lang 2005]. Ideomotor apraxia is best elicited through voluntary pantomime and/or imitation of hand gestures and tool use and is characterized by disturbances of spatial organization, sequencing and timing of gestural limb movements [Rothi et al. 1991]. Limb-kinetic apraxia (LKA) is defined as a loss of hand and finger dexterity resulting in a breakdown and awkwardness of distal movements [Kleist

1907]. The definitions used for conceptual/ ideational apraxia have been more variable.

Conceptual/ideational apraxia was defined in this thesis as impairment in object/tool or action knowledge [Stamenova et al. 2009]. However, some studies have distinguished between conceptual and ideational apraxia with the latter being defined as a failure to sequence tasks related to tool use correctly. This has resulted in phenomenological/ taxonomic confusion across studies [Stamenova et al. 2009]. More research is required to better localize the regions of the brain involved in limb apraxia associated with the CBS and to better understand the network involved in this phenomenology.

1.5 Neuroimaging in CBS

The cognitive and physical symptoms and signs of the disorder correlate reasonably well with the location of the maximally affected brain regions, which can often be identified in vivo using structural neuroimaging (e.g., brain MRI) and functional neuroimaging (e.g., brain perfusion

SPECT or glucose metabolism PET).

1.5.1 Structural neuroimaging studies

Riley et al. [Riley et al. 1990] conducted one of the earliest clinical studies in a case series of 15 patients with CBS that examined brain computed tomography and MRI imaging. On visual inspection of CT and/ or MRI of the brain, asymmetric atrophy worse contralateral to the most affected side of the body was observed in eight patients, whereas six patients demonstrated symmetric atrophy [Riley et al. 1990]. One patient did not have any notable atrophy on CT of the brain [Riley et al. 1990]. Several years later Yamauchi and colleagues [Yamauchi et al. 1998b] observed that, compared to controls, a group of eight CBS patients had atrophy on MRI of the corpus callosum, which was most severe in the middle-posterior > middle-anterior > anterior > posterior regions. The degree of callosal atrophy also was correlated with glucose metabolism as measured by PET and the latter tended to be asymmetric [Yamauchi et al. 1998b]. Another MRI- based study comparing 16 patients with CBS to 28 patients with PSP demonstrated that atrophy on T1-weighted MRI images was most prominent in frontoparietal regions contralateral to the most affected side of the body in approximately 14/16 (87.5%) of the CBS patients and was not present in any patients with PSP, who demonstrated mainly midbrain atrophy [Soliveri et al.

1999]. This group also observed the presence of cortical and subcortical white matter signal changes involving or underlying the atrophic region on proton density and T2-weighted images in six (37.5%) CBS cases [Soliveri et al. 1999]. Similar findings were observed by a Japanese group that demonstrated that the parietal, anterior middle, and inferior frontal lobes, and paracentral regions were significantly more atrophic and tended to be asymmetric in CBS than in

PSP, whereas the brainstem was more atrophic in PSP using MRI-based hemisphere surface display and volumetry [Taki et al. 2004]. Another study that compared 18 patients with CBS to

33 with PSP found that CBS cases as a group had reduced whole brain volumes, and more selective atrophy involving the parietal lobes and the corpus callosum [Groschel et al. 2004]. The most severe atrophy was observed in the white matter of the parietal lobes; however, in contrast to prior studies, there was no tendency for the atrophy to be localized contralateral to the most affected side of the body [Groschel et al. 2004]. Similar to Soliveri et al. [Soliveri et al. 1999], midbrain atrophy also differentiated PSP from CBS [Groschel et al. 2004]. Finally, using a discriminant function analysis in a subset of the sample with pathologically proven CBD or PSP as well as controls, the combined volumes of the midbrain, brainstem, pons, frontal and parietal white matter and temporal grey matter were found to differentiate the groups with high accuracy

[Groschel et al. 2004]. The first published voxel-based morphometry (VBM) study that compared 14 patients with CBS to 15 with PSP identified that atrophy in CBS was more prominent on the left than the right and involved bilateral premotor regions, superior parietal lobes and the striatum whereas PSP patients had prominent atrophy involving the midline subcortical structures including the midbrain, pons, thalamus and striatum as well as minimal involvement of the frontal lobes [Boxer et al. 2006]. Using a voxel-wise discriminant function analysis, they were able to correctly distinguish between CBS and PSP patients with 93% accuracy by using the severity of atrophy in the dorsal pons, midbrain tegmentum and left frontal eye field [Boxer et al. 2006].

There have been only a few published case series of pathologically proven CBD studied with

MRI. One study examined 17 patients with a clinical diagnosis of CBS of which six had a pathologically confirmed diagnosis of CBD and 11 had other pathological diagnoses including

PSP, FTD, AD, and Creutzfeldt Jakob Disease [Josephs et al. 2004]. Using a semi-quantitative

23 visual assessment of pre-selected regions of interest bilaterally on MRI, they confirmed findings of previous studies that demonstrated atrophy on T1-weighted imaging involving the posterior frontal, superior parietal and middle corpus callosum in both groups and subcortical/ periventricular white matter changes on T2-weighted imaging [Josephs et al. 2004]. However, there was no difference between the MRI findings in the CBD group vs. that with other pathologies suggesting that it is the location and distribution of the pathology and not the specific pathology itself that predicts the CBS [Josephs et al. 2004]. The same group later demonstrated in a larger series of pathologically proven CBD patients (n=11) compared to controls that atrophy predominated in the cortical regions bilaterally including the superior, middle and posterior inferior frontal lobes, the posterior temporal and parietal lobes, and the superior premotor cortex [Josephs et al. 2008]. The insular cortex and supplementary motor area also demonstrated atrophy in CBD patients and subcortical grey matter atrophy was observed in the globus pallidus, putamen and caudate head [Josephs et al. 2008]. There was also a small amount of white matter atrophy identified in the posterior frontal lobes, the corpus callosum, the external capsule and the right midbrain in the CBD group [Josephs et al. 2008].

Several recent studies have employed diffusion tensor imaging (DTI) to better characterize the integrity of the white matter in CBS in order to follow up on prior studies that demonstrated T1- weighted atrophy and T2-weighted hyperintensities of the white matter in this condition. Borroni et al. [Borroni et al. 2008b] compared 20 patients with CBS to 21 normal controls using DTI

MRI and demonstrated reduced fractional anisotropy in the long frontoparietal connecting tracts, the intraparietal associative fibers, and the corpus callosum. Reductions in fractional anisotropy were also observed in the sensorimotor projections of the cortical hand areas [Borroni et al.

2008b]. Another study used tract-based statistics to study 10 patients with CBS and 10 normal controls and found that CBS patients had higher average apparent diffusion coefficient values and lower average fractional anisotropy values in the corticospinal tract in the most affected hemisphere and also in the posterior trunk of the corpus callosum [Boelmans et al. 2009]. The same group has more recently observed that higher mean diffusivity and lower fractional anisotropy within the posterior trunk of the corpus callosum can distinguish CBS from

Parkinson‟s disease [Boelmans et al. 2010]. MRI studies to date demonstrate heterogeneity across CBS patients in terms of the degree and localization of the cortical and subcortical grey matter atrophy observed and also in the involvement of the white matter and this may be, in part, responsible for the variability in clinical presentations.

1.5.2 Functional neuroimaging studies: PET and SPECT

Sawle et al. [Sawle et al. 1991] using PET to measure regional cerebral oxygen metabolism were the first to demonstrate that patients with CBS have hypometabolism predominantly in the posterior and superior temporal, inferior parietal, and occipital (association) cortices; frontal association regions also demonstrated reduced metabolism although they did not achieve statistical significance. This pattern of hypometabolism tended to be asymmetric being lower contralateral to the most affected side of the body. Following this initial study, several other PET studies using fluorodeoxyglucose (FDG) as the tracer demonstrated similar findings. One of the first studies using FDG-PET also demonstrated asymmetric uptake of FDG in five patients with

CBS compared to PD patients and normal controls involving the thalamus, hippocampus and inferior parietal lobule [Eidelberg et al. 1991]. Asymmetry of parietal lobe metabolic reduction

25 of 5% or more was found in the CBS group whereas the PD and normal control groups manifested less than 5% reductions [Eidelberg et al. 1991]. Another study demonstrated significant reductions in FDG uptake in frontal, temporal, sensorimotor and parietal association cortices in five CBS patients compared to controls and additionally showed involvement of subcortical structures including the caudate and lentiform nuclei and thalami; reductions were noted predominantly contralateral to the most affected side of the body [Blin et al. 1992]. Similar findings were observed in a Japanese study with asymmetric involvement of the parietal cortex, including the primary sensorimotor and lateral parietal regions, the caudate, putamen and thalamus contralateral to the most severely affected side in the CBS group [Nagasawa et al.

1996]. Another FDG-PET study compared nine patients with CBS to nine with PSP and observed that CBS patients had significant metabolic reductions involving the inferior parietal, lateral temporal, sensorimotor cortices as well as the striatum that were worse contralateral to the most affected side of the body [Nagahama et al. 1997]. Symmetrical hypometabolism involving the frontal and parietal lobes and thalami has also been demonstrated in some cases even though asymmetry was present on the clinical exam suggesting the presence of heterogeneity in the imaging findings [Taniwaki et al. 1998].

The previous FDG-PET studies described had sample sizes that were small, typically less than

10 CBS patients, and employed mainly region of interest approaches. Garraux et al. [Garraux et al. 2000] conducted a voxel-based analysis of 22 patients with CBS, 21 PSP patients, and 46 healthy controls. They largely confirmed findings of earlier studies demonstrating asymmetric metabolic reductions involving the thalamus, putamen, supplementary motor and lateral premotor areas, the dorsolateral prefrontal cortex and the anterior part of the inferior parietal

26 lobule, which includes the intraparietal sulcus [Garraux et al. 2000]. PSP patients could be differentiated from CBS patients based on metabolic reductions involving the midbrain, anterior cingulate and orbitofrontal regions, whereas CBS patients had reductions in posterior frontal regions including the supplementary motor area as well as the inferior parietal lobule in contrast to PSP [Garraux et al. 2000]. These FDG-PET findings were also observed in several smaller case series using voxel-based approaches [Hosaka et al. 2002;Juh et al. 2005;Klaffke et al.

2006]. Finally, an FDG-PET study using a visual assessment method as opposed to semi- quantitative or quantitative techniques demonstrated the clinical utility of visual assessment in detecting asymmetric hypometabolism involving the peri-rolandic area, striatum and thalamus

[Coulier et al. 2003].

Other PET tracers have been used to characterize patients with CBS. Sawle et al. [Sawle et al.

1991] were the first to report reductions in basal ganglia uptake of 18F-6-fluorodopa (18F-dopa) in

CBS. They found that uptake of 18F-dopa was most reduced in the caudate contralateral to the most affected side of the body in all patients [Sawle et al. 1991]. Putaminal reduction of 18F-dopa uptake was also most prominent contralateral to the most affected side of the body in all but one patient who demonstrated bilateral reductions [Sawle et al. 1991]. This study provided the first in vivo evidence of nigrostriatal dopaminergic denervation, that is, reduction in the number of functioning nigrostriatal dopaminergic neurons, in CBS. Other studies have confirmed the finding of reduced 8F-dopa uptake in CBS [Laureys et al. 1999;Nagasawa et al. 1996].

Perfusion SPECT, using a variety of tracers, has also been used to image series of patients with

CBS and largely show similar cortical and subcortical involvement as that observed with FDG-

PET metabolic studies. Markus et al. [Markus et al. 1995] were the first to demonstrate that, in eight CBS patients compared to controls, markedly reduced perfusion on

Hexamethylpropyleneamine Oxime (HMPAO)-SPECT was present bilaterally, but worse contralateral to the most affected side of the body in subcortical regions including the caudate, putamen and thalamus, and in cortical regions including the posterior frontal cortex and in all divisions of the parietal cortex (anterior, superior, posterior, and inferior). In comparison to PD patients, perfusion was also reduced in the most affected hemisphere in the thalamus, posterior frontal, as well as the anterior and inferior parietal cortices [Markus et al. 1995]. Using N- isopropyl-p[123I]iodoamphetamine (IMP) SPECT in nine patients with CBS, limb apraxia was the most common clinical finding and hypoperfusion contralateral to the most affected limb was most prominent in the sensorimotor cortex and posterior parietal cortex [Okuda et al. 1999]. In a small IMP-SPECT study, asymmetric reductions in regional perfusion were observed in the frontoparietal regions including the inferior prefrontal, posterior parietal and sensorimotor cortices in CBS, but not in PSP [Okuda et al. 2000b]. Medial prefrontal perfusion reductions, however, were seen in both disorders [Okuda et al. 2000b]. Another SPECT study using

Technetium-99m ethyl cysteinate dimer (99mTc-ECD SPECT) as a tracer compared nine patients with CBS to nine with PSP and found that asymmetrical hypoperfusion in the frontal, parietal, and temporal cortex and basal ganglia, as well as, to a lesser degree, the occipital cortex differentiated CBS from PSP [Zhang et al. 2001]. The first ECD-SPECT study of CBS to employ an unbiased, whole brain, voxel-wise analytical technique (statistical parametric mapping; SPM) demonstrated more widespread brain hypoperfusion than previously observed by

28 region of interest studies, including the frontal, parietal and temporal cortices, as well as the basal ganglia, thalamus and pontocerebellar regions [Hossain et al. 2003]. However, to our knowledge, this study did not correct for multiple testing using more modern techniques such as correcting for the family-wise error [Hossain et al. 2003]. Using HMPAO SPECT and a factor discriminant analysis applied to regions of interest, Kreisler et al. [Kreisler et al. 2005] identified seven variables, including the more affected temporoinsular region, the more affected medial frontal region, the less affected and more affected lateral frontal regions, the less affected temporoparietal region, and the lateral frontal and parietal asymmetry indices, that correctly classified patients as having CBS or PD with 100% and 95% accuracy, respectively. A more recent clinical study that performed both MRI and ECD-SPECT in 16 patients with CBS that were read by two neuroradiologists blinded to the diagnosis and clinical information found that

SPECT was more sensitive than MRI in detecting asymmetries [Koyama et al. 2007].

Frisoni et al. [Frisoni et al. 1995] were the first to report reductions in uptake of the SPECT tracer 123I-iodobenzamide (IBZM; binds to post-synaptic dopamine D2 receptors) in the right basal ganglia in a case of CBS with prominent left-sided motor involvement and proposed that reduction in the number of D2 receptors may account for the lack of levodopa responsiveness seen in CBS. This finding was largely refuted by two papers showing that IBZM uptake on

SPECT was mostly normal in most patients with CBS as combined results across the studies demonstrated that only three of 17 patients had reductions in IBZM uptake in the basal ganglia

[Klaffke et al. 2006;Plotkin et al. 2005].

Finally, several studies have demonstrated the value of using dopamine transporter (DAT)-

SPECT imaging in CBS and other parkinsonian disorders. One of the earliest studies used the

DAT-SPECT tracer, 2-β-carbomethoxy-3-β-(4-iodophenyl)-tropane (β-CIT), to compare 18 patients with multiple system atrophy (MSA), eight with PSP, four with CBS and 48 with PD in terms of their striatal binding [Pirker et al. 2000]. They found that all patient groups demonstrated reduced β-CIT striatal binding compared to controls and this tended to be asymmetric in the PD and CBS groups [Pirker et al. 2000]. Another study by the same group followed longitudinal changes in striatal β-CIT striatal dopamine transporter binding over time in

36 patients with PD, 10 patients with atypical parkinsonian syndromes including three CBS cases, and nine patients with essential tremor [Pirker et al. 2002]. They found that the uptake of

β-CIT was reduced in PD and the atypical parkinsonian syndromes, but not in essential tremor compared to controls [Pirker et al. 2002]. They also observed that the β-CIT striatal uptake declined more rapidly in those with atypical parkinsonian syndromes compared to PD [Pirker et al. 2002]. These initial findings of reduced β-CIT striatal uptake in CBS were supported by other studies [Klaffke et al. 2006;Plotkin et al. 2005]. A single case of pathologically proven CBD did not demonstrate any reductions in β-CIT striatal uptake visually after four years from disease onset refuting prior studies [O'Sullivan et al. 2008]. The largest and most recent study using β-

CIT SPECT in 36 patients with CBS, 37 patients with PD and 24 healthy controls demonstrated that striatal binding reduction was variable across CBS cases and more uniformly reduced with more hemispheric asymmetry than that observed in PD [Cilia et al. 2011]. There was also no correlation between striatal β-CIT and clinical features of the disease including severity [Cilia et al. 2011]. Four CBS patients had normal striatal uptake compared to controls, while four had strictly unilateral uptake despite all showing bilateral extrapyramidal signs [Cilia et al. 2011].

Many of the neuroimaging studies described employed small sample sizes and only a few of the studies provided a detailed clinical and neuropsychological characterization of their CBS subjects. Therefore, further neuroimaging studies in a prospective sample that has been well- characterized clinically are required in order to further understand the heterogeneity observed in the presentation of CBS and how this correlates with neuroimaging features.

1.6 Pathological Heterogeneity in CBS

CBS is not only heterogeneous in its clinical presentation and in its neuroimaging as previously described, but there is also substantial pathological heterogeneity that can produce the syndrome.

Ball and colleagues [Ball et al. 1993] described a case of CBS presenting with an alien left limb, memory loss, cortical myoclonus and bilateral parietal dysfunction with a pathological diagnosis of AD. Lang and colleagues [Lang et al. 1994] described a case of pathologically proven

“parietal” Pick‟s disease that presented as corticobasal syndrome. Since these two early studies, there have been several case series published that have demonstrated similar pathological heterogeneity underlying the corticobasal syndrome. This section reviews these case series and provides images that demonstrate the variety of pathologies that have been associated with the

CBS.

Figure 4. Macroscopic brain specimen showing left frontal > temporal atrophy of Pick’s disease

A study of 11 cases with pathologically confirmed CBD identified overlapping pathology of one or more of AD, PD, hippocampal sclerosis and PSP in six cases (54%) suggesting that mixed pathology can be associated with CBD [Schneider et al. 1997]. The cases with mixed CBD and other pathologies all presented with early memory loss in comparison to those with CBD alone and the authors proposed that the mixed pathology may account for the variable clinical presentations observed in CBS [Schneider et al. 1997]. Another study presented clinical vignettes of several different pathologically confirmed cases of neurodegenerative disease, including ten with CBD, to six neurologists who then had to make a clinical diagnosis based on the information that was provided [Litvan et al. 1997]. The accuracy of the clinical diagnosis

32 was then determined and it was observed that the specificity was high at 99.6% meaning that less than 1% of patients without CBD were diagnosed as having it [Litvan et al. 1997]. However, the sensitivity was low at 48.3% meaning that only about 50% of patients were accurately diagnosed with CBD in life [Litvan et al. 1997]. Boeve and colleagues [Boeve et al. 1999] identified 13

Figure 5. Microscopic Lewy body pathology showing Lewy bodies, cytoplasmic stippling, neuropil grains and Lewy neurites immunostained by antibodies to alpha-synuclein patients from the Mayo clinic records with a diagnosis of CBS who also had a neuropathological examination at autopsy and demonstrated that seven cases had a pathological diagnosis of CBD

(53.8%) while six had other diagnoses (46.2%; two with AD, one with Creutzfeldt-Jakob disease

(CJD), one with PSP, one with Pick‟s disease and one with non-specific histopathology) [Boeve et al. 1999]. Frontotemporal lobar degeneration (FTLD) with motor neuron disease-like

33 inclusions, today known as FTLD-Ubiquitin (U)/TAR DNA-binding protein 43 (TDP43) with motor neuron disease (MND), has also been documented to produce the CBS [Grimes et al.

1999a]. CJD has also been observed to cause the CBS [Kleiner-Fisman et al. 2004], as has agyrophilic grain inclusion disease [Rippon et al. 2005]. Another case report identified that bilateral strokes involving the frontoparietotemporal and occipital regions, worse on the right, due to ipsilateral occlusion of the distal internal carotid and middle cerebral arteries and severe stenosis of the left middle cerebral artery was associated with a corticobasal syndrome [Kim et al. 2009].

An important pathological study that screened all archival data from the Queen Square Brain

Bank over a 20 year period identified 19 pathologically confirmed cases of CBD and 21 clinically diagnosed cases of CBS [Ling et al. 2010]. Of the pathologically confirmed cases, only five were accurately diagnosed as having CBD in life yielding a sensitivity of 26.3% [Ling et al.

2010]. Alternative clinical diagnoses were eight cases with PSP, two with PD, two with FTD, one with spastic quadriparesis with myoclonus of unknown etiology, and one incidental case with Tourette‟s syndrome who died before symptoms of CBS manifested [Ling et al. 2010].

From the clinical standpoint, of the 21 cases diagnosed as having CBS in life, only five had confirmed CBD pathology, while the rest had alternative pathological diagnoses including six with PSP, five with AD, two with PD, one with frontotemporal lobar degeneration-

Ubiquitin/TDP43 (FTLD-U/TDP43) with MND, one with FTLD-U/TDP43 subtype 2, and one with dementia lacking distinctive histopathology resulting in a positive predictive value of 23.8%

[Ling et al. 2010]. Finally, a larger study of 18 cases with pathologically proven CBD and 40

34 cases of CBS due to other histopathologies will now be discussed [Lee et al. 2011]. The pathologically confirmed cases of CBD presented with four distinct clinical syndromes including

Figure 6. Microscopic pathology of CBD stained with Gallyas demonstrating (A) oligodendroglial coils, (B) neuronal pre-tangles in the precentral region, (C) ballooned neurons, and (D) astrocytic plaques in the basal ganglia executive-motor (n=7; 38.9%), progressive non-fluent aphasia (n=5; 27.8%), behavioural variant

FTD (n=5; 27.8%), and posterior cortical atrophy (n=1; 5.5%) [Lee et al. 2011]. Conversely, those presenting with a CBS had various underlying pathologies including AD (n=9; 22.5%),

CBD (n=14; 35%), PSP (n=5; 12.5%), FTLD-U/TDP43 (n=5; 12.5%), mixed pathologies (n=5 comprised of two PSP+AD, one CBD+AD, and one FTLD-U/TDP43+AD; 12.5%), Pick‟s disease (n=1; 2.5%), and one with multiple system tauopathy without agyrophilia (n=1; 2.5%)

[Lee et al. 2011]. As can be seen from this review of prior clinicopathological studies of CBS

Figure 7. Microscopic agyrophilic grain disease pathology showing (A) branched astrocytes in the amygdale, and (B) agyrophilic grains and coiled bodies in the prosubiculum and CBD, the rate at which CBD pathology is predicted based on having a CBS is highly variable and in general is low. The variability is likely explained by the small samples sizes used in even the larger studies. Future studies are required that follow patients longitudinally to death and characterize them with multiple modalities including clinical examination, neuropsychological and neuroimaging with subsequent pathological analyses as only this type of study will improve our ability to predict the specific pathological diagnosis underlying the CBS in life.

Figure 8. Microscopic Alzheimer’s pathology showing (A) astrocytic plaques in frontal regions, and (B) neurofibrillary tangles in the CA1 region of the hippocampus

1.7 Genetics of CBS and CBD

Genetic analysis of complex syndromes, such as CBS, may be complicated by many factors such as incomplete penetrance, multiple disease susceptibility loci, gene-environment interactions and diagnostic uncertainties [Nothen et al. 1993]. The latter is particularly important given the significant pathological heterogeneity underlying the CBS described in the preceding section.

Two main strategies have been utilized for the genetic study of complex illnesses, such as CBS: a) linkage or candidate gene studies involving families or affected pairs of relatives, and b) association studies using candidate genes or genome wide approaches in unrelated cases and controls. Traditional family-linkage studies follow the segregation of marker alleles in pedigrees that are multiply affected with the disease phenotype of interest. A model is then proposed to explain the inheritance pattern of phenotypes and genotypes in the pedigree [Lander and Schork

1994]. Although this is the method of choice for simple Mendelian traits, linkage analysis of complex traits has limited power in identifying disease susceptibility loci, that is, estimating the large number of unknown parameters required to model complex traits is extremely difficult [Ott

1990]. A genetic association study design does not require specification of a genetic model and therefore overcomes many of the limitations inherent in the linkage-based familial approaches

[Crowe 1993;Kidd 1993]. Risch & Merikangas [Risch and Merikangas 1996] have suggested that association analyses have far greater power than linkage analyses to identify genes involved in complex genetic diseases. Both approaches have been applied to elucidate genetic factors contributing to the etiology of CBD.

From a genetic epidemiologic perspective, CBD is mainly a sporadic disorder with very few reported familial cases [Mahapatra et al. 2004]. In sporadic cases, genetic association studies have identified a particular haplotype that spans the MAPT gene among several other loci, as being associated with CBD and PSP pathology. Please refer to Caffrey and Wade-Martins

[Caffrey and Wade-Martins 2007] for a comprehensive review. The MAPT gene is localized to chromosome 17q21 and encodes for the Microtubule-Associated Protein Tau (MAPT)

[Andreadis et al. 1992]. Tau is highly expressed within both central and peripheral nervous system neurons where it is involved in the assembly and stabilization of microtubules, signal transduction and maintaining neuronal polarity [Shahani and Brandt 2002]. Hyperphosphorylated

Tau can aggregate in neurons producing pathological Tau inclusions called neurofibrillary tangles, which are present in several neurodegenerative diseases including AD, PSP, CBD, agyrophilic grains, FTD Parkinsonism-17 (FTDP-17), and Pick‟s disease [Caffrey and Wade-

Martins 2007].

MAPT is comprised of 16 exons and alternative splicing of exons 2, 3 and 10 yields six mRNA transcripts that are translated into unique protein isoforms [Goedert et al. 1988;Goedert et al.

1989]. Exons 9 through 12 of MAPT encode for imperfect repeat sequences that code for microtubule-binding domains and thus play an important role in the main function of the protein

[Caffrey and Wade-Martins 2007]. When exon 10 is spliced out, three repeat (3R) sequences are generated, whereas the presence of exon 10 results in the generation of four repeat (4R) sequences [Goedert et al. 1988;Goedert et al. 1989]. The major tangle isoform observed in CBD is comprised of 4R Tau [Caffrey and Wade-Martins 2007]. MAPT is located within the largest known block of linkage disequilibrium in the human genome that spans approximately 1.8

38 megabases (Mb) [Caffrey and Wade-Martins 2007]. Two major haplotypes, H1 and H2, have been defined based on tagging with eight single nucleotide polymorphisms (SNPs; inherited with

H1 haplotype) and a 238 base pair (bp) deletion (inherited with rarer H2 haplotype). Two early studies demonstrated that the H1 haplotype is over-represented in sporadic CBD cases compared to controls [Di Maria E. et al. 2000;Houlden et al. 2001].

Figure 9. Schematic representation of the MAPT genomic region and 3-repeat and 4-repeat Tau transcripts. Adapted from Caffrey and Wade-Martins [Caffrey and Wade-Martins 2007].

In the rare event that CBS and/or CBD pathology are observed to segregate in a family, other members are typically affected with FTD and/or PSP demonstrating overlap in these conditions

[Boeve et al. 2002;Brown et al. 1996;Brown et al. 1998;Bugiani et al. 1999;Gallien et al.

1998;Tuite et al. 2005;Casseron et al. 2005;Uchihara and Nakayama 2006]. Brown et al. [Brown et al. 1996] described two families in which a progressive dementia segregated in 15 affected individuals. Ten individuals were clinically studied and the main presenting features were that of

39 personality change or memory loss with invariable progression into a frontal dementia [Brown et al. 1996]. Additional features observed were aphasia, limb clumsiness, parkinsonism and gait imbalance [Brown et al. 1996]. Pathological examination in two individuals revealed swollen achromatic cortical neurons and corticobasal inclusion bodies in the basal ganglia [Brown et al.

1996]. Patients had features of frontotemporal dementia and/or CBS and the pathology most closely resembled CBD [Brown et al. 1996]. A similar family was described by the same group with one member clinically having CBS and a sibling having FTD, while the mother presented with an early onset dementia with features of a movement disorder [Brown et al. 1998]. The pathological diagnosis of the individual with CBS was dementia lacking distinctive histopathological features confirming heterogeneity even in familial cases of the syndrome

[Brown et al. 1998]. Pathological findings of non-distinctive histopathology has also been found in another study of a kindred that included a patient with CBS and that also segregated FTD in other individuals [Boeve et al. 2002]. An Italian group identified a family with two afflicted members, the father presenting as FTD and the son presenting as CBS, with the etiological cause being a Proline301Serine (P301S) mutation in exon 10 of MAPT that lead to extensive filamentous hyperphosphorylated Tau pathology [Bugiani et al. 1999]. A Japanese group identified three siblings, all of whom presented with parkinsonism and frontal dementia, with typical CBD pathology [Uchihara and Nakayama 2006]. Tuite et al. [Tuite et al. 2005] identified a consanguinous family with members having clinical diagnoses of PSP and CBS. In two members who presented as CBS, one had confirmed CBD pathology while the other demonstrated PSP pathology demonstrating the overlap at both the clinical and pathological level

[Tuite et al. 2005]. Interestingly, no MAPT mutations were identified and only the H1/H1 haplotype was found in the four affected individuals studied [Tuite et al. 2005]. CBS has also

40 been associated with a leucine-rich repeat kinase 2 mutation [Chen-Plotkin et al. 2008]. Given the pathologic and genetic heterogeneity observed in CBS, future genetic studies of families with this syndrome and association studies of unrelated individuals are required to identify other causative genes and/ or genetic risk factors that predispose to this syndrome.

Figure 10. (A) H1 and H2 linkage disequilibrium blocks showing a 900 kb region of inversion, and (B) sub- structure of the MAPT gene and associated H1 and H2 haplotypes. Adapted from Caffrey and Wade-Martins [Caffrey and Wade-Martins 2007].

1.8 Synopsis and Overall Research Objective

Despite the rarity of CBS compared to other neurodegenerative disorders such as AD and PD, it represents an important neurodegenerative syndrome to study given the substantial heterogeneity that is observed in its initial presentation and evolution over time. Understanding this heterogeneity in CBS may facilitate our understanding of heterogeneity in the more common neurodegenerations. From a clinical perspective, it is unclear why some CBS patients present with an early motor syndrome while others present initially with mainly symptoms of dementia.

Comparing these different presentations of CBS in terms of clinical, neuropsychological and neuroimaging features may help to shed light on the brain regions involved in determining the type of symptom onset in CBS and this may help to determine which patients presenting with an early dementia would be at risk of evolving into a CBS. Additionally, there have been few CBS samples that have been extensively characterized to allow for brain-behaviour correlations using structural and functional neuroimaging, and this remains an important line of investigation to understand the localization of some of the observed phenomenology in the brain. Novel genetic studies are required in order to elucidate additional genes that can cause or increase risk for the

CBS. Finally, more work needs to be done in understanding how clinical and neuroimaging features map on to the various pathologies that can underlie the CBS, as this will provide insight into clinicopathological correlations, which may help in the prediction of underlying in vivo pathological state. Therefore, the overall objective of this thesis is to characterize a prospective sample of CBS patients in terms of the heterogeneity observed across clinical, neuropsychological, and neuroimaging features and, in a subset of the sample, to describe 42 genetic and pathological features and how these relate to the clinical phenotype and neuroimaging findings.

1.8.1 Specific Objectives

The specific objectives of this thesis and related hypotheses are as follows:

A) Objective 1: To characterize the genetic and pathological heterogeneity observed in a

family segregating corticobasal syndrome.

Hypothesis 1: Affected patients will harbor a mutation in one of the genes known to

cause diseases occurring along the spectrum of frontotemporal dementia, including CBS,

and will have associated pathological features that are typical of the identified genetic

mutation.

B) Objective 2: To characterize members of a family that segregate a novel mutation in the

progranulin gene (PGRN) associated with FTD spectrum disorders, including CBS, and

to contrast the heterogeneity observed in their clinical presentation, neuropsychological

testing, and neuroimaging findings.

Hypothesis 2: There will be significant heterogeneity in clinical, neuropsychological, and

neuroimaging features among patients with the same PGRN mutation and this will be

dependent on the hemisphere and lobar region most prominently affected in the early

stages of the disease.

C) Objective 3: To identify brain SPECT perfusion and neuropsychological correlates of

severity of ideomotor apraxia in CBS.

Hypothesis 3A: Compared to controls, CBS patients will demonstrate reduced perfusion

on SPECT in an asymmetrical fashion in frontoparietotemporal cortical and subcortical

regions.

Hypothesis 3B: Hypoperfusion within the left frontoparietal network will correlate with

severity of ideomotor apraxia in CBS.

Hypothesis 3C: Patients with more severe apraxia will demonstrate more impairment on

language-based measures.

D) Objective 4: To describe the initial neuropsychological and neuropsychiatric, MRI, and

pathological features of a prospective sample of CBS patients.

Hypothesis 4A: Compared to controls, CBS patients will demonstrate reduced

performance globally on neuropsychological testing with worse performance on

measures assessing executive, visuospatial, language and praxis functions.

Hypothesis 4B: Compared to controls, asymmetric atrophy on MRI contralateral to the

most affected side of the body will be observed.

Hypothesis 4C: In a subset of the CBS sample that came to autopsy, underlying

neuropathological diagnoses will be heterogeneous.

Hypothesis 4D: Atrophy and white matter hyperintensities on MRI in vivo will be

associated with the underlying neuropathology.

E) Objective 5: To compare the clinical, neuropsychological, MRI, and SPECT features of

CBS presenting with early dementia versus those presenting with early motor features.

Hypothesis 5A: The CBS group with early dementia will be more likely to have their

right side of the body affected by motor signs, have more profound language deficits, and

have hemispheric atrophy and reduced perfusion in left frontotemporal regions.

Hypothesis 5B: The CBS group with early motor features will be more likely to have

their left side of the body affected by motor signs, and have hemispheric atrophy and

hypoperfusion on SPECT that is more pronounced on the right.

1.9 Description of Chapters

1.9.1 Chapter 2: Novel splicing mutation in the progranulin gene causing familial corticobasal syndrome

This study was the first to report that mutations in the PGRN gene, discovered in 2006 as a major cause of frontotemporal dementia, can also cause familial corticobasal syndrome. It provides a detailed account of two family members afflicted with corticobasal syndrome and characterizes one of the family members from the clinical, neuropsychological, neuropsychiatric and neuroimaging perspective. The other sibling is characterized from the pathological standpoint as having underlying FTLD-U pathology, which, shortly after this publication was accepted, was found to be a marker of TDP43 pathology [Neumann et al. 2006]. This study extends the literature on genetic and phenotypic heterogeneity associated with FTD and set the stage for several follow-up papers confirming our initial findings.

1.9.2 Chapter 3: Intra-Familial Clinical Heterogeneity due to FTLD-U with TDP43

Proteinopathy Caused by a Novel Deletion in Progranulin Gene (PGRN)

This paper identifies a novel mutation in the PGRN gene that caused neurodegenerative presentations in a kindred originally from Poland and characterizes two affected brothers from the clinical, neuropsychological, and neuroimaging perspective comparing important differences in presentation and how these correlate with heterogeneous neuroimaging findings between them. One of the brothers, who initially presented with symptoms of progressive non-fluent aphasia (PNFA) and then evolved into CBS, is studied from the pathological perspective demonstrating the expected FTLD-U/ TDP43 pathology. This study extends on the literature demonstrating allelic and phenotypic heterogeneity in FTD and proposes molecular mechanisms, which likely underly some of this heterogeneity.

1.9.3 Chapter 4: Ideomotor Apraxia in Corticobasal Syndrome: Brain Perfusion and

Neuropsychological Correlates

This paper replicates findings from previous studies showing that perfusion reductions on

SPECT occur in frontoparietotemporal regions in CBS compared to controls and is the first to identify that severity of ideomotor apraxia in CBS correlates strongly with reduced perfusion in the left inferior parietal lobule in patients afflicted with this syndrome. It is the largest SPECT study of CBS that attempts to understand the neuroanatomical correlates of ideomotor apraxia and also identifies that several other posterior cognitive functions are more impaired in the CBS group with significant apraxia compared to those without this feature. The study is one of the first to provide a comprehensive discussion of limitations in the field of apraxia research and

46 identifies that many of the limitations originate from variable definitions that are currently applied to the different types of apraxia.

1.9.4 Chapter 5: Clinical, neuropsychological, MRI and SPECT characterization of a prospective sample of patients with corticobasal syndrome

This paper provides a comprehensive and multi-modal assessment of a prospective sample of

CBS patients using clinical and neuropsychological assessments, MRI and brain SPECT neuroimaging. It then compares a subgroup of CBS patients presenting with early dementia to one presenting with early motor features identifying a tendency for the early dementia group to have symptoms involving the right side of the body and to have more severe language disturbances whereas the early motor group has symptoms prominently involving the left side of their body. A subset of the patients came to autopsy and heterogeneity in pathological diagnoses was observed. The burden and location of the pathology mostly correlated with neuroimaging features irrespective of the specific underlying pathological diagnoses.

2.0 Novel splicing mutation in the progranulin gene causing

familial corticobasal syndrome

Mario Masellis,* Parastoo Momeni,* Wendy Meschino, Reid Heffner Jr., Christine Sato, Yan

Liang, Peter St George-Hyslop, John Hardy, Juan Bilbao, Sandra Black, and Ekaterina Rogaeva

As published in: Brain (2006); 129: 3115-3123

* These authors contributed equally to the work as co-first authors

2.1 SUMMARY

Corticobasal Syndrome (CBS) is a rare cognitive and movement disorder characterized by asymmetric rigidity, apraxia, alien-limb phenomenon, cortical sensory loss, myoclonus, focal dystonia, and dementia. It occurs along the clinical spectrum of Frontotemporal Lobar

Degeneration (FTLD), which has recently been shown to segregate with truncating mutations in progranulin (PGRN), a multifunctional growth factor thought to promote neuronal survival. This study identifies a novel splice donor site mutation in the PGRN gene (IVS7+1G>A) that segregates with CBS in a Canadian family of Chinese origin. We confirmed the absence of the mutant PGRN allele in the RT-PCR product which supports the model of haploinsufficiency for

PGRN-linked disease. This report of mutation in the PGRN gene in CBS extends the evidence for genetic and phenotypic heterogeneity in FTLD spectrum disorders.

Keywords: Corticobasal Syndrome; Frontotemporal Lobar Degeneration; progranulin; gene; mutation

Abbreviations: CBS = Corticobasal Syndrome; CBD = Corticobasal Degeneration; FTD =

Frontotemporal Dementia; MAPT = microtubule-associated protein tau; FTDP-17 = FTD with parkinsonism linked to chromosome 17; PSP = Progressive Supranuclear Palsy; MMSE = Mini-

Mental Status Examination; PGRN = progranulin; MND = Motor Neuron Disease; CHMP2B =

Chromatin-modifying protein 2B; MRI = Magnetic Resonance Imaging; SPECT = Single Photon

Emission Computed Tomography; FTLD = Frontotemporal Lobar Degeneration; LFB = luxol fast blue

2.2 INTRODUCTION

In 1967, Rebeiz and colleagues [Rebeiz et al. 1967] described three cases of a progressive, perceptuo-motor disorder characterized by an asymmetric akinetic-rigid syndrome and apraxia.

They termed the disorder “corticodentatonigral degeneration with neuronal achromasia” based on identified pathological features. Since then, a variety of terms have been applied to this enigmatic disorder of interest to cognitive and movement disorder neurologists worldwide including corticonigral degeneration with neuronal achromasia, cortical degeneration with swollen chromatolytic neurons, cortical basal ganglionic, corticobasal ganglionic, and the most common designation, corticobasal degeneration (CBD) [Mahapatra et al. 2004]. This terminology has caused considerable nosological confusion over the years presumably because some terms refer to the underlying pathological changes, while others refer to the neural substrates causing the recognized clinical syndrome.

To add to this nosological uncertainty, extensive research has demonstrated significant clinical and pathological heterogeneity in CBD [Boeve et al. 2003;Lang 2003]. Specifically, cases presenting with the “classical” clinical syndrome of CBD often have alternative pathologies (i.e., not CBD) underlying the clinical manifestations such as Progressive Supranuclear Palsy (PSP),

Frontotemporal Dementia (FTD), Alzheimer‟s Disease (AD), Dementia with Lewy Bodies

(DLB), and Creutzfeldt-Jacob Disease (CJD). Conversely, pathologically-confirmed cases of

CBD [Dickson et al. 2002] may present with a variety of clinical phenotypes in addition to

“classical” CBD including Primary Progressive Aphasia (PPA) and Frontotemporal Dementia

(FTD). As a result, it has been suggested that the term Corticobasal Syndrome (CBS) be applied

50 to clinically-diagnosed cases presenting with the “classical” features of asymmetric rigidity, apraxia, alien-limb phenomenon, cortical sensory loss, myoclonus, and focal dystonia [Boeve et al. 2003;Kertesz et al. 2000a;Lang 2003;Litvan et al. 2003]. Herein, we use the term

Corticobasal Syndrome (CBS) to refer to clinically diagnosed cases without proof of typical

CBD pathology conforming to the clinical diagnostic criteria [Boeve et al. 2003]. Included in this syndromic definition are patients presenting with early dementia, for which there is evidence suggesting this to be the most common initial presentation [Bergeron et al. 1998;Grimes et al.

1999b;Mathuranath et al. 2000]. The cognitive symptoms and underlying pathologies of CBS have many overlapping features with those of FTD prompting current nosological classification to include CBS as part of the spectrum of FTD [Josephs et al. 2006;Kertesz et al. 2000b;Neary et al. 1998]. Similar to CBS, several terms have been applied to describe this heterogeneous disorder including FTD/Pick Complex [Kertesz 2003], Frontotemporal Lobar Degeneration

[Neary et al. 1998], Pick‟s Disease [Pick 1892], and FTD [The Lund and Manchester Groups

1994]. We have adopted the term Frontotemporal Lobar Degeneration (FTLD) in this paper.

FTLD encompasses a wide spectrum of clinical entities ranging from FTD, Primary Progressive

Aphasia, Semantic Dementia, CBS, PSP, FTD-Motor Neuron Disease (FTD-MND), and FTD with Parkinsonism linked to chromosome 17 (FTDP-17) [Kertesz 2003;Kertesz 2005]. It represents a group of primary degenerative dementias with predominant frontal and/or temporal lobe symptoms (e.g. decline in social and personal behavior, apraxia, stereotyped behavior, hyperorality and aphasia) [Kertesz 2005] and consensus diagnostic and neuropathological criteria have been proposed [McKhann et al. 2001;Neary et al. 1998]. The neuropathological characteristics of FTLD include variable frontal, temporal, and basal ganglia atrophy with

51 neuronal loss and gliosis (with tau or ubiquitinated inclusions). The deposition and abnormal processing of tau encoded by the gene named microtubule-associated protein tau (MAPT) play an important role in the development of several forms of FTLD, including CBS [Goedert et al.

2000;Hutton 2001;McKhann et al. 2001]. However, up to 60% of FTLD cases lack tau-positive neuronal inclusions, primarily displaying a microvacuolization of the superficial neuropil in the cortex (often with ubiquitin-positive inclusions in cortical neurons) [Ince and Morris 2006;Ince and Morris 2006;Kertesz et al. 2000a].

FTLD is a genetically complex disorder with at least three known causal genes. The aberrant splicing mutation in Chromatin-modifying protein 2B (CHMP2B) is responsible for autosomal dominant FTLD in a large Danish family [Skibinski et al. 2005]. However, the CHMP2B is not a common cause of FTLD since several large series of FTLD patients failed to detect any

CHMP2B mutations [Cannon et al. 2006;Momeni et al. 2006]. Many of the autosomal dominant

FTDP-17 families are explained by mutations in the MAPT gene [Hutton et al. 1998;Poorkaj et al. 1998;Spillantini et al. 1998]. However, in several FTLD families linked to chromosome

17q21, MAPT mutations were excluded. Recently the disease in many of these families was explained by truncating mutations in the progranulin gene (PGRN) which was mapped ~1.7 Mb centromeric of the MAPT locus [Baker et al. 2006;Cruts et al. 2006]. The PGRN gene encodes a secreted multifunctional growth factor involved in development, wound repair and inflammation.

Patients with PGRN mutations do not have tau-pathology. Instead there are ubiquitin- immunoreactive neuronal cytoplasmic and intranuclear inclusions, the protein identity of which remains unknown [Baker et al. 2006;Cruts et al. 2006;Mackenzie et al. 2006].

Neurodegeneration in mutation carriers is caused by PGRN haploinsufficiency due to nonsense- mediated decay since transcript analysis demonstrated the absence of the mutant allele.

Herein, we describe the clinical, neuropathological and genetic findings of a CBS-like disease which is segregating a novel PGRN mutation in a Canadian family of Chinese origin. This finding extends knowledge on the clinical, pathologic and genetic heterogeneity of CBS and

FTLD.

2.3 METHODS

2.3.1 Subjects

The proband (Case 4150) was recruited through the Linda C. Campbell Cognitive Neurology

Research Unit at Sunnybrook Health Sciences Centre in Toronto as part of the Sunnybrook

Dementia Study. This is a prospective, longitudinal study of dementia and aging with well over

800 subjects enrolled to date approved by the local Research Ethics Boards. Patients or their substitute decision makers provide written, informed consent to participate in accordance with the Declaration of Helsinki. The proband underwent a detailed clinical assessment including: history and physical examination, and standardized behavioural neurology assessment. Routine biochemical screening was done to exclude any other causes for their presentation. The patient was seen every six months for routine clinical follow-up and had yearly prospective longitudinal assessments which included: detailed neuropsychological battery (measures of general intelligence and cognition, language, praxis, visuospatial ability, attention and working memory, and executive functions), measures of neuropsychiatric symptoms and of functional status.

Structural and functional neuroimaging of the brain with Magnetic Resonance Imaging (MRI) and Single Photon Emission Computed Tomography (SPECT), respectively, were performed.

The sister of the proband (Case #4993) was identified through clinical history from the proband.

Information pertaining to this case is limited to that ascertained through a telephone interview with her caregiver and through an autopsy report as this patient was residing out of country. The normal control group consisted of 200 unrelated subjects of North American origin (mean age at time of examination of 72.7  8.4 years).

2.3.2 Neuropathology

Neuropathological examination was carried out by two of the authors (R.H.; J.B.). Paraffin- embedded sections were stained with Hematoxylin and Eosin, luxol fast blue (LFB),

Bielschowski and Gallyas. Immunostains using commercial antibodies for tau (Dako, #A0024) and ubiquitin (Vector Labs, #ZPU576) were performed.

2.3.3 Genetic Analysis

Genomic DNA and RNA were extracted from whole blood using Qiagen kits. Two affected members of the family (Case 4150 and Case 4993) were tested for mutations in exons 1 and 9-13 of the MAPT gene by direct sequencing as previously described [Kertesz et al. 2000a]. The entire open reading frame with the exon-intron boundaries of the CHMP2B and PGRN genes were sequenced in both affected individuals as previously described [Baker et al. 2006;Skibinski et al.

2005]. RT-PCR primers were designed for PGRN exon 3 (5‟- GCCACTCCTGCATCTTTACC-

3‟) and exon 8 (5‟-TTCTCCTTGGAGAGGCACTT-3‟). The RT-PCR conditions were 94C for

5 min, followed by 40 cycles of 94C for 30 sec, 58C for 30 sec, 72C for 30 sec, and 7 min at

72C. Mutations were detected by direct inspection of the fluorescent chromatographs and by analysis using the SeqScape software version 1.0 (Applied Biosystems, Foster City, CA).

2.4 RESULTS

2.4.1 Clinical features and autopsy results

This family of Chinese origin presented with inheritance of a progressive neurodegenerative disorder characterized by dementia and motor decline, including rigidity, dystonia, apraxia, cortical sensory loss, visuospatial dysfunction and behavioural changes (Figure 1A). Family records indicate that two out of 12 siblings have been affected with Corticobasal Syndrome. A third family member has developed early parkinsonism. Two patients were available for the genetic and clinical study.

Figure 1.

(A) The pedigree structure of the Canadian family showing the inheritance of the disease (with age-at-onset). Affected individuals are shown as filled symbols and the arrow points to the proband. The gender of the individuals has been masked to protect family confidentiality; (B) Genomic DNA (gDNA) and RT-PCR (cDNA) sequence fluorescent chromatograms around the PGRN mutation (IVS7+1G>A) observed in the patients and the sequence around common synonymous variation rs25646; (C) An agarose gel photograph of the PGRN product from RT-PCR, using RNA obtained from white blood cells of the affected family member (#4150) and normal control (the 586bp band corresponds to the PGRN fragment containing exons 3-8 confirmed by sequencing analysis).

2.4.1.1 Case #4150 (Proband)

This 71 year old right-handed woman with a previous history of hyperthyroidism treated with radioablation and requiring thyroid replacement presented at age 62 with the insidious onset of behavioural changes including increased irritability, depression, social withdrawal, and suspiciousness. Subsequently, she began to experience difficulties with short-term memory, planning, attention, word-finding difficulties, and getting lost in familiar environments.

Abnormalities on her initial examination (age 65) were a left visual field defect which was thought to be, in part, secondary to profound left hemi-neglect, left cortical sensory loss

(specifically, sensory extinction and agraphesthesia), left-hand ideomotor apraxia, and a dressing apraxia. These exam features are consistent with right parieto-occipital dysfunction. She scored

20/30 on the Mini-Mental Status Exam (MMSE) putting her in the moderate range of dementia severity. Cognitive testing confirmed severe visual perceptual dysfunction and also revealed short-term memory deficits, impaired executive functions, anomic aphasia and apraxia. The results of the neuropsychological battery and standardized scores are summarized in Table 1. An

MRI of the brain revealed right greater than left hemispheric cortical atrophy and ventricular dilatation, slightly more prominent in the posterior regions; there were also some periventricular white matter changes (Figure 2A). A brain SPECT scan demonstrated a large right

Demographics, Neuropsychological Battery and Raw Scores for Standardized Category Functional Measures (Test name /maximum raw Case #4150 Score score) Age of Onset 62 - - Age at this testing 65 - - Duration of disease at testing 3 - - Years of education 12 - - General cognition Folstein’s Mini-Mental Status Examination /30 20 ≥ 28 (NCO) Impaired Mattis Dementia Rating Scale /144 92 2 (SS) Impaired Memory California Verbal Learning Test – Long Delay Free 4 -2 (ZS) Impaired Recall /16 Delayed Visual Reproduction /41 0 1st percentile Impaired Language Western Aphasia Battery – total /100 83 -2 (ZS) Impaired Western Aphasia Battery – comprehension /10 8 -2 (ZS) Impaired Boston Naming /30 19 2 (SS) Impaired Semantic Fluency /20 6 < 10th percentile Borderline-Impaired Praxis Western Aphasia Battery – praxis /60 48 -2 (ZS) Impaired Attention & working memory Digit span – forward /12 6 30th percentile Normal Digit span – backward /12 2 5th percentile Borderline Visuospatial abilities Rey Osterieth Complex Figure – copy /36 0 < 1st percentile Impaired Benton Line Orientation /30 N/A ≤ 4 (SS) Impaired Executive functions Phonemic fluency (F-, A-, S-words) 16 3 (SS) Impaired Wisconsin Card Sort Test – categories /6 0 > 1 (NCO) Impaired Wisconsin Card Sort Test – perseverative errors 20 0 (NCO) Impaired Activities of daily living Disability Assessment for Dementia (%) 53 100 (NCO) Impaired Neuropsychiatric symptoms Neuropsychiatric Inventory – total /144 24 0 (NCO) Abnormal Neuropsychiatric Inventory – apathy /12 8 0 (NCO) Abnormal Neuropsychiatric Inventory – depression /12 8 0 (NCO) Abnormal Neuropsychiatric Inventory – disinhibition /12 0 0 (NCO) Normal Cornell Depression Scale (%) 53 < 25 (NCO) Depressed Table 1. Scores on neuropsychological and functional measures for case #4150 compared to standardized scores calculated based on normal population matched for age and years of education. Abbreviations: NCO = Normal cut-off; SS = Scaled score (Mean = 10; 1 standard deviation = 3); ZS = Z-score; N/A = Not assessable parieto-occipital perfusion deficit extending into the temporal and frontal regions with a milder decrease in perfusion in the left parietal lobe (Figure 2B). The neuropsychological data was

58 collected within a one month time period of the MRI and SPECT images. The provisional diagnosis was thought to be posterior cortical atrophy, a possible variant of Alzheimer‟s disease.

She was initiated on a cholinesterase inhibitor with no major change in symptoms apart from some improvement in attention.

Figure 2. Corresponding (A) T1-weighted Magnetic Resonance Imaging (MRI) and (B) Technetium 99m-ethyl cysteinate dimer (99mTc-ECD) Single Photon Emission Computed Tomography (SPECT) scans of the brain of Case #4150. Areas of relative atrophy on MRI and decreased cerebral perfusion on SPECT in the right inferior frontal (IF), inferior parietal (IP), superior frontal (SF), superior parietal (SP), and occipital (O) regions are demonstrated. There is a clear asymmetry in cortical atrophy and regional perfusion with the right being more affected than the left. For the SPECT images, orange-yellow colours represent areas of high perfusion while blue-purple colours represent areas of low perfusion.

A year after her initial assessment (age 66), the patient‟s cognitive performance continued to decline (MMSE = 12) and she required assistance in all activities of daily living. She also was

59 developing an asymmetric akinetic-rigid syndrome including prominent rigidity of the left upper and lower extremities, bradykinesia and a stooped posture with a shuffling gait. The provisional diagnosis was changed to CBS based on the emergence of an asymmetric akinetic-rigid syndrome and severe left-sided ideomotor apraxia. She met clinical criteria for CBS [Boeve et al.

2003]. Over the next three months, the patient became verbally and physically aggressive towards her day-time caregiver. Cognitively, her dementia had progressed into the severe range and she was completely dependent for self care. She had a positive glabellar tap and bilateral grasp reflexes consistent with frontal release phenomena. At this point, she was observed to be constantly biting her finger nails, likely representing repetitive, stereotyped behavior. Clinically, her akinetic-rigid syndrome had progressed and she now developed dystonic posturing of her left hand, and worsening left-sided apraxia, the combination of which produced a useless left arm.

Approximately three years after her initial assessment (age 68), she lost the ability to ambulate and developed corticospinal tract findings on the left side of her body (i.e., left hyperreflexia and extensor plantar response). Her verbal output declined and she would often repeat phrases such as “you‟re killing me”. She continued to decline and four years after the initial assessment (age

69), her speech output diminished to the point where she was only able to grunt to indicate her needs, with relative preservation of verbal comprehension. Eventually, she became mute and lost the ability to comprehend and interact with others. Recently, she developed dysphagia to liquids and is able to eat only pureed foods. Currently (age 71), she is bed-bound with end-stage CBS about nine years into the course of her illness.

2.4.1.2 Case #4993 (sister of proband)

This deceased 61 year old woman had a history of dementia and motor decline since age 57 consisting of axial and extremity rigidity and aphasia. She had significant contractures and flexion posturing of her upper extremities and right lower extremity. She required complete personal care and gastrostomy tube feeds for nutrition towards the end of her disease course. Her clinical diagnosis by a neurologist was CBS. She passed away at age 61 from medical complications related to her neurodegenerative disorder. Disease duration in this patient was about four years.

2.4.1.3 Neuropathology (Case #4993).

Gross: The whole brain weighed 940 grams unfixed. Examination of the right half of the fixed brain demonstrated mild to moderate sulcal widening in the frontotemporal regions. Coronal sections showed a well-defined and regular cortical ribbon without focal defects. Significant widening of the circular sulcus and Sylvian fissure was noted. The caudate nucleus and putamen were atrophic. The hippocampus was normal in size. The substantia nigra was normally pigmented. There were no gross abnormalities of the cerebellum, pons, medulla, or cervical spinal cord.

Microscopic: Severe pancortical micro-vacuolation associated with neuronal loss and gliosis was seen in the frontal cortex. Similar changes were seen in the insular and temporal cortices and in the basal ganglia. The vacuoles varied in size and were more numerous in the superficial layers of the cortex. Vacuoles were not encountered in the thalamus, brainstem, cerebellum and spinal cord. The vacuoles were located within neuronal cytoplasm and the neuropil. Patchy myelin

61 pallor was demonstrated in the white matter underlying the atrophic cortical areas. This finding was best seen in LFB stains. The hippocampus was well-preserved. There was some neuronal loss in the substantia nigra with an absence of Lewy bodies in the brainstem or cerebral cortex.

Bunina bodies were not seen in the motor nuclei of the cranial nerves. Bielschowski stains demonstrated no neocortical senile plaques but rare, probably age-related, plaques were identified in the hippocampus. No astrocytic plaques were observed in Gallyas stains. There were no axonal spheroids. Immunostains for tau protein were performed and showed no reactivity in neurons or other cells. Immunostains for ubiquitin demonstrated ubiquitin-reactive neuronal cytoplasmic inclusions. Ubiquitin-reactive neuronal intranuclear inclusions were not seen. Scattered neurites in the frontotemporal cortex were also ubiquitin positive. These findings are compatible with the diagnosis of FTD with ubiquitin-only positive inclusions also referred to as FTD-MND-type inclusion or FTD-U pathology [Lipton et al. 2004;Mackenzie and Feldman

2005;Mann et al. 2000;Taniguchi et al. 2004].

The third affected family member (brother of proband), after retiring at age 65, experienced

“dizzy” spells and did not feel well. He saw a number of doctors and he was told that he had early Parkinsonism. Although he was never diagnosed with a dementing illness, he has been unable to drive or prepare meals for himself. Information pertaining to this brother was limited to history from a family member. There was no history of dementia or Parkinsonism in either parent. The father died in his sixties from tuberculosis. The mother died at age 65 from

“pulmonary edema”. The other siblings are unaffected.

2.4.2 Genetic Analysis

Due to the clinical course and strong family history of disease, we performed mutation analysis of all three known FTLD genes (MAPT, CHMP2B and PGRN) for patients #4150 and #4993. We did not observe any sequence variations in the MAPT and CHMP2B genes. However, in the

PGRN gene we identified a novel heterozygous single nucleotide G-to-A mutation in the invariant “GT” splice donor site 3‟of exon 7 (genomic position 5680; Accession Number

AC003043) (Figure 1B). The exon numbering was according to the National Center for

Biotechnology Information (http://www.ncbi.nlm.nih.gov/) and our exon 7 corresponds to exon 6 in the published report [Baker et al. 2006]. The IVS7+1G>A mutation segregates with the disease in the two affected family members (#4150 and #4993) tested and was not found in 200 normal controls.

The mutation is predicted to have dramatic consequences on the maturation process of PGRN mRNA leading to the removal of exon 7 which would create a frame shift and truncate the protein to half its normal length (amino acid position 236). Likely such a transcript will be destroyed by nonsense mediated decay. In agreement with this, the result of the RT-PCR, using

RNA isolated from the white blood cells of patient #4150, revealed only the wildtype product on agarose gel (Figure 1C). The specificity of this RT-PCR product, containing exons 3-8 of the

PGRN gene, was confirmed by direct sequencing analysis. Importantly, this patient, who is heterozygous (T>C) for a common synonymous polymorphism in exon 5 (D128D; rs25646) using genomic DNA, showed only the “C” allele in the RT-PCR product (Figure 2B). Hence, the

RT-PCR result demonstrates the absence of the mutant PGRN transcript.

2.5 DISCUSSION

In the family reported here, the novel splice donor site mutation in the PGRN gene

(IVS7+1G>A) affects the sequence that is important in the recognition of the intron/exon boundary and removal of the intron [Berget 1995]. There are no doubts about the pathological nature of this mutation. It segregates with the disease in two affected family members and was absent in 200 unrelated normal controls. The predicted consequence of the splicing mutation is either the expression of the truncated protein or the haploinsufficiency of PGRN due to nonsense mediated decay. According to the published reports the second possibility is more likely [Baker et al. 2006;Cruts et al. 2006]. Indeed, our attempt to evaluate the pathological consequences of the IVS7+1G>A mutation by RT-PCR using RNA from the blood cells of patient #4150 did not identify aberrant PGRN transcripts (Figure 1C). Instead we confirmed the absence of the mutant

PGRN allele in the RT-PCR product (Figure 1B). Hence, the progression from normal function to the disease state would result from the reduction of the PGRN level, further supporting the model of haploinsufficiency for PGRN-linked FTLD. Previously, a different splicing mutation

(named IVS8+1G>A) was reported in one family; however, a source of RNA was not available to confirm the haploinsufficiency mechanism [Baker et al. 2006].

The cases described in this family met clinical criteria for CBS [Boeve et al. 2003]. Pathology in one affected individual demonstrated ubiquitin-positive, tau-negative cytoplasmic inclusions consistent with the pathology reported in the original FTLD families in which PGRN mutations co-segregate with disease [Baker et al. 2006;Cruts et al. 2006]. To our knowledge, this is the first report of mutation in PGRN causing familial CBS with underlying FTD-MND-type inclusion pathology. This type of pathology has been demonstrated previously in sporadic cases of CBS

[Grimes et al. 1999a;Kertesz et al. 2005]. One could surmise that these previously reported

“sporadic” cases may come from families with PGRN mutations that were non-penetrant.

Previous familial studies have demonstrated that CBS coexists with PSP, and/or FTLD [Boeve et al. 2002;Brown et al. 1996;Brown et al. 1998;Bugiani et al. 1999;Gallien et al. 1998;Tuite et al.

2005;Uchihara and Nakayama 2006]. Only two of these studies had more than one affected individual with CBS making this a relatively uncommon presentation in FTLD families [Tuite et al. 2005;Uchihara and Nakayama 2006]. Our study extends this literature in that two of the affected family members have CBS, while one has early parkinsonism which may be evolving into a dementing condition based on history. Perhaps, the novel splice donor site mutation in

PGRN identified in this family predicts the phenotypic expression of CBS as opposed to FTLD or PSP. However, this would be unlikely given the current haploinsufficiency model proposed for PGRN mutation. Another possibility may be that the FTLD phenotype may be differentially expressed in Asians such that CBS is more likely to occur. Reasons for this might include epigenetic factors, modifier genes, and/or environmental influences that “tip the balance” in favour of one particular manifestation of FTLD over another.

The proband in our study presented initially with behavioural symptoms consisting of increased irritability, depression, social withdrawal, and suspiciousness. Prominent visuospatial dysfunction was present early on in the clinical course. Subsequently, she had difficulties with short-term memory, executive functions, and expressive language. MRI and SPECT imaging of the brain (Figures 2A and 2B) demonstrated cortical atrophy and reduced perfusion, respectively,

65 in the right parieto-occipital greater than right frontotemporal regions which was clearly asymmetrical when compared to the left hemisphere. This suggested an initial diagnosis of posterior cortical atrophy although clinically there were also deficits of anterior cerebral dysfunction. Once the extrapyramidal features evolved, the diagnosis of CBS became clear.

We have previously reported a case of a patient with sporadic CBS who presented initially with prominent visuospatial dysfunction and a hemi-neglect syndrome similar to the proband in the current study [Kleiner-Fisman et al. 2003]. Interestingly, final pathological diagnosis in this patient confirmed ubiquitin positive, tau negative inclusions consistent with FTD-MND-type inclusion pathology (unpublished data) similar to the pathology observed in the current study.

Visuospatial dysfunction in CBS has also been observed rarely [Mendez 2000;Okuda et al.

2000a] with one study demonstrating underlying typical CBD pathology [Tang-Wai et al. 2003].

Therefore, CBS presenting with prominent visuospatial dysfunction does not necessarily predict the specific underlying pathological diagnosis.

Although both cases described in this family were diagnosed with CBS, there were significant differences in their clinical course. The proband presented at age 62 with behavioural symptoms and posterior cerebral dysfunction and evolved over a few years into CBS and is still living nine years after disease onset, although nearing end-stage disease. The sister of the proband presented at a younger age (57 years old) and had early and prominent motor features which eventually lead to death at age 61, four years after symptom onset. Unknown environmental or genetic

66 factors or stochastic events must contribute to this variability in age of onset and disease severity within families and will require further investigation.

Clinical diagnosis along the FTLD spectrum is challenging and frequently longitudinal follow-up of patients is required to ascertain the most likely provisional diagnoses. Take, for example, the prospective, clinic-based cohort of FTLD patients of Kertesz et al. [Kertesz et al. 2005] that was followed longitudinally to autopsy. In this cohort, the authors describe patients presenting with initial syndromes ranging from behavioural variants of FTLD, CBS, PSP, to primary progressive aphasia. The majority of these patients then went on to develop second and/or third syndromes with significant clinical overlap along the FTLD spectrum. Added to this complexity is the fact that there were a variety of pathologies underlying each of the clinical phenotypes ranging from tau positive to tau negative types. For the most part, the clinical syndrome of FTLD observed is dependent more on “the distribution of the underlying pathological state rather than on its nature” [Lang 2003]. It is hoped that as we learn more about the underlying molecular pathogenic mechanisms of FTLD spectrum disorders, diagnostic accuracy in life will improve and this will also lead to potential therapies to prevent or cure these debilitating disorders.

2.6 ACKNOWLEDGEMENTS

This work was supported by grants from the Japan-Canada and Canadian Institutes of Health

Research Joint Health Research Program, Parkinson Society of Canada (ER), Canadian Institutes of Health Research (PSGH; SB – MT13129), Howard Hughes Medical Institute, Canada

Foundation for Innovation (PH). MM is supported by a Medical Scientist Training Fellowship

67 from the McLaughlin Centre for Molecular Medicine, University of Toronto. JH was supported by the NIA/NIH Intramural Program. The authors would like to thank Edward Huey, MD for constructive criticisms; Mr. Shahryar Rafi-Tari for his assistance in preparing the neuroimaging figure; Ms. Isabel Lam for her assistance in preparing the table of neuropsychological data.

2.7 ADDENDUM

Since this original paper was published, we have subsequently confirmed that the proband described in this study had a pathological diagnosis of FTLD-U/ TDP43 consistent with that observed in association with PGRN mutation.

3.0 Intra-Familial Clinical Heterogeneity due to FTLD-U

with TDP43 Proteinopathy Caused by a Novel Deletion in

Progranulin Gene (PGRN)

Tomasz Gabryelewicz*, Mario Masellis*, Mariusz Berdynski*, Juan M. Bilbao, Ekaterina

Rogaeva, Peter St. George-Hyslop, Anna Barczak, Krzysztof Czyzewski, Maria Barcikowska,

Zbigniew Wszolek, Sandra E. Black and Cezary Zekanowski

As published in: J Alzheimers Dis (2010); 22: 1123-1133.

Mario Masellis extracted the clinical information on the brother of the proband, interpreted and integrated the clinical, neuropsychological, neuroimaging, genetic and pathological data, and was responsible for writing a significant proportion of the manuscript with contribution from Tomasz Gabryelewicz. Sequencing of the proband and genotyping of the controls was performed by Mariusz Berdynski and Cezary Zekanowski. Ekaterina Rogaeva performed the genotyping in the brother of the proband. Pathological analysis was done by Juan Bilbao.

* These authors contributed equally to the work as co-first authors

3.1 ABSTRACT

Frontotemporal dementia (FTD) is one of the commonest forms of early-onset dementia, accounting for up to 20% of all dementia patients. Recently, it has been shown that mutations in progranulin gene (PGRN) cause many familial cases of FTD. Members of a family affected by

FTD spectrum disorders were ascertained in Poland and Canada. Clinical, radiological, molecular, genetic, and pathological studies were performed. A sequencing analysis of PGRN exons 1-13 was performed in the proband. Genotyping of the identified PGRN mutation and pathological analysis was carried out in the proband‟s brother. The onset of symptoms of FTD in the proband included bradykinesia, apathy, and somnolence followed by changes in personality, cognitive deficits, and psychotic features. The proband‟s clinical diagnosis was FTD and parkinsonism (FTDP). DNA sequence analysis of PGRN revealed a novel, heterozygous mutation in exon 11 (g.2988_2989delCA, P439_R440fsX6). The mutation introduced a premature stop codon at position 444. The proband‟s brother with the same mutation had a different course first presenting as progressive non-fluent aphasia, and later evolving symptoms of behavioral variant of FTD. He also developed parkinsonism late in the disease course evolving into corticobasal syndrome. Pathological analysis in the brother revealed

Frontotemporal Lobar Degeneration-Ubiquitin (FTLD-U)/TDP43 positive pathology. The novel

PGRN mutation is a disease-causing mutation and is associated with substantial intra-familial clinical heterogeneity. Although presenting features were different, rapid and substantial deterioration in the disease course was observed in both family members.

Keywords: corticobasal syndrome, frontotemporal dementia, haploinsufficiency, parkinsonism, progranulin mutation, progressive non-fluent aphasia

3.2 INTRODUCTION

Frontotemporal dementia (FTD) is a clinically, genetically, and neuropathologically heterogeneous disorder, accounting for 20% of early-onset dementia [Neary et al.

1998;Neumann et al. 2009]. FTD is characterized by behavioral and language dysfunction, without amnesia, and consensus clinical and pathological diagnostic criteria have been proposed

[McKhann et al. 2001;Neary et al. 1998;The Lund and Manchester Groups 1994;Cairns et al.

2007].

Progranulin gene (PGRN, GRN [OMIM 138945]) mutations were shown to be common in sporadic and familial FTD [Baker et al. 2006;Cruts et al. 2006;Gass et al. 2006]. PGRN is a 593 amino acid glycoprotein, composed of 7.5 evolutionary conserved tandem repeats, which are cleaved, forming a family of granulin peptides. It is a growth factor important in neural development [Ahmed et al. 2007]. A haploinsufficiency mechanism was identified to be the etiology underlying PGRN-associated neurodegeneration, which causes frontotemporal lobar degeneration with ubiquitin-positive, tau-negative inclusions (FTLD-U) [Baker et al. 2006;Cruts et al. 2006]. TDP43 was found to be the major pathological protein underlying FTLD-U pathology [Neumann et al. 2006].

From a clinical perspective, there is much to learn about how specific symptoms of FTD map onto FTLD pathological subtypes. PGRN mutations have been associated with substantial phenotypic heterogeneity in clinical presentation with a variety of diagnoses being observed: behavioral variant FTD (bvFTD), progressive non-fluent aphasia (PNFA), corticobasal syndrome

(CBS), Alzheimer's disease, parkinsonism, and FTD with Parkinson‟s (FTDP) [Benussi et al.

2008;Kelley et al. 2009;Masellis et al. 2006;Rademakers et al. 2007;Rohrer et al. 2009;Yu et al.

2010]. This clinical heterogeneity results from the same PGRN mutation causing pathology in different hemispheres and lobar regions [Rademakers et al. 2007]. The molecular mechanism underlying this clinical variability in different family members is unknown.

In this report, we describe the clinical, neuropsychological, and radiographic features at onset and longitudinally in two brothers from the first Polish kindred identified to have a novel PGRN mutation. Pathological characterization was performed in the index case‟s brother.

3.3 MATERIALS AND METHODS

3.3.1 Subjects

Genealogical data was ascertained in Poland. The proband was living in Warszawa, Poland. His brother was living in Toronto, Canada. They underwent assessment in specialized dementia clinics. Clinical evaluation included history, physical examination, and cognitive screening.

Routine biochemical screening was done. Brain MRI and SPECT were performed.

Neuropsychological, neuropsychiatric, and functional measures were completed. Baseline and ten month follow-up data are presented for the brother.

The case-control group for genetic analysis consisted of 90 patients with familial or sporadic

FTD (age-matched) and 200 elderly, neurologically healthy controls from the Polish population.

All participants or their relatives provided written, informed consent in accordance with the

Helsinki Declaration and the study was approved by the appropriate ethics committees.

3.3.2 Genetic analysis

DNA was isolated from peripheral leukocytes using standard procedures [Zekanowski et al.

2003b]. Intronic primers were used to amplify and sequence all (1-13) PGRN exons [Baker et al.

2006;Cruts et al. 2006]. Additionally, all exons of MAPT and PSEN1 were amplified and sequenced to exclude mutations or rare polymorphisms [Zekanowski et al. 2003b;Zekanowski et al. 2003a]. Amplification products were purified with ExoSAP IT (USB) and sequencing was carried out using the BigDye Terminator v1.1 Cycle Sequencing Kit (Applied Biosystems) and the ABI PRISM 3130 Genetic Analyzer (Applied Biosystems). RNA was extracted from the proband‟s leukocytes using TRIzol reagent (Ambion) according to manufacturer's instructions. cDNA prepared from 5 μg of RNA using Superscript II (Invitrogen) was used as a template for quantitative PCR with Power SYBR Green PCR Master Mix on an ABI PRISM 7500 instrument

(Applied Biosystems) according to manufacturer‟s` protocols. Relative 2-ΔΔCt method with ACTB as a reference gene was used to estimate levels of PGRN mRNA. Primers were designed for

PGRN cDNA: forward 5'-ATCCAGAGTAAGTGCCTCTCCAA-3', reverse 5'-

CTCACCTCCATGTCACATTTCAC-3', and for ACTB: 5'-CCGCAAAGACCTGTACGCCA-3' and 5'-TGGACTTGGGAGAGGACTGG-3'.

Absence of mutated mRNA was confirmed using the PCR method with reverse primer specific for the frameshifted region (5‟-GTCTGCTGCTCGGACCAC-3‟ and 5‟-

GTCACAGCCGATGTCTCG-3‟). Absence of the mutation in the case-control groups was confirmed using restriction fragment length polymorphism analysis with AvaI (Fermentas) or direct sequencing.

3.3.3 Neuropathological analysis

Paraffin-embedded sections were stained with haematoxylin and eosin, Luxol fast blue,

Bielschowski and Gallyas. Immunostains using commercial antibodies for tau (Dako, A0024), ubiquitin (Vector Labs, ZPU576), α-synuclein (Vector Labs), and subsequently with TDP43

(ProteinTech Group, Inc.) were performed.

3.4 RESULTS

3.4.1 Clinical, neuropsychological, and radiographic features

Proband (III:1). The proband was a 65-year-old right-handed male with no medical history. He had 16 years of education and worked as a managing director of a company. At age 62, the first symptoms were slowness, apathy and somnolence. The patient became withdrawn, less talkative, gave up hobbies and had trouble handling familiar objects. After several months, his social judgment deteriorated with a breakdown in formalities. He became disinhibited and significant personality changes were observed. He developed cognitive symptoms thereafter including aphasia and memory impairment.

Two years later (age 64), he stopped working and driving. Urinary incontinence occurred. The patient underwent neurologic assessment and had evidence for dementia and parkinsonism.

Mini-Mental State Exam (MMSE) was 20. The patient deteriorated rapidly over the next few months with insomnia and psychotic symptoms. He had significant irritability when opposed.

Motor re-examination showed moderately impaired monotone, slurred speech; minimal hypomimia; resting tremor of upper extremities, moderate in amplitude; moderate rigidity;

74 severe motor slowness and multi-step turning with postural instability. The symptoms progressed throughout the ensuing observation period.

Neuropsychological assessment (age 64) showed impairment of executive functions, speech, attention, and visuospatial functions (Table 1, III:1). He had spared autobiographical memory.

Word-finding difficulties were pronounced both in spontaneous speech and in verbal fluency tasks with perseverations. He had problems switching between categories. The proband‟s verbal learning was impaired, with a flat, plateau-like curve, and with intact delayed memory. Working memory was severely disturbed. Copy of the Rey-Osterrieth Complex Figure was disorganized with visuospatial and perseverative errors; most details were omitted on its delayed reproduction.

Naming and visual gnosis was intact. The patient had problems with gesture and spatial praxis because of difficulties in motor switching. Sequencing of motor learning was severely impaired with perseverations. This was also manifest as disturbed reciprocal coordination with a strong tendency to repeat only one motor action without inhibition. The patient required help in dressing and showering, and falls occurred daily. The patient manifested loss of initiative and a lack of interest in daily routine activities. He had difficulties with speech and his handwriting became illegible. Psychiatric examination showed psychotic features, including visual hallucinations (faces on windows), bizarre delusions, and misidentifications.

Table 1. Raw scores on neuropsychological and functional measures for proband (III:1) and proband’s brother (III:2). III:2 III:2 III:1 Session1 Session 2 Demographics Age of Onset (years) 55 - 62 Age at testing (years) 57 58 64 Duration of disease at testing (years) 2 3 2 Years of education 18 - 16

Neuropsychological Battery and Functional Measures (Test name/Maximum raw score) General cognition Folstein’s Mini-Mental Status Examination /30 19** 9** 22** Mattis Dementia Rating Scale /144 96** N/T N/A Blessed Information, Memory and Concentration Scale N/A N/A 32 /37 Memory California Verbal Learning Test - Long Delay Free Recall 1** N/T N/A /16 Delayed Visual Reproduction /41 0** N/T N/A Auditory verbal learning of 10 words list /First N/A N/A 4/6/5 attempt/last attempt/delayed reproduction Rey Osterieth Complex Figure - reproduction /36 N/A N/A 6** Address item from BIMC/5 N/A N/A 2 Language Western Aphasia Battery - total /100 67.8** 40.4** N/A Western Aphasia Battery – Aphasia Category Anomic Broca’s N/A Western Aphasia Battery – Spontaneous Speech 7** 2** N/A Content Western Aphasia Battery – Spontaneous Speech Fluency 5** 1** N/A Western Aphasia Battery - comprehension /10 7.9** 5.8** N/A Western Aphasia Battery – Repetition /10 8.4** 6.9** N/A Western Aphasia Battery – Naming /10 5.6** 4.5** N/A Boston Naming /30 13** N/T N/A Boston Naming /20 N/A N/A 20 Semantic Fluency 6* 1** 10* Praxis Western Aphasia Battery - praxis /60 52** 34** N/A Praxis of gesture /5 N/A N/A 4* Reciprocal coordination I/10 N/A N/A 5 II/10 10* Motor sequences learning I/5 N/A N/A 3

II/5 2 Attention & working memory Digit span - forward /12 6 N/T 4* Digit span - backward /12 3* N/T 3* Serial 7’s test /14 N/A N/A 1** Visuospatial abilities Rey Osterieth Complex Figure - copy /36 36 26.5 28 Benton Line Orientation /30 28 22 N/A Visual gnosis /17 N/A N/A 14 Executive functions Phonemic fluency (F-, A-, S-words) 3** N/T N/A Phonemic fluency (K-words) N/A N/A 2** Wisconsin Card Sort Test - categories /6 1** N/T N/A Wisconsin Card Sort Test - perseverative errors 40** N/T N/A Activities of daily living Disability Assessment for Dementia (%) 96 24** N/A Neuropsychiatric symptoms Neuropsychiatric Inventory - total /144 4** 14** 23** Neuropsychiatric Inventory- delusions/12 0 0 2** Neuropsychiatric Inventory- hallucinations/12 0 0 6** Neuropsychiatric Inventory - euphoria /12 2** 4** 0 Neuropsychiatric Inventory- anxiety/12 0 0 2** Neuropsychiatric Inventory - apathy /12 2** 4** 6** Neuropsychiatric Inventory – depression /12 0 0 1** Neuropsychiatric Inventory - disinhibition /12 0 2** 2** Neuropsychiatric Inventory – Irritability /12 0 0 4** Neuropsychiatric Inventory - appetite /12 0 4** 0 Cornell Depression Scale (%) 8 3 11 Session 2 scores were obtained 10 months after session 1 scores for III:2. Unmarked scores are normal based on comparison to healthy population matched for age and years of education. *Borderline-Impaired; **Impaired; N/T = Not testable; N/A = Not available

He was diagnosed with FTDP based on neurologic, psychiatric, physical, and neuropsychological examinations. Brain MRI and SPECT results supported this diagnosis and correlated with his symptoms and findings (Figure 3). Specifically, there was atrophy in the right anterior temporal region and bifrontally, more prominent on the right. There was reduced perfusion bifrontally, more prominent on the right and extending into the right superior parietal region.

Proband’s brother (III:2). The proband‟s brother was a right-handed male with no relevant medical history. He was assessed at age 57. He spoke Polish and English fluently. He had 18 years of education. Two years prior, he first presented with the insidious onset and gradual decline in speech fluency; he had frequent word-finding difficulties that interrupted verbal output. He often reverted to his native tongue. Comprehension was intact. However, he continued to work as an engineer.

MMSE was 22/30. Spontaneous speech revealed word-finding difficulties with no paraphasic errors. Comprehension, repetition, naming, and reading were intact. A written description of the

Cookie Theft Picture revealed use of simplified sentences with a sparse, but accurate description.

There was mild impairment in working memory and executive functions. His neurological exam was normal except for mild increase in tone in the right arm with contralateral limb activation.

The initial diagnosis was PNFA.

Four months later, neuropsychological testing revealed moderate impairments in most domains with relative sparing of visuospatial and visuoconstructive tasks (Table 1, session 1, III:2). On the Western Aphasia Battery (WAB), his category was anomic. The aphasia over-estimated his deficits. He, however, remained independent functionally with only minor troubles having a phone conversation and taking messages. Initial MRI revealed bilateral frontal > anterior temporal atrophy, which was prominent on the left (Figure 1A, B, C). Corresponding brain

SPECT revealed left > right bifrontal hypoperfusion extending into the left parietal region

(Figure 1B, C).

Figures 1 and 2. T1-weighted brain MRI and corresponding 99mTc-ECD brain SPECT images of proband‟s brother (III:2) in radiographic axial orientation. Asymmetric atrophy on MRI is seen affecting the left frontal > parietal regions with ventricular enlargement (III:2 – Session 1) which progresses as seen in Figure 2 (III:2 – Session 2). Perfusion deficits in the left > right frontoparietal regions in Figure 1 (III:2 – Session 1) also progress to more bilateral involvement along with left temporal involvement seen in figure 2 (III:2 – Session 2). Figure 3. T1-weighted brain MRI and corresponding 99mTc-HMPAO (800MBq) brain SPECT images of proband (III:1) in standard radiographic axial orientation. Bilateral frontal and temporal regions demonstrate significant atrophy with ventricular enlargement seen on axial slices of T1 weighted images in MRI. Corresponding axial images of functional SPECT showing perfusion defect in frontal and temporal regions, bilaterally. There was a predilection for the right hemisphere both in terms of atrophy and perfusion deficits. Orange-yellow color represents areas of normal perfusion on SPECT, while blue-purple color represents relative decreases in perfusion. AT=anterior temporal; PT=posterior temporal; O=occipital; IF=inferior frontal; IP=inferior parietal; SF=superior frontal; SP=superior parietal.

Clinical assessment seven months later (age 58) revealed deterioration in multiple spheres of cognition, behavior and function. He perseverated and had difficulties shifting sets. He giggled excessively. He ate quickly cramming food into his mouth and pocketing it in his cheeks. He developed a craving for chocolate. He became disinhibited and impulsive. He stopped maintaining oral hygiene and had trouble eating with utensils. He had difficulties arising from a chair and climbing stairs. His gait was slow with decreased arm swing on the right. Formal

79 testing of praxis revealed both conceptual and ideomotor deficits. His score on the Frontal

Behavioral Inventory was 29, above the cut-off indicating FTD. The diagnosis remained PNFA, but his syndrome evolved to include bvFTD.

Prospective re-evaluation on neuropsychological testing ten months after his first session revealed significant deterioration (Table 1, session 2, III:2). MMSE was 9/30. WAB category indicated a Broca‟s aphasia. He remained within normal limits on visuospatial tasks. From the neuropsychiatric perspective, there was evidence for euphoria, disinhibition, apathy, and appetite dysregulation. A repeat brain MRI demonstrated worsening atrophy of left > right frontotemporoparietal regions (Figure 2A-C).

He became incontinent. He spoke with one word at a time. He was unable to follow instructions.

He required constant supervision. Physical exam revealed worsening parkinsonism with hypomimia, right > left rigidity, difficulties arising from a chair, decreased right arm swing, stooped posture and festinating gait. Re-evaluation on SPECT revealed progressive global perfusion deficits with occipital sparing (Figure 2A-C). With the emergence of an asymmetric akinetic-rigid syndrome associated with apraxia, his final diagnosis evolved to include CBS.

Eventually, he progressed to full mutism. At age 60, he was bed-ridden. He developed progressive dysphagia. He passed away six years after disease onset (age 61) from complications due to his neurodegeneration.

3.4.2 Neuropathology (III:2)

The brain weighed 1,230 grams. Macroscopic examination disclosed atrophy of the frontal lobes, worse on the left. Temporal and left parietal involvement was present. There was atrophy of the caudate head.

Microscopic examination revealed severe pan-cortical atrophy, worse in anterior frontal regions with microvacuolization. There was cell loss, gliosis, and pallor of the subcortical white matter.

Ubiquitin-positive threads co-localized with the microvacuolar changes. Many neurons displayed

“comma”-shaped perinuclear inclusions. Rare ubiquinated intranuclear inclusions were demonstrable. Ubiquinated inclusions were abundant in the cingulum, mesiofrontal lobe, precentral gyrus, temporal and parietal lobes, but less so in the latter with segmentally spared areas. A dramatic decrease in ubiquitin pathology was noted in transition from the precentral to postcentral gyrus. Primary visual cortex was spared. Silver stain and immunostaining for tau and

α-synuclein was negative.

Subcortical grey matter revealed neuronal ubiquinated granular and filamentous inclusions in caudate, putamen, thalamus, posterior hypothalamus and nucleus accumbens. Globus pallidus and nucleus basalis of Meynert were spared. In limbic regions, the cornu ammonis of CA1 was severely gliotic and shrunken. Microvacuolar changes involving the parahippocampal gyrus were noted, with sparing of the perirhinal cortex. Ubiquitin positive inclusions were observed in neurons of the fascia dentata.

The anterior 1/3 of the cerebral peduncles bilaterally was degenerated. The midbrain was small.

Estimated cell loss in the substantia nigra was 60% with severe gliosis and macrophages present.

There was no immunostaining for tau or α-synuclein. The pons was small with pallor of the descending tracts. In the medulla, there was no α-synuclein staining. Rare neurons in the inferior olive contained filamentous ubiquinated inclusions. Medullary motor nuclei were intact with no ubiquinated inclusions observed. The cerebellum was unremarkable. Motor neurons in the spinal cord were not affected.

Autopsy sections were re-examined with immunostains for TDP43 (Figure 4). TDP43 positive neuropil threads, neuronal cytoplasmic stippled staining, neuronal cytoplasmic filamentous inclusions, glial [oligo] cytoplasmic and neuronal intranuclear inclusions were found in the frontal cortex, anterior striatum, fascia dentata, substantia nigra, and CA1 region. Final pathological diagnosis was FTLD-U/TDP43 proteinopathy.

Figure 4. Micrographs demonstrating a large number of TDP43 inclusions (neuropil threads, neuronal cytoplasmic stippled staining, neuronal cytoplasmic filamentous inclusions, glial [oligo] cytoplasmic and neuronal intranuclear inclusions) found in the fascia dentata, substantia nigra, and CA1 region.

3.4.3 Family history

There was a strong family history of early-onset dementia and parkinsonism, suggesting autosomal dominant inheritance (Figure 5). The proband‟s mother (II:2) died at age 64, with a surmised progressive aphasia. Age of onset was 60. The maternal aunt had parkinsonism and

82 dementia and died ca. 65 years (II:3). The proband‟s father (II:1) was neurologically intact and died at age 62 of lung cancer.

3.4.4 Genetic analysis

A novel PGRN dinucleotide deletion in exon 11 (g.2988_2989delCA, c.1536_1537delCA,

P439_R440fsX6) was identified in the proband and his affected brother (Figure 5). Both the sense and the complementary DNA strand were sequenced. The mutation causes a frameshift at codon 441, and introduces a stop codon at position 444. The mutation was absent in a group of

90 Polish patients with FTD (mean age=59.7 ± 13 years) and 200 ethnically matched neurologically healthy controls (mean age=72.7 ± 7 years; MMSE≥28).

RT-PCR analysis of PGRN mRNA levels in peripheral leukocytes from the proband revealed a two-fold decrease of the cDNA transcript as compared to controls without the mutation. PCR using a primer specific to the mutant cDNA resulted in absence of amplification product (Figure

6).

Figure 5. Detection of PGRN mutation P439_R440fsX6. A) Pedigree showing family history of neurodegenerative condition. Black symbols: patients affected with FTD and neurodegeneration; white symbols: unaffected individuals or individuals with no clinical diagnosis available. B) Electropherogram showing start of deletion marked with an

84 arrow. The resulting PGRN mutation P439_R440fsX6 is shown at the bottom of the chromatogram of the proband and his affected brother.

Figure 6. Amplification from genomic DNA (gDNA; lane 1) using primers specific for the mutant allele demonstrate the mutant fragment of 153 bp as expected. Amplification from cDNA (lane 2) shows an absence of the expected product supportive of non-sense mediated decay. Ladder: GeneRuler 1kb DNA Ladder (Fermentas), the lowest band is 250 bp (lane 3); positive control: cDNA amplified 84 bp fragment of β-actin gene (lane 4).

3.5 DISCUSSION

We describe a novel PGRN mutation causing a frameshift introducing a premature stop codon.

RT-PCR analysis of PGRN mRNA levels confirmed the PGRN transcript decrease in the proband as compared to normal. Additionally, no amplification products of the mutant allele were detected suggesting that mRNA with the premature stop codon is rapidly degraded as a result of non-sense mediated decay [Baker and Parker 2004]. There are no doubts about the pathological nature of this mutation. It segregates with the disease in two affected family members, it is absent in 200 normal controls, and immunohistochemistry confirms FTLD-

U/TDP43 pathology associated with mutant PGRN.

Patients affected with different PGRN mutations showed a broad range of age of onset (AOO;

48-83 years), with a mean of 597 years, often resulting in no family history recorded [Brouwers et al. 2008;Gass et al. 2006]. Another study also showed a highly variable AOO ranging from 49 to 88 years, with variable disease duration ranging from one to 14 years [Kelley et al. 2009].

This novel PGRN deletion is associated with a rapid disease course and clear inheritance pattern.

Consistent with other studies, AOO was variable with the proband‟s brother developing symptoms seven years earlier.

The clinical course of FTD in the two siblings was different, particularly at illness onset (Table

1). The proband‟s clinical features suggested early medial and dorsolateral prefrontal involvement with slowing, lack of motivation, and apathy. Shortly thereafter, social impairment and disinhibition were present, suggesting progression to orbitofrontal and right anterior temporal structures. Parkinsonism was also present. The behavioral disturbance correlated well

86 with bifrontal and anterior temporal atrophy and hypoperfusion, worse on the right (Figure 3).

Language problems were observed later than behavioral impairment. The proband also had psychotic features including hallucinations, which is atypical in FTD.

In contrast, language disturbances came first in the proband‟s brother. These were expressive with an early anomia progressing to a Broca‟s aphasia and then to mutism. The initial symptoms of PNFA correlate with atrophy and hypoperfusion predominantly in the left frontal region

(Figure 1A-C). Later on, behavioral disturbance developed, which suggested progression to orbitofrontal and right anterior temporal regions, with social impropriety and disinhibition, culminating in apathy and cognitive decline. As the disease progressed from PNFA to include bvFTD so did the atrophy and perfusion deficits involving frontotemporal regions bilaterally

(Figure 2A-C). Apraxia was likely accounted for by the left frontoparietal involvement (Figures

1C and 2C) and these findings supported the third diagnosis of CBS. Visuospatial function was relatively preserved, correlating well with intact perfusion and absent pathology in the occipital regions, bilaterally.

The most common clinical presentation of PGRN mutation includes behavioral symptoms, with apathy as the dominant feature [Beck et al. 2008], similar to the proband. However, as is the case with his brother, clinical presentations of PNFA due to PGRN mutation are also frequent

[Snowden et al. 2006]. Several studies have confirmed this strong association between PNFA and PGRN mutations with the typical FTLD-U/TDP43 pathology [Beck et al. 2008;Moreno et al. 2009;Pickering-Brown et al. 2008;Skoglund et al. 2009]. In particular, similar to the brother‟s pathological findings, FTLD-U, type 3 pathology, was found to be most commonly associated

87 with the clinical phenotype of PNFA [Snowden et al. 2007]. In one series, semantic dementia cases were associated with MAPT mutations whereas PNFA with associated apraxia predicted

PGRN mutations [Pickering-Brown et al. 2008]. These particular case series were enriched with familial forms of FTD or were selected for based on a priori identification of PGRN mutation.

Studies of predominantly sporadic cases of primary progressive aphasia selected for based on availability of pathological material demonstrated the opposite trend. Specifically, non-fluent presentations were associated with Tau pathology [Josephs et al. 2006;Knibb et al. 2006], while fluent cases were associated with ubiquitin pathology [Knibb et al. 2006]. Longitudinal studies of familial and sporadic aphasic variants of FTD followed clinically until death with subsequent pathological characterization are warranted to clarify these apparent discrepant findings.

In the current study, both the proband and his brother developed parkinsonism. Indeed, FTD and parkinsonism due to PGRN mutation is common [Josephs et al. 2007;Wong et al. 2009] and is more variable than that due to FTDP-17 with MAPT mutations [Boeve and Hutton 2008]. In the former, there are often posterior features, such as limb apraxia and visuospatial dysfunction, which results in a wider clinical spectrum of diagnoses including dementia with Lewy bodies or

CBS [Boeve and Hutton 2008].

In general, the clinical heterogeneity and course of the affected siblings with this novel

P439_R440fsX6 dinucleotide deletion resembles the course of other FTD patients with short- segment nucleotide PGRN deletions [Benussi et al. 2008;Borroni et al. 2008a;Llado et al.

2007;Skoglund et al. 2009]. The particular type of mutation does not predict the clinical syndrome, but rather it is the location of the pathology which is most significant.

To date, significant progress has been made in understanding the allelic heterogeneity of PGRN mutation in FTD. This paper extends the literature on the allelic and phenotypic heterogeneity of

FTD. However, progress in terms of understanding the variable clinical presentation of FTD, i.e., specific diagnoses, age of onset, hemispheric and specific lobar involvement and duration of disease remain to be explained. Studies examining polymorphism within PGRN miRNA binding sites and peripheral expression levels of PGRN may help to shed light on this phenotypic heterogeneity [Finch et al. 2009;Rademakers et al. 2008]. Using the approach of early identification of those at risk of developing FTD by imaging and CSF biomarkers coupled with a better understanding of genetic, epigenetic, and environmental modulators of disease will facilitate future development of preventative treatments and/or disease-modifying therapies for these devastating FTD syndromes.

3.6 ACKNOWLEDGMENTS

The authors would like to thank Dr Jaroslaw B.Cwikla from Dep. of Radiology and Diagnostic

Imaging, Medical Centre for Postgraduate Education and CSK, MSWiA in Warsaw for comments and creating figure 1. The authors would also like to thank Mr. Mike Misch, Gregory

Szilagyi, and Mark Gravely for creating Figures 1, 2, and 3 and Ms. Isabel Lam for creating

Table 1. MM is supported by a Canadian Health Institutes of Research (CIHR) Clinician

Scientist Award and the Department of Medicine, Sunnybrook Health Sciences Centre. This research is supported by operating grants from the CIHR (SEB, MT13129; PSGH & ER,

MT417763) and the Ontario Research Fund (PSGH). ZW is supported by NIH/NINDS

1P50NS072187-01, 1RC2NS070276-01, 1R01NS057567-01A2; Carl Edward Bolch, Jr. and

Susan Bass Bolch Gift, and Mayo Clinic Florida Research Committee. CZ and MB are supported by grant PBZ-MEiN-0/2/20/17.

4.0 Ideomotor Apraxia in Corticobasal Syndrome: Brain

Perfusion and Neuropsychological Correlates

Mario Masellis, Philip L. Francis, Kie Honjo, Bradley J. MacIntosh, Isabelle Guimont, Gregory

M. Szilagyi, Wendy R. Galpern, Galit Kleiner-Fisman, James L. Kennedy, Robert Chen, Eric A.

Roy, Curtis B. Caldwell, Anthony E. Lang, Sandra E. Black

As submitted to: Cortex

Mario Masellis clinically assessed several of the patients included in this study, extracted the clinical information, designed the study, performed the data analysis and wrote the manuscript. Philip Francis assisted with the SPM analysis of SPECT data and Kie Honjo assisted with the MRI segmentation procedure. Brad J. MacIntosh assisted with the atrophy correction procedure. Wendy R. Galpern, Galit Kleiner-Fisman and Anthony E. Lang assessed and collected clinical data on patients ascertained from a movement disorders clinic.

4.1 Abstract

Ideomotor apraxia is one of the most common clinical features of corticobasal syndrome and is associated with disability and reduced quality of life. Previous electrophysiological and neuroimaging studies of apraxia implicated a role of the left frontoparietal network. However, the specific nodes within this network have yet to be fully elucidated. The current study provides the first direct correlative analysis between the severity of ideomotor apraxia in corticobasal syndrome and cerebral perfusion imaging using brain SPECT. Reductions in perfusion within the left inferior parietal lobule (t=5.7, p=0.03, Family-Wise Error [FWE] corrected), including the left angular gyrus (t=5.7, p=0.02, FWE corrected), were associated with more severe ideomotor apraxia as measured by the Western Aphasia Battery praxis scale. Results remained significant even after controlling for the most affected side of the body. After categorizing the patients into those with or without apraxia, language, visuospatial and visual memory functions were more impaired in those with apraxia suggesting the involvement of overlapping networks, specifically, bilateral occipitoparietal and left peri-Sylvian, subserving these related higher cognitive processes. This study provides further evidence for the importance of the left inferior parietal lobule in the dominant hemisphere frontoparietal praxis network and provides new insights into associated cognitive dysfunction.

Keywords: apraxia; SPECT; perfusion; neuropsychology; corticobasal syndrome

4.2 Introduction

Corticobasal Syndrome (CBS) is a rare and debilitating neurodegenerative syndrome characterized by asymmetric rigidity, apraxia, dystonia, myoclonus, alien-limb phenomenon, cortical sensory loss, frontosubcortical dementia, behavioral disturbances, and speech and language abnormalities including apraxia of speech and progressive non-fluent aphasia (PNFA).

There is significant pathological heterogeneity that can produce the syndrome including corticobasal degeneration, progressive supranuclear palsy (PSP), frontotemporal lobar degeneration (FTLD)-Tau (Pick‟s disease) and FTLD-Ubiquitin/TDP43, and Alzheimer‟s disease [Kertesz et al. 2005;Lee et al. 2011;McMonagle et al. 2006;Wadia and Lang 2007].

Apraxia is the hallmark that distinguishes CBS from other parkinsonian disorders in the early stages of disease and it is the most common clinical feature occurring cross-sectionally in 70% and longitudinally in 100% of cases [Leiguarda et al. 1994;Rinne et al. 1994;Stamenova et al.

2009]. Apraxia is defined as a higher-order “neurological disorder characterized by loss of the ability to execute or carry out skilled movements and gestures, despite having the desire and the physical ability to perform them” (http://www.ninds.nih.gov/disorders/apraxia/apraxia.htm).

There are many types of apraxia observed in CBS including apraxia of speech, limb-kinetic apraxia, ideomotor apraxia (IMA), and ideational/conceptual apraxia, which have been extensively described elsewhere [Gross and Grossman 2008;Josephs and Duffy 2008;Leiguarda and Marsden 2000;Stamenova et al. 2009;Zadikoff and Lang 2005].

Ideomotor apraxia, best elicited by asking a patient to pantomime and/or imitate hand gestures and tool use, is characterized by disturbances of timing, sequencing and spatial organization of

93 gestural movement of the limbs [Rothi et al. 1991]. It has been the most extensively studied apraxia type in CBS [Zadikoff and Lang 2005], although there have been only a paucity of studies that have directly correlated specific measures of praxis with brain imaging findings in this disorder. Peigneux et al. [Peigneux et al. 2001] were the first to examine the association of upper limb apraxia with fluorodeoxyglucose positron emission tomography (FDG-PET) imaging in a case series of 18 patients with CBS. Their sample was stratified into CBS with and without apraxia based on a standardized praxis measure. Using a global praxis performance score, the bilateral anterior cingulate gyri demonstrated mild reductions in metabolism in the apraxic group

(uncorrected p < 0.001) [Peigneux et al. 2001]. Alternatively, stratification using a praxis correction score resulted in hypometabolism contralateral to the most affected body side in the superior parietal lobule, medial frontal gyrus and supplementary motor area, as well as the middle frontal gyrus in the apraxic group (uncorrected p < 0.001) [Peigneux et al. 2001]. This study, however, did not correlate PET images of metabolism with praxis measures, nor did their imaging analysis correct for multiple comparisons on a voxel-by-voxel basis. Other small PET and single photon emission computed tomography (SPECT) studies of apraxia have been conducted in CBS samples, but these did not specifically look at the relationship between praxis measure and functional imaging; rather they were a comparison of CBS versus controls and only indirect associations with apraxia were made [Zadikoff and Lang 2005].

Functional neuroimaging studies with FDG-PET and SPECT have shown reduced metabolism and perfusion, respectively, in frontoparietotemporal regions in CBS patients compared to controls [Eidelberg et al. 1991;Garraux et al. 2000;Markus et al. 1995;Okuda et al. 1999]. The hypoperfusion tends to be contralateral to the most affected side of the body. Similarly,

94 asymmetrical atrophy on MRI can also be seen in CBS and the atrophy usually is more prominent contralateral to the most affected side of the body [Riley et al. 1990;Savoiardo et al.

2000;Soliveri et al. 1999]. Perfusion SPECT was used in the current study given that prior research of focal cortical atrophy syndromes, such as FTLD, have shown that perfusion reductions on SPECT are more extensive than atrophy detected on MRI in the early stages of disease and in longitudinal follow-up, indicating increased sensitivity of this modality as a potential biomarker [Gregory et al. 1999;Gabryelewicz et al. 2010;Mendez et al. 2007].

The primary objectives of the current study were 1) to identify regions of reduced perfusion using brain SPECT in a prospectively recruited sample of CBS cases compared to controls, 2) to determine which of these regions directly correlate with performance on a standardized global measure of ideomotor praxis using the Western Aphasia Battery (WAB) [Kertesz and Poole

1974;Kertesz 2007] accounting for effects of lateralization of motor symptoms and underlying atrophy on MRI, and 3) to compare the demographic, clinical, neuropsychological and SPECT characteristics of CBS patients with significant apraxia to those without this feature defined based on performance on the WAB praxis subscale. The secondary objective of this study was to explore the different subcomponents of the WAB praxis scale and their association with brain

SPECT perfusion.

4.3 Materials and Methods

4.3.1 Subjects:

Thirty-one patients with a clinical diagnosis of CBS according to proposed diagnostic criteria

[Boeve et al. 2003] were recruited through two academic clinics: the Linda C. Campbell

Cognitive Neurology Research Unit at Sunnybrook Health Sciences Centre and the Movement

Disorders Centre at the Toronto Western Hospital, University Health Network. Patients provided informed consent to participate according to the Declaration of Helsinki and were followed as part of a prospective, longitudinal study of dementia and ageing approved by the local Research

Ethics Board. The patients underwent a detailed neurological exam, including a screening assessment for apraxia in both upper limbs comprised of asking them to pantomime five gestures

(two intransitive and three transitive ones). Patient handedness was determined using a standardized questionnaire [Bryden 1977]. The side of greatest rigidity and/or apraxia on clinical examination by a cognitive and/or movement disorders neurologist with expertise in the clinical assessment and diagnosis of CBS defined the motor-onset of symptoms. Although one side of the body was more prominently affected than the other side in all patients initially, motor signs, including apraxia, were indeed present bilaterally and became more evident as the disease progressed. Diagnostic consensus was achieved through review by at least two neurologists

(AEL, MM and/or SEB). All patients were followed longitudinally to ensure diagnostic accuracy. This is important because in advanced disease, the proportion of patients fulfilling three of the most commonly applied diagnostic criteria for CBS was similar at approximately

90%, indicating that all criteria could be applied equally well in late stage disease [Mathew et al.

2011].

4.3.2 Description of neuropsychological measures:

Neuropsychological tests assessing general cognitive functions included Folstein‟s Mini-Mental

State Examination (MMSE) [Folstein et al. 1975], and Mattis Dementia Rating Scale (DRS)

[Mattis 1976]. Measures of language function and naming included: the Western Aphasia

Battery (WAB), which calculates an aphasia quotient based on combined subscores of fluency, content, comprehension, repetition and naming, with a maximum score of 100 and lower scores representing more severe impairment [Kertesz and Poole 1974;Kertesz 2007]; the Boston

Naming Test (BNT) [Williams et al. 1989]; and semantic/categorical fluency [Gladsjo et al.

1999]. The visual reproduction subtest of the Wechsler Memory Scale-Revised (WMS-R) was used to assess visual memory [Lezak 1983]. Visuospatial function was assessed using the Rey-

Osterrieth Complex Figure Test [Lezak 1983;Osterrieth 1944;Rey 1941], and the Benton Line

Orientation task, which is motor-free and assesses visuospatial orientation and attention [Lezak

1983]. Additional standardized neuropsychological, neuropsychiatric and functional measures were performed as previously described [Masellis et al. 2006].

Praxis was assessed using the WAB praxis scale [Kertesz and Poole 1974;Kertesz 2007]. The

WAB is a valid and reliable measure of language and other higher cortical functions [Kertesz

2007] and the WAB praxis scale has been used to correlate stroke lesion localization and size with severity of ideomotor apraxia [Kertesz and Ferro 1984]. Briefly, patients were asked to pantomime gestures using their bucco-facial musculature and their less affected limb. Since CBS is strikingly asymmetric in presentation and since the apraxia most often co-exists in the same

97 limb where the extrapyramidal and cortical sensory features reside, the less affected limb was selected for praxis assessment scoring to avoid contamination with the sensory and extrapyramidal signs. Gestures fall into four categories: upper limb intransitive or transitive, bucco-facial, or complex. Examples include waving goodbye (intransitive gesture), using a toothbrush (transitive gesture), blowing out a match (bucco-facial gesture), and pretending to drive a car (complex gesture). There were five different gestures asked in each of the four categories. For all of the gestures, if the patient was unsuccessful at pantomime, they were asked to imitate the gestures. For transitive gestures, if they were unsuccessful at both pantomime and imitation, they were then handed the tool and asked to demonstrate how to use it. Three points were given if the gesture was performed correctly on pantomime; two points were given if there was approximate performance on pantomime or good performance on imitation only; one was given if there was approximate performance on imitation or if performed correctly with the actual tool or object; and no points were given if the patient was unable to perform the task, the gesture was unrecognizable or unrelated, and for erroneous use of the actual object. Approximate performance on gestural tasks was defined by the occurrence of the following types of errors: inaccurate positioning of the hand or limb in space, improper finger configuration, a breakdown in the core characteristics of the movement, and/or deficits in the sequence of an action, such as omission or addition of movement elements, as well as a change in the order in which an action should be carried out. Apraxia was scored out of 60 with lower scores indicating more severe

IMA [Kertesz and Poole 1974;Kertesz 2007]. Trained psychometrists with a Bachelor‟s or

Master‟s degree in Psychology administered all the tests, including the WAB, and were completely blind to all neuroimaging measures.

4.3.3 Brain SPECT acquisition and processing:

SPECT imaging employed a triple-head gamma camera (Prism 3000XP; Phillips Medical

Systems Inc., Cleveland, Ohio) and was performed a minimum of 30 minutes and a maximum of

120 minutes after injection of 20 mCi (740 MBq) of Technetium-99m ethyl cysteinate dimer

(99mTc-ECD SPECT). Patients were asked to rest quietly during the acquisition phase. 120 views were acquired uniformly over 360 degrees using all three detectors fitted with ultra-high resolution fan-beam collimators. Each view consisted of a 128 × 128 pixel image. Total imaging time was 19 minutes. Reconstruction was performed using a ramp-filtered back-projection algorithm followed by a 3-dimensional restoration post-filter (Wiener filter, multiplier 1.0).

Reconstructed image resolution was typically 9.7 mm full width at half maximum (FWHM).

Ellipses were fit to the approximate location of the outline of the head in each transaxial image, and a calculated attenuation correction applied [Matsuda et al. 1995]. Voxel dimensions were

2.18 × 2.18 × 3.56 mm.

4.3.3.1 Regional perfusion ratios:

Reconstructed SPECT images were co-registered to a template that was an average of 14 healthy, elderly control scans. A T1-weighted MRI with dimensions similar to the SPECT template was the source of 79 bilateral regions of interest (ROI) as previously described

[Lobaugh et al. 2000]. To obtain ROI intensity values, we used a common transformation to move from the SPECT template space to MRI space. The cerebellum is frequently used to normalize SPECT counts in studies of dementia [Stamatakis et al. 2001]. However, crossed cerebellar diaschisis may lead to relative differences in perfusion between the left and right cerebellar hemispheres, and, if whole cerebellum is used as the reference region in these cases,

99 regional cerebral blood flow (rCBF) may be overestimated in a particular ROI. We, therefore, applied the following rule: if there was more than a 5% difference in counts between left and right cerebellar hemispheres, we use the hemisphere that was more perfused as the reference region. If there was no difference then the whole cerebellum was used as the reference region. In this way, semiquantitative perfusion ratios can be derived and regional Z scores calculated

[Lobaugh et al. 2000].

4.3.4 Data analysis:

Statistical analysis of demographic, clinical, neuropsychological and ROI SPECT variables was performed using the Statistical Package for the Social Sciences (SPSS), version 16.

4.3.4.1 Demographic, clinical and neuropsychological measures:

Categorical demographic and clinical data were analyzed using chi-square or Fisher exact tests.

Normality of continuous demographic and neuropsychological data was assessed based on examination of Q-Q probability plots. Normally distributed data were analyzed using independent sample t-tests or ANOVA, otherwise, Mann Whitney U tests were performed.

4.3.4.2 Statistical Parametric Mapping (SPM) SPECT analysis:

SPECT scans were converted to Analyze 7.5 format. Statistical Parametric Mapping version 5

(SPM5, Wellcome Department of Imaging Neuroscience, University College London) was used for all imaging processing. Images were spatially normalized to a standard SPECT template in

Montreal Neurological Institute (MNI) space [Stamatakis et al. 2001] with re-sampling of voxel dimensions of 2 × 2 × 2 mm. Images were then smoothed using an isotropic Gaussian kernel (12

100 mm FWHM). A thresholded mean voxel value was chosen for global calculation, and global normalization was achieved by proportional scaling to an arbitrarily chosen constant value set at

50 mL/100 g/min. Voxel-by-voxel regression analysis was performed between perfusion and praxis measures. Alternatively, voxel-by-voxel analyses were performed using unpaired t-tests to compare 1) CBS patients to controls, and 2) apraxic to borderline/non-apraxic patients.

Covariates were incorporated if they were significantly different between groups. We reported significance using a voxel-wise p-value threshold (p < 0.05) corrected for multiple comparisons and an extent threshold of at least 20 contiguous voxels. Our correction methodologies included either controlling the family-wise error (FWE) rate [Worsley et al. 1996] or controlling the false discovery rate (FDR) [Genovese et al. 2002]. Controlling the FWE rate is more conservative but is known to be associated with type II errors. A whole brain mask was used to exclude extracranial voxels from the analysis. The maximal peak coordinates of the perfusion differences were converted to Talairach space using the Yale Non-linear MNI to Talairach

Converter [Lacadie et al. 2008] (http://www.bioimagesuite.org/Mni2Tal/index.html). These converted coordinates were translated into anatomical brain regions and Brodmann Areas (BAs) using Talairach Daemon Client [Lancaster et al. 2000] (http://www.talairach.org/client.html).

4.3.4.3 Region of interest (ROI) SPECT analysis:

4.3.4.3.1 Comparison of CBS cases to controls

Normality of ROI SPECT data was confirmed as described above. Independent sample t-tests between CBS and control groups were conducted to compare mean perfusion ratios of individual

ROIs within frontal, parietal, temporal, basal ganglia, and thalamic regions previously shown to be affected in CBS [Markus et al. 1995;Okuda et al. 1999]. ROIs that were statistically

101 significant on the t-test analyses were included in a multivariate analysis of covariance

(MANCOVA). „Years of education‟ was included as a covariate since it was the only demographic variable that differed significantly between the cases and controls.

4.3.4.3.2 Confirmation of voxel-wise correlation with WAB praxis using ROI method

Regions of hypoperfusion identified on the voxel-by-voxel regression analysis allowed us to pre- select the most relevant ROI to perform a Pearson correlation and linear regression analysis with

WAB praxis scores. This analysis provided for an independent confirmation of the voxel-wise findings using the bottom up approach for data reduction given the sample size.

Since there is left hemisphere specialization for praxis control, and since CBS is typically asymmetric in presentation, the potential confounding effect of symptom lateralization was controlled for in the SPM and ROI regression analyses by incorporation of right- versus left- sided motor presentation as a covariate.

4.3.4.4 Brain MRI acquisition and processing:

Structural MRI was obtained in 29 of the 31 patients using a standard protocol. Images were acquired on a 1.5T Signa MR imager (GE Medical Systems, Milwaukee, Wis.) and consisted of the following acquisitions: 1) T1-weighted (axial 3D spoiled gradient [SPGR] echo, with echo time [TE] 5 ms, repetition time [TR] 35 ms, flip angle 35°, number of excitations [NEX] 1, field of view [FOV] 22 × 16.5 cm, in-plane resolution 0.859 × 0.859 mm and slice thickness 1.2–1.4 mm), 2) proton-density and 3) T2-weighted images (interleaved axial spin echo, with TEs 30 and

102

80 ms, TR 3 s, NEX 0.5, FOV 20 × 20 cm, in-plane resolution 0.781 × 0.781 mm and slice thickness 3 mm).

4.3.4.4.1 Brain Extraction and Automated Tissue Segmentation:

Twenty-one of the 29 MRI scans were of sufficient quality to undergo semi-automated image analysis. Poor image quality was primarily due to head motion artifacts. Brain extraction and automated tissue segmentation were based on previously described methods [Kovacevic et al.

2002]. Images were co-registered to the T1-weighted image using the Automated Image

Registration package (AIR, v.5.2.3). T2/PD images were used collectively to extract brain and subdural/ventricular CSF, then the masked T1 was segmented using a T1-based protocol whereby local intensity histograms are fitted to four Gaussian curves to derive cut-offs for classifying each voxel as white matter, grey matter, or cerebrospinal fluid (CSF) [Kovacevic et al. 2002]. This is important for calculating the Total Intracranial Volume in correcting for head size, especially in focal atrophy syndromes like CBS. The methods of Kovacevic et al. have been updated and more details of the MRI image processing pipeline have been described [Ramirez et al. 2011].

4.3.4.4.2 Post-hoc MRI analysis:

A post-hoc analysis was performed on the FWE- and FDR-corrected group statistical maps.

These two maps were outputted as masks and transformed into each participant‟s MRI T1- weighted image space. Registering of the group SPECT SPM masked results into participant coordinate space allowed for the characterization of tissue types that were defined by the group

SPM result. The proportion of grey matter, white matter, and CSF underlying the SPM masks

103 was calculated for each patient. Mean and standard deviation values for each tissue class were calculated across all patients to estimate the degree of atrophy underlying the SPM masks.

4.4 Results

4.4.1 CBS vs. controls

4.4.1.1 Demographic data

Demographic feature CBS Controls (n=31) (n=31) Gender 19 (61.3%) F 19F 12 (38.7%) M 12M Handedness 29 (93.5%) R 29R 2 (6.5%) L 2L Age of Onset (mean  SEM years) 65.2  1.7 N/A Age at Investigation (mean  SEM years) 68.5  1.7 70.0  1.2 Duration of symptoms (mean  SEM years) 3.3  0 .4 N/A Years of Education (mean  SEM years)* 12.4  0.6 14.5  0.5 Initial body side most affected 16 (51.6%) R N/A 15 (48.4%) L Table 1. Demographics of patients with corticobasal syndrome (CBS) and control group. F=female; M=male; R=right; L=left; SEM = standard error of mean; N/A = not applicable. * t(60) = -2.7, p = 0.008; initial body side most affected was defined based on where most prominent motor symptoms were observed.

104

4.4.1.2 Clinical features

Clinical characteristics Frequency (%) at Frequency (%) at time of investigation follow-up (N=31) (N=31) Extrapyramidal features Rigidity (asymmetric) 28 (90.3%) 31 (100%) Dystonia (asymmetric) 16 (51.6%) 18 (58.1%) levodopa trial with poor 13 (41.9%) 13 (41.9%) response* Tremor – postural/action 8 (25.8%) 11 (35.6%) Cortical features Apraxia (asymmetric) 28 (90.3%) 31 (100%) Cortical sensory loss 19 (61.3%) 19 (61.3%) Alien-limb phenomenon 1 (3.2%) 3 (9.7%) Limb levitation 7 (22.6%) 10 (32.3%) Myoclonus 9 (29.0%) 13 (41.9%) Early dementia 22 (71.0%) 22 (71.0%) Language disturbance 24 (77.4%) 24 (77.4%) Table 2. Clinical characteristics of CBS sample. * 13 patients had a trial of levodopa and all responded poorly based on clinical assessment. Average time for emergence of additional signs on follow-up was 1.0 ± 0.3 years. All findings described above are based on clinical examination.

Rigidity and IMA were asymmetric in the early stages of the disease and eventually occurred in all patients. Early dementia was defined clinically according to DSM-IV criteria. Cortical sensory loss in this study was defined by the presence of one or more of the following abnormalities: extinction to double simultaneous tactile stimuli and/or astereognosis and/or agraphesthesia. Limb levitation was distinguished from true alien limb phenomenon.

4.4.1.3 SPM and ROI SPECT analyses

Two methods, SPM and ROI analyses, were used to compare perfusion differences between all

CBS cases and controls, CBS cases with left-sided symptoms (CBS-L) vs. controls, and CBS

105 cases with right-sided symptoms (CBS-R) vs. controls. See figure 1 and supplementary table 1, which provides anatomical locations of the reduced perfusion and the statistical results.

Figure 1. Statistical parametric maps (SPM) of bilateral frontal, parietal and temporal surface regions of the brain showing decreased perfusion in (A) all CBS cases compared to controls and (B) CBS cases with predominant symptoms on their left side (CBS-L) compared to controls overlaid on brain MRI template. N.B. Refer to Supplementary Table 1 for details of analysis and results. Green areas are corrected for multiple testing using Family-Wise Error methods, while red areas are corrected using False Discovery Rate methods.

Areas of significantly reduced perfusion among CBS individuals compared to controls using ROI and voxel-wise approaches were: bilateral dorsolateral prefrontal association cortices, bilateral primary sensorimotor cortices, bilateral anterior cingulate regions, right superior and inferior parietal lobules, left superior parietal lobule, right superior and middle temporal gyri, right

106 fusiform gyrus and right insula. The left supramarginal gyrus ROI showed reduced perfusion in

CBS vs. controls in the individual t-test analysis, but when incorporated into the GLM multivariate analysis, it showed a trend for significance (p=0.06). Three of these regions remained significant after correcting the FWE: right middle frontal gyrus, left superior frontal gyrus and left superior parietal lobule (Figure 1 and Supplementary Table 1). CBS patients with predominant left-sided symptoms as compared to controls demonstrated reduced perfusion in the same cortical regions as in the entire patient sample using the FDR correction except that these regions lateralized mainly to the right hemisphere. Areas of reduced perfusion in CBS-L versus controls were: superior and middle frontal gyri and post-central gyrus all lateralized to the right hemisphere (FWE corrected; Supplementary Table 1 and Figure 1). An area of reduced perfusion in CBS-R versus controls was the dorsal aspect of the left inferior frontal gyrus (ROI method).

When the stringency of the SPM analysis was reduced (uncorrected p-value <0.001), left frontoparietal regions including the left inferior parietal lobule demonstrated reduced perfusion in CBS-R vs. controls (data not shown). No significant areas of relative hyperperfusion were observed in the CBS group.

4.4.1.4 CBS sample with praxis scores available

WAB praxis data were available on 87.1% (27/31) of the CBS patients. Severe dementia was the reason for three CBS patients being unable to complete the WAB praxis task; MMSE scores were 10 or less in these patients and they were unable to complete any other neuropsychological tests as a result. One patient with an MMSE score of 24/30 and a DRS score of 99/144 was unable to complete the WAB and most other tests due to poor effort secondary to severe apathy

107

(apathy score on the Neuropsychiatric Inventory = 8) [Cummings 1997]. The body side most affected by CBS symptoms and signs was right in 51.9% (14/27) of CBS patients and left in

48.1% (13/27; Table 3). Mean MMSE and total DRS scores were 23.2 ± 1.0 and 114.1 ± 1.2

(DRS cut-off for dementia in this age group = 123/144), respectively, suggesting that the CBS patients were on average only mildly demented.

The mean WAB praxis scale total score of the CBS sample was 53.2 ± 1.6. There were no statistically significant differences in mean WAB praxis scale total scores between those presenting with their right side of the body most affected compared to those with the left side most affected. Based on a normal control group matched for age and education obtained through our longitudinal study, scores of greater than 57.1 are considered in the normal range (between 0 and -1.5 standard deviations [SD]), whereas scores of between 57.1 and 56.1 are considered borderline apraxic (between -1.5 to -2 SD). Scores of less than or equal to 56.1 are considered in the apraxic range; that is, -2 SD and below. Based on these cut-offs, 29.6% (8/27) of CBS patients had no apraxia at the time of their initial investigation; 18.5% (5/27) had borderline apraxia, while more than half (51.9%; 14/27) had clear IMA of varying severity. Importantly, all patients eventually developed apraxia as their disease progressed over the longitudinal observation period.

108

Demographic variable CBS-APX (n=14) CBS-nAPX (n=13) Summary (n=27) Gender 11 (78.6%) F 5 (38.5%) F 16 (59.3%) F 3 (21.4%) M 8 (61.5%) M 11 (40.7%) M Handedness 13 (92.9%) R 12 (92.3%) R 25 (92.6%) R 1 (7.1%)L 1 (7.7%) L 2 (7.4%) L Site of recruitment 9 (64.3%) Cog 8 (61.5%) Cog 17 (63.0%) Cog 5 (35.7%) MD 5 (38.5%) MD 10 (37.0%) MD Dementia vs. motor 10 (71.4%) Dem 8 (61.5%) Dem 18 (66.7%) Dem onset 4 (28.6%) Motor 5 (38.5%) Motor 9 (33.3%) Motor Age of Onset 68.4  2.3 64.7  2.2 66.6 ± 1.6 (mean  SEM years) Age at Investigation 71.7  2.3 68.1  2.2 70 ± 1.6 (mean  SEM years) Duration of symptoms 3.4  0.6 3.4  0.5 3.4 ± 0.4 (mean  SEM years) Years of Education 12.1  0.5 12.8  0.1 12.4 ± 0.6 (mean  SEM years) Body side most affected 7 (50.0%) R 7 (53.8%) R 14 (51.9%) R 7 (50.0%) L 6 (46.2%) L 13 (48.1%) L Table 3. Demographic features of CBS presenting with apraxia (CBS-APX) vs. those without significant apraxia (CBS-nAPX). F = Female; M = Male; Cog = Cognitive Neurology Clinic; MD = Movement Disorders Clinic; R = Right; L = Left; Dem = Dementia onset; Motor = Motor onset

4.4.1.5 Comparison of apraxic to borderline/non-apraxic CBS patients: Neuropsychological and

SPECT analysis

The CBS group was stratified into 1) those with apraxia and 2) those with borderline/no apraxia.

There were no significant differences in demographic features between CBS patients with apraxia and those with borderline or no apraxia (Table 3). There were also no significant differences in any of the clinical features between the two groups (data not shown). Table 4 compares neuropsychological, neuropsychiatric and functional measures between the two groups. Mean MMSE scores were slightly lower in the apraxic vs. borderline/non-apraxic group

(21.2 ± 1.6 vs. 25.3 ± 1.1, respectively; t(25) = -2.1, p = 0.04). However, no significant differences were observed on the mean DRS scores indicating that the groups did not differ

109 significantly in terms of dementia severity. Mean scores on the delayed visual reproduction were also worse in the apraxic vs. borderline/non-apraxic group (5.3 ± 1.9 vs. 14.4 ± 3.4, respectively; t(15) = -2.4, p = 0.03 adjusted for unequal variances). A similar finding was observed for the mean scores on the immediate visual reproduction task (apraxic: 12.9 ± 3.0 vs. non-apraxic: 22.6

± 3.0; t(17) = -2.2, p = 0.04). There was a good correlation noted between WAB praxis and immediate visual reproduction scores (Pearson r=0.50, p=0.03) suggesting that the degree of apraxia may account for some of the variance in this relationship. However, there was no correlation observed between scores on the delayed visual reproduction and the WAB praxis scale (Pearson r=0.32, p=0.17) suggesting that this association was mostly independent of degree of motor impairment. Another significant difference between the apraxic and non-apraxic groups was in the Benton Judgement of Line Orientation test (7.3 ± 2.9 vs. 18.6 ± 3.2, respectively; t(19) = -2.6, p = 0.02) suggesting more right parieto-occipital involvement in the group with apraxia.

110

Psychometric Measures CBS-APX (n) CBS-nAPX (n) General cognition MMSE /30 [n=27]* 21.2 ± 1.6 (14) 25.3 ± 1.1 (13) Clock total /10 [n=8] 6.3 ± 1.5 (4) 7.5 ± 1.3 (4) MDRS /144 [n=25] 105.9 ± 6.0 (12) 121.6 ± 5.2 (13) NART /127.8 [n=19] 108.4 ± 2.6 (8) 106.8 ± 2.9 (11) Raven‟s Progressive Matrices [n=22] 19.6 ± 1.3 (10) 23.8 ± 2.6 (12) Memory CVLT Long Delay Free Recall /16 [n=21] 6.2 ± 0.8 (9) 6.8 ± 0.9 (12) Delayed Visual Reproduction /41 [n=19]* 5.3 ± 1.9 (8) 14.4 ± 3.4 (11) Language WAB total /100 [n=23]* 81.1 ± 3.8 (13) 91.8 ± 1.9 (10) WAB content /10 7.7 ± 0.5 8.8 ± 0.3 WAB fluency /10 7.8 ± 0.6 9.1 ± 0.2 WAB comprehension /10* 8.7 ± 0.3 9.8 ± 0.1 WAB repetition /10 8.4 ± 0.4 9.4 ± 0.2 WAB naming /10* 8.0 ± 0.3 8.9 ± 0.3 Boston Naming /30 [n=22] 23.8 ± 1.6 (10) 24.2 ± 1.4 (12) Semantic Fluency /20 [n=26]* 7.4 ± 1.0 (14) 12.6 ± 2.2 (12) Praxis WAB praxis /60 [n=27] ¥ 48.7 ± 2.7 (14) 58.0 ± 0.3 (13) Attention & working memory Digit span - forward /12 [n=23] 7.1 ± 0.9 (11) 7.0 ± 0.8 (12) Digit span - backward /12 [n=23] 4.6 ± 1.0 (11) 4.5 ± 0.8 (12) Visuospatial abilities Rey Osterieth Complex Figure – Copy /36 [n=20] 12.6 ± 4.1 (9) 20.5 ± 4.1 (11) Benton Line Orientation /30 [n=21]* 7.3 ± 2.9 (10) 18.6 ± 3.2 (11) Executive functions Phonemic fluency (FAS) [n=21] 14.2 ± 2.3 (10) 23.6 ± 4.4 (11) Trail Making Test A (time in seconds) [n=19] 133.4 ± 27.2 (8) 90.8 ± 14.7 (11) Trail Making Test B (time in seconds) [n=13] 206.0 ± 43.4 (4) 218.2 ± 55.6 (9) WCST categories /6 [n=22] 1.6 ± 0.4 (9) 2.0 ± 0.4 (13) WCST perseverative errors [n=22] 16.9 ± 6.3 (9) 8.2 ± 1.9 (13) Neuropsychiatric features Neuropsychiatric Inventory – Total /144 [n=25] 11.8 ± 3.2 (12) 7.9 ± 2.3 (13) Cornell Depression Scale (%) [n=26] 25.6 ± 4.7 (13) 19.0 ± 2.9 (13) Functional measures Disability Assessment for Dementia (DAD; %) [n=26] 70.4 ± 7.7 (13) 79.4 ± 6.8 (13) DAD-Activities of Daily Living (%) 81.1 ± 8.0 86.8 ± 6.3 DAD-Instrumental Activities of Daily Living (%) 65.2 ± 9.3 73.9 ± 8.1 Table 4. Mean scores (± SEM) on neuropsychological, neuropsychiatric and functional measures in CBS presenting with apraxia (CBS-APX) vs. those without significant apraxia (CBS-nAPX). The number of patients (n) tested is listed next to individual measures. Missing data is secondary to the inability of the patient/caregiver to complete the test. MMSE = Folstein‟s Mini-Mental State Exam; NART = National Adult Reading Test; MDRS = Mattis Dementia Rating Scale; CVLT = California Verbal Learning Test; WAB = Western

111

Aphasia Battery; FAS = F-, A-, and S-phonemic fluency; WCST = Wisconsin Card Sort Test. Independent samples t-tests were used to compare MMSE, NART, Clock, MDRS, Boston naming, semantic fluency, visual reproduction, forward and backward digit span, CVLT, Benton, Trails A and B mean scores between groups. Mann Whitney U test was used to compare scores on FAS, WAB, Rey and WCST between groups. *p≤0.05; ≦≤0.005

The Western Aphasia Battery (WAB) praxis scores were correlated with WAB total scores

(Spearman‟s rho = 0.52, p<0.01). Furthermore, WAB total scores were significantly lower in the apraxic compared to the non-apraxic group (81.1 ± 3.8 vs. 91.8 ± 1.9, respectively; Mann-

Whitney U test, p = 0.05; Table 4). This may account for the reduced MMSE scores in the apraxic group since MMSE is heavily weighted towards language function. In support of this, a strong correlation was observed between the WAB total and MMSE scores (Spearman‟s rho =

0.77, p<0.0005). Although all WAB subscores tended to be lower in the apraxic group, the comprehension and naming subscores were significantly worse (Table 4). Figure 2 demonstrates that CBS patients with apraxia tended to have more severe aphasic disturbances than those without apraxia consistent with the mean WAB total score differences between the groups. Mean semantic fluency scores were also lower in the apraxic vs. non-apraxic group (7.4 ± 1.0 vs. 12.6

± 2.2, respectively; t(16) = -2.2, p = 0.05; Table 4). With respect to the SPECT perfusion data, no significant differences were observed between the apraxic and non-apraxic groups, after correcting for multiple comparisons.

112

Figure 2. Frequency of different aphasia categories on the Western Aphasia Battery (WAB) distributed according to the CBS group with apraxia versus those with borderline/no apraxia.

4.4.1.6 Perfusion versus ideomotor apraxia

The SPECT scans were, on average, acquired within 3.9 ± 1.4 weeks of the neuropsychological assessment including the WAB praxis measurement. Severity of IMA was positively correlated with perfusion in the left inferior parietal lobule, including the left angular gyrus (i.e., WAB praxis scores decrease as perfusion decreases). This was seen on the FWE- and FDR-corrected maps shown in Figure 3 (see Table 5 for details). There were no negative correlations or areas of relative hyperperfusion observed in association with the praxis measure. The use of the „body side most affected‟ covariate did not significantly change the results.

113

Anatomical locus Talairach No. of SPM t-score ROI (Brodmann area) Coordinates voxels (p-value) x y z Parietal region – SPM (FWE-corr) Left angular gyrus (39) -42 -70 31 46 5.7 (p=0.02) Yes Left inferior parietal lobule (40) -50 -52 47 59 5.7 (p=0.03) Yes

Parietal region – SPM (FDR-corr) Left inferior parietal lobule (39) -44 -64 38 632 5.3 (p=0.01) Yes Table 5. Areas of hypoperfusion on SPECT in the CBS group that correlate with WAB praxis scores in the regression analyses. „Body side most affected‟ was included as a covariate in both Statistical Parametric Mapping (SPM) and Region of Interest (ROI) analyses. All p-values were corrected for multiple testing in SPM analysis using Family Wise Error (FWE-corr) and False Discovery Rate (FDR-corr), and in ROI analysis within a general linear model. Column denoted ROI refers to overlapping regions of decreased perfusion between the SPM and ROI analyses.

114

Figure 3. Statistical parametric map of surface regions of the brain showing decreased perfusion in the left inferior parietal region, including the angular gyrus, that correlate with WAB praxis scores in the regression analyses. These have been overlaid on the „Collin‟ brain MRI template. Red areas are corrected for multiple testing using the False Discovery Rate, while green areas are corrected using the more conservative Family Wise Error method. Refer to text and Table 5 for details of analysis and results

ROI analysis was also performed to serve as an independent confirmation of the SPM result.

Pearson correlation analysis revealed a significant positive correlation between perfusion in the left inferior parietal ROI (comprised of left angular and supramarginal gyri) and WAB praxis scores (r = 0.64, p < 0.001). To gauge the independent contribution of perfusion within the left inferior parietal region in predicting WAB praxis scores and to control for the potential bias of symptom lateralization in CBS, we conducted a hierarchical regression analysis with stepwise variable entry. In this model, perfusion in the left inferior parietal region and „most affected side of body‟ served as the independent predictors. WAB praxis scores represented the dependent variable. These results were similar to the SPM analysis. Specifically, reduced perfusion in the left inferior parietal region significantly predicted reduced performance on the WAB praxis scale accounting for 42% of the variance in the relationship (F[1, 25] = 17.7, R2 = 0.42, p < 0.001).

„Body side most affected‟ did not enter into the overall model as significant.

To further explore the relationship between perfusion and praxis performance, separate SPM regression analyses were conducted using individual subscores on the WAB praxis scale. No individual WAB praxis subscores, including intransitive, transitive, bucco-facial, or complex gestures, correlated with hypoperfusion or hyperperfusion after correction for multiple testing. In an exploratory analysis set out to better delineate neural components of the praxis network, the stringency of the SPM analyses was subsequently reduced by setting the threshold voxel level p- value to <0.001, uncorrected. Intransitive gesture scores demonstrated correlation with reduced perfusion in more posterior regions including the left inferior occipital, fusiform, and lingual gyri, as well as the left and right superior parietal lobules. Transitive gesture scores were associated with reduced perfusion in angular and supramarginal gyri, the superior parietal lobule, the precentral and postcentral gyri, and the inferior occipital gyrus all located only within the left hemisphere. Complex gesture scores were correlated with the same left posterior regions as in the transitive gesture regression; however, more left anterior regions demonstrated reduced perfusion including the superior and middle frontal gyri. Finally, bucco-facial gestures were correlated with hypoperfusion in the left inferior and middle frontal gyri as well as the left precentral gyrus.

4.4.1.7 Post-hoc atrophy analysis

The average tissue type in the FWE-corrected SPM mask was found to be: 50% in white matter,

37% in grey matter, and 13% in CSF. Similar results were obtained with the FDR-corrected SPM mask (white matter: 51%; grey matter: 35%; and CSF: 14%). Please refer to Supplementary

Figure 1.

4.5 Discussion

To our knowledge, this is the first brain SPECT study demonstrating that perfusion in the left inferior parietal lobule is significantly correlated with severity of IMA in CBS. This result was identified using a whole brain voxel-by-voxel SPM regression analysis that accounted for multiple comparisons and was corroborated by a region of interest linear regression analysis.

Left inferior parietal atrophy, that is, the effect of partial volume averaging, is unlikely to be a major contributor to this result based on the estimate that approximately 85% of the tissue underlying the hypoperfused region was classified as brain parenchyma.

117

Several models of apraxia have emerged in the literature based on original case studies and series, and the majority of these implicate a role of the left parietal lobe [Geschwind

1975;Goldenberg 2009;Heilman and Rothi 1993;Liepmann 1920;Roy 1996]. Several animal and human studies have attempted to identify the underlying neural substrates of IMA.

Neuroanatomical and electrophysiological studies in monkeys demonstrate the importance of the parietofrontal circuit in transforming visual and tactile sensory information into knowledge for limb movements [Leiguarda and Marsden 2000]. A functional MRI study of pantomiming use of tools in healthy adults implicated the dominant, left intraparietal and dorsolateral frontal cortices suggesting that these regions may be important in determining ideomotor praxis [Moll et al.

2000]. Lesional studies also confirm the role of the left hemisphere in apraxia, in particular the inferior parietal and premotor/supplementary motor areas [Gross and Grossman 2008;Leiguarda and Marsden 2000;Stamenova et al. 2009]. The majority of these studies have included patients with strokes and CBS [Buxbaum et al. 2007;Goldenberg and Spatt 2009;Jacobs et al.

1999;Kertesz and Ferro 1984].

The main types of limb apraxia identified in CBS are IMA, limb-kinetic apraxia and, less often, conceptual/ideational apraxia [Gross and Grossman 2008;Stamenova et al. 2009;Zadikoff and

Lang 2005]. Limb-kinetic apraxia (LKA; loss of hand and finger dexterity resulting in a breakdown and awkwardness of distal movements) [Kleist 1907] is thought to reflect sensory- motor control dysfunction [Liepmann 1920]. In CBS and in one study of pathologically-proven corticobasal degeneration (CBD), it has been associated with involvement of the ventral

118 premotor cortex bilaterally, although worse on the side contralateral to the LKA deficit

[Leiguarda et al. 2003;Tsuchiya et al. 1997;Zadikoff and Lang 2005]. Conceptual/ideational apraxia, defined in this paper as impairment in object/tool or action knowledge, has been less well studied in CBS [Stamenova et al. 2009]. We summarized data across nine studies that looked for both ideomotor and conceptual/ideational apraxia in CBS, and approximately 27%

(30/112) of CBS patients demonstrated a conceptual deficit [Chainay and Humphreys

2003;Graham et al. 1999;Jacobs et al. 1999;Kertesz et al. 2000b;Leiguarda et al. 1994;Pillon et al. 1995;Soliveri et al. 2005;Spatt et al. 2002;Stamenova et al. 2011]. There was a high degree of variability in the occurrence of conceptual/ideational apraxia with several studies demonstrating no conceptual deficit [Chainay and Humphreys 2003;Graham et al. 1999;Jacobs et al.

1999;Pillon et al. 1995;Soliveri et al. 2005;Stamenova et al. 2011], while three studies demonstrated frequencies ranging from 30% to 60% [Kertesz et al. 2000b;Leiguarda et al.

1994;Spatt et al. 2002]. It is likely that these discrepancies across the studies occurred as a result of differences in definition of this apraxia type, diagnostic heterogeneity, and/or methodological differences in the assessments utilized. In terms of anatomical localization, conceptual knowledge of tool use and action has been suggested to reside in the left inferior parietal lobule

[Heilman et al. 1982], and this was later shown to be restricted to mechanical knowledge

[Ochipa et al. 1992]. In contrast, semantic knowledge on the prototypical use of tools has been shown to localize to the left temporal lobe [Hodges et al. 1999].

More recently, two studies have directly correlated structural changes on MRI to standardized measures of ideomotor praxis in CBS [Borroni et al. 2008b;Huey et al. 2009b]. In 20 patients with CBS, the first of these studies demonstrated a significant positive correlation between total

119 score on the de Renzi test of praxis and grey matter density in the bilateral parietal operculum

[Borroni et al. 2008b]. Using a pre-specified hypothesis, they also found that total scores positively correlated with fractional anisotropy in the left dorsolateral parietofrontal associative fibers on diffusion tensor imaging [Borroni et al. 2008b]. A smaller study of 16 patients with progressive non-fluent aphasia including three with CBS found that limb apraxia as assessed by the Apraxia Battery for Adults-2 (ABA-2) correlated with loss of gray matter volume in the left inferior parietal lobe [Rohrer et al. 2010b]. Notwithstanding important differences between these studies and our current one (e.g., imaging modalities, praxis assessment tools, and diagnostic heterogeneity), consistent findings are that the left hemisphere is invariably involved in IMA and that the majority of studies identify the dominant inferior parietal lobule as an important neuroanatomical correlate. From our structural MRI analysis, approximately 50% of the tissue underlying the hypoperfused left inferior parietal region was white matter while approximately

35% was grey matter. These findings support those of Borroni et al. [Borroni et al. 2008b] suggesting the importance of underlying white matter disease (either perfusion abnormalities or loss of white matter tract integrity) as a potential contributor to IMA in CBS. The current finding is also consistent with contemporary theories of apraxia previously described as well as a correlational study using the subtraction method of lesion overlap in stroke, in which the critical area of overlap in apraxic compared to non-apraxic patients was in the centrum semiovale deep to the parietal cortex including the long association tracts, such as the superior longitudinal and frontal occipital fasciculi [Roy et al. 1998].

Given that a deficit in tool or action knowledge (i.e., ideational/conceptual apraxia) is an uncommon finding in CBS, why do our results demonstrate such a strong association between

120 hypoperfusion in the left inferior parietal lobule where the purported „praxicons‟ are thought to reside, and IMA as assessed by the WAB praxis scale? The reason, in part, may be the result of what the WAB praxis assessment tool is actually evaluating. Although the total score on the

WAB praxis scale best represents severity of IMA, the score in patients with more severe impairment is partly accounted for by failure of actual tool use, which reflects a conceptual deficit. The definition of conceptual/ideational apraxia has also been variable from study to study resulting in some degree of phenomenological/taxonomic confusion. Some studies distinguish between ideational apraxia defined as a failure to sequence tasks related to tool use correctly and

“conceptual apraxia” defined (as in this paper) as a loss of knowledge relating to tool and action use. Given the common feature of tool use across different assessments of apraxia (ideomotor, conceptual and ideational), it is likely that there will be some degree of overlap in the brain regions most correlated with deficits across the studies.

An alternative way of putting our finding into context is to explain the association of left inferior parietal lobule hypoperfusion with IMA in CBS as being related to the dysfunction of a larger circuit or network that is involved in determining both simple and more complex gestural movements. Indeed, when we reduce the stringency of our analysis and examine different subcomponents of the WAB praxis assessment, a larger network emerges. For example, performance of transitive gestures correlate with hypoperfusion predominantly in the entire left parietal lobe (inferior and superior divisions) as well as within the left sensorimotor cortex. In examining complex gestures, the same regions are implicated; however, hypoperfusion extends into the left premotor and supplementary motor areas as well. These results suggest that performance of transitive and more complex gestures is more strongly linked to left hemispheric

121 function [Kroliczak and Frey 2009]. Based on our hypoperfusion results in this CBS sample, it is plausible that as gestures performed by healthy individuals increase in complexity more of the left parietofrontal network is recruited to carry out the task. Although the reduced stringency of this SPECT-WAB praxis subscore regression analysis increases the chances that the more extensive area shown is a false positive result, prior studies have shown that it is acceptable to use an uncorrected p-value of <0.005 in correlational analyses with SPM in a sample of this size

[Desgranges et al. 1998;Kas et al. 2011].

A large voxel-based morphometry study of 48 patients with CBS demonstrated that gray matter volume loss in the left middle frontal and precentral gyri as well as the left caudate nucleus correlated with reduced performance on the Test of Oral and Limb Apraxia (TOLA) [Huey et al.

2009b]. To some extent, these data are at odds with the results of the current study and we propose as one possible explanation for this discrepancy that atrophy on MRI in the left frontal- subcortical grey matter and hypoperfusion in the left inferior parietal lobule may each account for unique variance associated with IMA severity in CBS. We did not look at volumetric correlations and Huey et al. [Huey et al. 2009b] did not examine perfusion/metabolism so further studies will be necessary to confirm this hypothesis. Another possible explanation for these discrepant findings is diagnostic heterogeneity between the two studies. A prior study has shown that patients with posterior lesions or fluent aphasia have a more severe form of apraxia – including both ideomotor and conceptual/ideational types – than patients presenting with more anterior lesions or non-fluent aphasia [Heilman et al. 1982]. Since about 67% of our sample presented with early cognitive problems with the apraxic group having lower scores on the WAB and more severe forms of aphasia compared to the borderline/non-apraxic group, then this might

122 be one possible explanation for the discrepant findings. Although Huey et al. [Huey et al. 2009b] did not specify the proportion of their sample presenting with early cognitive symptoms or aphasia, scores on the Mattis DRS were comparable (Huey et al.: 116/144 vs. current: 114/144) indicating a similar degree of global cognitive impairment in the two patient samples. Huey et al.

[Huey et al. 2009b] did not include any specific assessments of aphasia so we were not able to compare the samples in this regard. The most affected side of the body was also similar between samples as well (Huey et al.: 46% right-sided vs. current: 52% right-sided).

We further subdivided our sample into those with and without significant apraxia based on a -2

SD cut-off from controls (i.e. WAB praxis score ≤ 56.1). Our WAB praxis cut-off score was higher than that of ≤ 49.7 used in one of the earliest studies of IMA in stroke patients [Kertesz and Ferro 1984]. The control group in this original study was derived from non-brain injured hospitalized patients, whereas our control group were healthy, elderly volunteers who were living in the community. It is conceivable that the „non-brain‟ medical conditions or drug therapies of the control group of Kertesz & Ferro [Kertesz and Ferro 1984] might have affected their overall cognitive performance on the WAB praxis task.

In our stratified analysis, several interesting observations were made. As expected, given that both language and praxis are most often lateralized to the dominant left hemisphere and are represented in overlapping neuroanatomical networks, apraxic patients demonstrated significantly lower WAB total scores and had more severe forms of aphasia than the borderline/non-apraxic group. Comprehension difficulties and anomia were significantly more

123 pronounced in the apraxic group in addition to an observed reduction in semantic fluency. This suggests more left temporal lobe involvement in the apraxic group. Visuospatial orientation and attention were also significantly worse in the apraxic group as evidenced by lower Benton judgement of line orientation scores indicating more prominent right parieto-occipital dysfunction. We have previously shown that performance on the Benton line orientation task correlated with reduced perfusion of the right parietal lobe in Alzheimer‟s disease of varying severity [Tippett and Black 2008]. The parietal lobes are prominently affected by the underlying pathology in CBS [Wadia and Lang 2007] and although the neurodegenerative process usually starts asymmetrically, it progresses relentlessly to involve bilateral structures. Indeed in the current study, there was a significant correlation between performance on the WAB praxis and

Benton line orientation task with even stronger correlations observed between perfusion of left and right parietal regions (data not shown) supporting our finding.

An unexpected finding was that the apraxic group demonstrated lower scores on the Wechsler

Memory Scale-Revised (WMS-R) delayed visual reproduction task than the non-apraxic group.

This finding could not be accounted for by the severity of the apraxia alone. In humans and in monkeys, two pathways have been identified for the processing of visual information: the occipitotemporal pathway or ventral stream and the occipitoparietal pathway or dorsal stream

[Ungerleider et al. 1998]. The ventral visual stream is important for object vision including characteristics such as pattern, shape and colour, while the dorsal visual stream is important for spatial perception (e.g., judging distance and orientation of objects relative to each other) and also is involved in visually guided reaching [Goodale and Milner 1992;Ungerleider et al. 1998].

The visual reproduction task asks subjects to examine four drawings of several geometric figures

124 oriented in space in relation to each other, each for ten seconds. After a ten second (immediate visual reproduction) and 30 minute delay (delayed visual reproduction) they are asked to draw the four pictures from memory. Scoring is based on the ability to accurately recall the shapes and patterns as well as the distance and orientation in relation to each other thereby taxing both the ventral and dorsal visual streams, respectively [Lezak 1983;Wechsler 1987]. A BOLD fMRI study in healthy volunteers demonstrated that activity in the posterior parietal cortex bilaterally is strongly correlated with the capacity limit to store visual information (i.e., visual short-term memory) [Todd and Marois 2004]. We hypothesize that the poor performance of the apraxic group on delayed visual reproduction may be due to involvement of the posterior parietal cortex within the dorsal visual stream, a network which also overlaps with the frontoparietal praxis system.

In the SPECT analyses of all CBS cases versus controls, there was reduced perfusion noted in bilateral dorsolateral and medial frontal/prefrontal regions, as well as bilateral parietal regions in the CBS group. Additionally, reduced perfusion was also evident to a lesser degree in right temporal regions and insula. Our results confirm in a larger sample, previous SPECT studies of

CBS demonstrating reduced perfusion in frontoparietotemporal regions [Hossain et al.

2003;Koyama et al. 2007;Markus et al. 1995;Okuda et al. 1999;Okuda et al. 2000b;Zhang et al.

2001]. We further investigated whether perfusion reductions tended to be lateralized opposite to the most affected side of the body by comparing CBS-L and CBS-R to their respective control groups. In CBS-L, the regions of hypoperfusion localized to the right hemisphere including dorsolateral prefrontal cortex, primary somatosensory cortex, superior parietal lobe and some temporal regions. In contrast, the CBS-R group demonstrated reduced perfusion only in the left

125 inferior frontal gyrus in the ROI analysis. However, when the stringency of the SPM analysis was lowered, hypoperfusion in the left frontoparietal region was seen in the CBS-R group.

Overall, then, our study confirms a lateralization of perfusion defects contralateral to the most affected side of the body in CBS.

Strengths of the present study include ascertainment of CBS cases from both cognitive and movement disorders clinic, use of standardized neuropsychological assessments including a language battery, use of brain SPECT perfusion that attempted to account for effects of underlying atrophy on MRI, and the combined approach of an unbiased, whole brain voxel-by- voxel analysis followed by confirmation using a more robust region of interest method. Although the CBS sample size was relatively large considering the rarity of this diagnosis, from a statistical perspective it is indeed a small sample. Although we did not have pathological confirmation of CBD diagnosis on the entire sample, 25% of the sample came to autopsy with pathological confirmation of CBD in 63% of cases (5/8 cases; unpublished data); this rate of diagnostic accuracy is similar to prior studies [Wadia and Lang 2007]. Other pathologies included PSP (12.5%; 1/8 cases), FTLD-U/TDP43 proteinopathy (12.5%; 1/8 cases), and combined dementia with agyrophilic grains, CBD and cerebral amyloid angiopathy

(AGD/CBD/CAA; 12.5%; 1/8 cases). Other limitations include a cross-sectional design and assessment of predominantly IMA. Furthermore, we only had volumetric MRI data on 21 of the

27 patients who had WAB and SPECT completed. Therefore, we were unable to estimate the degree of atrophy underlying the SPM mask in these six patients. However, qualitative visual examination of their MRI data did not reveal any tendency for the apraxic subgroup to have more left parietal atrophy than the non-apraxic subgroup. Another limitation is that the WAB praxis

126 measure does not provide a full picture of the nature of the disruption to limb praxis since it confounds pantomime, imitation and tool use. The same limitation also applies to the ABA-2 and de Renzi apraxia tests used in the other studies [Borroni et al. 2008b;Rohrer et al. 2010b].

This study suggests that severity of left inferior parietal lobule hypoperfusion corresponds to

IMA as it becomes more severely affected in CBS supporting a central role for this structure in the dominant hemisphere frontoparietal praxis network. Dysfunction in language, visuospatial and visual memory performance is more frequent in CBS patients with apraxia due to involvement of overlapping brain networks that subserve these related higher cognitive processes. Future work will involve use of a comprehensive assessment of apraxia using a conceptual model [Stamenova et al. 2011] together with SPECT and MRI imaging modalities in order to better identify the neuroanatomical correlates of the different apraxia types.

4.6 Acknowledgements

This work was supported by an operating grant from the Canadian Institutes of Health Research

[MT13129 to S.E.B.] and a New Investigator Award from the Parkinson Society Canada [2011-

19 to M.M.]. M.M. was supported by a Canadian Institutes of Health Research Clinician

Scientist Award. We thank the patients and their families for the time and effort that they committed to participate in this study.

127

Supplementary Figure 1A. Mean proportion of different MRI tissue classes underlying the FWE-corrected SPM mask. Error bars denote standard deviation.

128

Supplementary Figure 1B. Mean proportion of different MRI tissue classes underlying the FDR-corrected SPM mask. Error bars denote standard deviation.

129

Group Anatomical locus Talairach No. of SPM t-score (Brodmann area) Coordinates voxels (p-value) or x Y z ROI F-score (p-value) CBS-all Frontal regions – SPM vs. Right middle frontal gyrus (6) 48 8 44 10972 5.7 (p=0.003) controls Right superior frontal gyrus (10) 30 56 23 184 4.2 (p=0.005) Right inferior frontal gyrus (47) 44 17 -3 225 4.0 (p=0.006) Left superior frontal gyrus (6) -18 9 68 10972 5.3 (p=0.003) Left superior frontal gyrus (8) -8 43 48 48 3.6 (p=0.008) Left precentral gyrus (44) -52 12 3 100 3.7 (p=0.007)

Frontal regions – ROI Left middle frontal gyrus - dorsal - - - 1740 3.5 (p=0.04) Left inferior frontal gyrus - dorsal - - - 1105 8.0 (p=0.001) Left anterior cingulate - middle - - - 623 4.8 (p=0.01) Right precentral gyrus - - - 2723 3.4 (p=0.04) Right inferior frontal gyrus - dorsal - - - 1128 5.8 (p=0.005)

Parietal regions – SPM Left superior parietal lobule (7) -32 -55 60 10972 5.0 (p=0.003) Right postcentral gyrus (2) 50 -25 42 1246 4.6 (p=0.004) Right inferior parietal lobule (40) 65 -24 29 1246 4.0 (p=0.006) Right angular gyrus (39) 50 -74 33 57 3.8 (p=0.007)

Parietal regions – ROI Left postcentral gyrus - - - 2675 3.4 (p=0.04) Right superior parietal lobule - - - 2132 3.3 (p=0.05) Right supramarginal gyrus - - - 1295 3.3 (p=0.05)

Limbic regions – SPM Right cingulate (24) 4 -8 41 107 3.9 (p=0.006) 8 8 35 107 3.4 (p=0.01) Left cingulate (24) 0 -12 41 52 3.7 (p=0.007)

Temporal regions – SPM Right middle temporal gyrus (21) 71 -45 2 138 3.8 (p=0.006) Right fusiform gyrus (37) 42 -44 -15 34 3.7 (p=0.007)

Temporal regions – ROI Right superior temporal gyrus lateral - - - 1473 3.5 (p=0.04)

Other regions – ROI Right insula - - - 1973 4.9 (p=0.01)

CBS-L Frontal regions – SPM vs. Right superior frontal gyrus (6) 28 -3 67 2331 5.4 (p=0.008) controls Right middle frontal gyrus (6) 48 8 44 2331 5.7 (p=0.008) Right inferior frontal gyrus (47) 46 17 -1 1300 5.3 (p=0.008) Left superior frontal gyrus (6) -18 4 70 122 4.9 (p=0.008)

Frontal regions - ROI Right superior frontal gyrus - dorsal - - - 1984 3.4 (p=0.05) Right precentral gyrus - - - 2723 7.3 (p=0.003) Left anterior cingulate - middle - - - 623 4.0 (p=0.03)

Parietal regions - SPM Right postcentral gyrus (1) 36 -36 66 2331 5.7 (p=0.008) Right postcentral gyrus (2) 46 -27 40 22 4.2 (p=0.009) Right superior parietal lobule (7) 24 -71 57 192 4.7 (p=0.008)

Temporal regions - SPM Right transverse temporal gyrus (41) 50 -23 12 1300 4.8 (p=0.008) Right superior temporal gyrus (22) 59 -6 4 1300 4.6 (p=0.008)

Temporal regions - ROI Right middle temporal gyrus lateral - - - 1962 4.5 (p=0.02)

Limbic – ROI Right insula - - - 1973 6.0 (p=0.007)

CBS-R Frontal regions – ROI vs. Left inferior frontal gyrus - dorsal - - - 1105 6.2 (p=0.006) controls Supplementary Table 1. Areas of hypoperfusion on SPECT in all CBS patients, CBS with left side of body most affected, and CBS with right side of body most affected relative to controls. CBS-all; n=31 cases and 31 matched, normal controls; refer to Figure 2A CBS-L; n=15 cases and 15 matched, normal controls; refer to Figure 2B CBS-R; n=16 cases and 16 matched, normal controls „Years of education‟ was included as a covariate in both Statistical Parametric Mapping (SPM) and Region of Interest (ROI) analyses. All p-values were corrected for multiple testing in SPM analysis using False Discovery Rate (FDR), or included in ROI analysis within a general linear model multivariate ANCOVA.

131

5.0 Clinical, neuropsychological, MRI and SPECT

characterization of a prospective sample of patients with

corticobasal syndrome

Mario Masellis, Philip L. Francis, Isabelle Guimont, Wendy Galpern, Juan Bilbao, Kie Honjo,

Fuqiang Gao1, Gregory Szilagyi, Farrell Leibovitch, James L. Kennedy, Galit Kleiner-Fisman,

Lisa Ehrlich, Robert Chen, Anthony E. Lang, Sandra E. Black

Mario Masellis clinically assessed several of the patients included in this study, extracted the clinical information, designed the study, performed the data analysis and wrote the manuscript. Philip Francis assisted with the SPM analysis of SPECT data and Kie Honjo assisted with an independent visual read of the MRI data. Wendy R. Galpern, Galit Kleiner-Fisman and Anthony E. Lang assessed and collected clinical data on patients ascertained from a movement disorders clinic. Juan Bilbao performed the neuropathological analysis.

132

5.1 Abstract

Corticobasal syndrome (CBS) is a rare and debilitating syndrome characterized by the unique combination of lateralized cortical and extrapyramidal features that occurs due to a variety of underlying neurodegenerative pathologies. In this paper, we describe the initial neuropsychological, MRI and SPECT imaging profile of a prospective series of 31 consecutive

CBS patients ascertained from a movement disorders and a cognitive neurology clinic. The sample was stratified into CBS presenting with early dementia (CBS-D; n=22) vs. early motor features (CBS-M; n=9), which identified that CBS-M had a higher occurrence of cortical sensory loss than CBS-D (100% vs. 45.5%, respectively; p=0.005). Conversely, the presence of aphasia, as determined by the Western Aphasia Battery, was found to be more common and severe in

CBS-D compared to CBS-M (88.2% vs. 33.3%, respectively; p=0.02). These findings are associated with lateralization of the motor signs to the right side in CBS-D. CBS-D also demonstrated more difficulties with simple attention span and visuospatial orientation/attention on neuropsychological testing. Atrophy patterns on MRI did not distinguish between CBS-D and

CBS-M. However, CBS-M patients had significantly reduced perfusion in the right supplementary and premotor areas compared to CBS-D (p<0.05). A subset of eight patients was followed to autopsy with 7 patients having a tauopathy and 1 patient exhibiting non-tau pathology, specifically, frontotemporal lobar degeneration-ubiquitin/TDP43 proteinopathy

(FTLD-U/TDP43). Atrophy and white matter changes on MRI correlated with the burden of underlying brain pathology. This study emphasizes the importance of performing detailed clinical and multimodal phenotyping to characterize heterogeneity in CBS. It also provides new insights into the neuropsychological and neuroimaging correlates of lateralized brain dysfunction in the syndrome.

133

5.2 Introduction

The first description of corticobasal syndrome (CBS) was in 1967 by Rebeiz and colleagues who later characterized three cases of this syndrome from the clinical and pathological perspective

[Rebeiz et al. 1967;Rebeiz et al. 1968]. The scientific literature on this topic was sparse until the late 1980s and early 1990s during which there were several case series published characterizing the unique clinical features of CBS [Gibb et al. 1989;Mahapatra et al. 2004;Riley et al.

1990;Rinne et al. 1994]. Since then, much has been learned about the clinical, neuroimaging, genetic and pathological heterogeneity of this enigmatic disorder.

The clinical diagnosis of CBS is made based on an insidious onset and progressive neurological decline including at least one cortical (e.g., apraxia, non-fluent aphasia/apraxia of speech, cortical sensory loss, myoclonus, alien limb phenomenon) and one extrapyramidal feature (e.g., rigidity, dystonia), which is not attributable to any other identifiable cause of brain dysfunction

[Boeve et al. 2003]. However, there have been no formally accepted, consensus clinical diagnostic criteria [Mahapatra et al. 2004]. There are two main early clinical presentations of

CBS. The first is the “classical” perceptuo-motor disorder without early dementia, which often presents to movement disorders clinics. The second subtype presents with an early dementia occurring along the spectrum of frontotemporal lobar degeneration (FTLD), most commonly the behavioural variant of frontotemporal dementia (bvFTD) or progressive non-fluent aphasia

(PNFA). This subtype is most likely to present first to dementia clinics. There is evidence suggesting that early dementia is the more frequent initial presentation of CBS [Bergeron et al.

1998;Grimes et al. 1999b;Mathuranath et al. 2000] yet, because the initial symptoms may be

134 non-specific, the movement disorder presentation is easier to recognize. This may have created a referral bias in CBS research particularly in many of the early studies, which ascertained patients predominantly from movement disorders clinics. Several studies have overcome this bias by examining patients with both types of presentations [Josephs et al. 2008;Kertesz et al.

2000b;Kertesz et al. 2005;McMonagle et al. 2006;Murray et al. 2007;Riley et al. 1990]. Few studies, however, have directly compared CBS patients presenting with early motor vs. early dementia features [Josephs et al. 2008;Kertesz et al. 2000b;McMonagle et al. 2006], and, to our knowledge, no studies have investigated whether perfusion SPECT can help differentiate between these two subtypes of CBS.

Both structural and functional neuroimaging studies may support a diagnosis of CBS. Early MRI studies have demonstrated asymmetrical cortical atrophy in frontoparietal regions and, frequently, subcortical white matter T2/FLAIR hyperintensities contralateral to the most affected side of the body [Riley et al. 1990;Savoiardo et al. 2000;Soliveri et al. 1999;Tokumaru et al.

1996;Winkelmann et al. 1999]. These initial findings have been confirmed by more recent MRI studies in larger patient cohorts [Boxer et al. 2006;Groschel et al. 2004;Grossman et al.

2004;Josephs et al. 2008;Koyama et al. 2007;Taki et al. 2004;Yekhlef et al. 2003]. Sawle et al.

[Sawle et al. 1991] using PET to measure regional cerebral oxygen metabolism were the first to demonstrate that patients with CBS have hypometabolism predominantly in the posterior and superior temporal, inferior parietal, and occipital (association) cortices; frontal association regions also demonstrated reduced metabolism although they did not achieve statistical significance. This pattern of hypometabolism tended to be asymmetric, being more prominent contralateral to the most affected side of the body. This frontoparietotemporal pattern of reduced

135 activity has been confirmed by other studies employing 18-fluoro-deoxyglucose (18-FDG)-PET

[Blin et al. 1992;Coulier et al. 2003;Eidelberg et al. 1991;Garraux et al. 2000;Hosaka et al.

2002;Juh et al. 2005;Klaffke et al. 2006;Laureys et al. 1999;Nagahama et al. 1997;Nagasawa et al. 1996;Taniwaki et al. 1998;Yamauchi et al. 1998a] and perfusion tracers (HMPAO, ECD and

IMP) using SPECT [Hossain et al. 2003;Koyama et al. 2007;Markus et al. 1995;Okuda et al.

1999;Okuda et al. 2000b;Zhang et al. 2001]. Several of these studies also demonstrated reduced asymmetric activity in the basal ganglia and thalamus contralateral to the most affected side of the body. Reduced dopamine transporter binding of TRODAT [Lai et al. 2004] and β-CIT

[Pirker et al. 2000;Plotkin et al. 2005] in the basal ganglia has also been demonstrated in CBS.

It is critically important to follow patients longitudinally to ensure that clinical criteria for CBS have been met, as the neurological features of the full syndrome may not be present at onset, but may develop over time. This was eloquently shown in a longitudinal, prospective cohort of patients with initial diagnoses ranging from bvFTD, CBS, Progressive Supranuclear Palsy (PSP) to PNFA, the majority of whom then went on to develop second and/or third syndromes with significant clinical overlap along the FTLD spectrum [Kertesz et al. 2005;McMonagle et al.

2006]. In addition to the clinical heterogeneity in presentation and evolution of CBS, there is also significant pathological heterogeneity [Lee et al. 2011] leading some to propose the term „Pick

Complex‟ to encompass the varying pathologies occurring along this disease spectrum [Kertesz et al. 2000b]. Given this pathological heterogeneity, prospective, longitudinal studies that follow patients with CBS to autopsy are required in order to obtain a more accurate estimate of the neuropathological substrates of this rare syndrome.

136

Our prior study provided the initial characterization of a prospectively recruited sample of CBS cases vs. controls in terms of demographics, clinical, and SPECT imaging features and identified that perfusion within the left inferior parietal lobule correlated with a measure of ideomotor apraxia (chapter 4). The purpose of the current study using this sample was threefold: 1) to describe the initial standardized neuropsychological and neuropsychiatric, and MRI

(qualitatively) profile of a prospective cohort of 31 CBS patients ascertained from either a movement disorders or a cognitive neurology clinic; 2) to compare the clinical, neuropsychological/neuropsychiatric, MRI, and, in particular, SPECT imaging features of CBS patients presenting with early dementia vs. early motor symptoms; and 3) to identify the underlying neuropathological substrates in a subset of this sample who came to autopsy. Novel aspects of this study include the comparison of SPECT perfusion measures in the early motor vs. early dementia subgroups and also the integration of clinical, neuropsychological, MRI, SPECT, and pathological data, whenever possible. This study is also unique in that it used two different techniques to analyze the SPECT data, namely, region of interest analysis and statistical parametric mapping (SPM).

5.3 Methods

5.3.1 Subjects:

31 subjects with a clinical diagnosis of CBS according to proposed diagnostic criteria [Boeve et al. 2003] were recruited consecutively through two academic clinics as previously described

(chapter 4). They were matched to 31 normal healthy controls as closely as possible with respect

137 to age, sex, and years of education. CBS subjects and controls were ascertained and followed as part of the Sunnybrook Dementia Study, a prospective, longitudinal study of dementia and ageing, approved by the local Research Ethics Board. Participants or their substitute decision makers provide written, informed consent to participate in accordance with the Declaration of

Helsinki. All subjects underwent detailed clinical evaluations including: history, general and neurological physical exam, routine laboratory investigations, and standardized behavioural neurology assessment [Darvesh et al. 2005]. The side of greatest rigidity and/or apraxia defined the motor-onset of symptoms. Patients were seen every 6 months for routine clinical follow-up and had yearly prospective, longitudinal assessments which included: standardized measures of neuropsychological performance, neuropsychiatric symptoms and functional status. Structural and functional neuroimaging of the brain with MRI and SPECT were also performed annually.

Additional inclusion criteria were: age between 40 and 90 years, have a knowledgeable caregiver, minimum educational attainment of grade 6 and fluent in English. Their SPECT and neuropsychological evaluations needed to be completed within three months of each other.

Exclusion criteria were: presence of secondary/reversible causes of dementia that were untreated, concomitant neurological or psychiatric illness/substance use and abuse, including clinically relevant depression, history of significant head trauma, early vertical gaze palsy, rest tremor, autonomic disturbances, sustained responsiveness to levodopa, and lesions on neuroimaging suggesting another pathological condition.

138

5.3.2 Neuropsychological, neuropsychiatric and functional measures:

Neuropsychological tests assessing general intelligence and cognition included Folstein‟s Mini-

Mental State Examination (MMSE) [Folstein et al. 1975]; Mattis Dementia Rating Scale (DRS), which ranges from 0 to 144, with lower scores representing more impairment [Mattis 1976];

Clock Drawing Test, which ranges from 0 to 10, with lower scores representing more impairment [Rouleau et al. 1992]; the National Adult Reading Test-Revised (NART-R), which provides a measure of premorbid verbal intelligence [Blair and Spreen 1989], and Raven‟s

Progressive Matrices, which provides a measure of premorbid non-verbal intelligence [Raven

1947]. Tests assessing learning and episodic memory included the California Verbal Learning

Test (CVLT), which assesses verbal memory [Delis et al. 1987], while the visual reproduction subtest of the Wechsler Memory Scale-Revised (WMS-R) assesses visual memory [Lezak 1983].

Measures of language function and naming included: the Boston Naming Test (BNT), which is scored out of 30 with lower scores representing more impairment [Williams et al. 1989]; semantic/categorical fluency [Gladsjo et al. 1999]; and the comprehension subscale of the

Western Aphasia Battery (WAB) [Kertesz and Poole 1974]. Initially, the full WAB was given to all patients, but in the last few years it has only been administered if there is anomia detected on the BNT. The WAB calculates an aphasia quotient based on combined subscores of fluency, content, comprehension, repetition and naming, with a maximum score of 100 and lower scores represent more severe impairment [Kertesz and Poole 1974]. Ideomotor praxis was assessed using the WAB praxis subscale, which is scored out of 60 with lower scores indicating more severe apraxia [Kertesz and Poole 1974]. Attention and working memory was assessed using the

Forward and Backward Digit Span tests from the WMS-R [Lezak 1983;Wechsler 1987]. Several assessments of executive function were employed including: phonemic (F-, A-, and S-word)

139 fluency [Gladsjo et al. 1999;Lezak 1983]; the Trail Making Test A and B (TMT-A and -B) that measure speed of psychomotor processing and mental flexibility [Lezak 1983]; and the

Wisconsin Card Sort Test (WCST) [Heaton 1981]. Visuospatial function was assessed using the

Rey-Osterrieth Complex Figure Test scored out of 36 with lower scores indicating worse visuospatial function [Lezak 1983;Osterrieth 1944;Rey 1941]; and the Benton Line Orientation task, which is motor-free and assesses visuospatial orientation and attention [Lezak 1983].

Behavioural function was investigated using the Neuropsychiatric Inventory (NPI-12), a caregiver-based interview assessing 12 common neuropsychiatric features of dementia; maximum score is out of 144 with lower scores indicating lesser degrees of psychopathology

[Cummings 1997]. Severity of depressive symptoms was assessed using the Cornell Scale for

Depression in Dementia (CSDD); higher scores indicate more severe depressive symptoms

[Alexopoulos et al. 1988]. Functional assessment was performed using the Disability

Assessment for Dementia (DAD), which assesses both basic and instrumental activities of daily living including subcomponents of initiation, planning and performance [Gelinas et al. 1999].

5.3.3 Brain MRI:

140 mm), 2) proton-density and 3) T2-weighted images (interleaved axial spin echo, with TEs 30 and

80 ms, TR 3 s, NEX 0.5, FOV 20 × 20 cm, in-plane resolution 0.781 × 0.781 mm and slice thickness 3 mm). MRI images were qualitatively interpreted by a neurologist (KH) blinded to all clinical, neuropsychological, neuropsychiatric, SPECT, and pathological data. Asymmetry and lobar localization of the maximal atrophy was described. Localization of T2/PD white matter changes was also noted.

5.3.4 Brain SPECT:

SPECT imaging was acquired with a triple-head gamma camera (Prism 3000XP; Phillips

Medical Systems Inc., Cleveland, Ohio) while the patient was resting comfortably and was performed a minimum of 30 minutes and a maximum of 120 minutes after injection of 20 mCi

(740 MBq) of Technetium-99m ethyl cysteinate dimer (99mTc-ECD SPECT). Each view consisted of a 128 × 128 pixel image with a typical reconstructed image resolution of 9.7 mm full width at half maximum. The total imaging time was 19 minutes. Reconstruction was performed by using a ramp-filtered back-projection algorithm followed by a 3-dimensional restoration post-filter (Wiener filter, multiplier 1.0). Ellipses were fit to the approximate location of the outline of the head in each transaxial image, and a calculated attenuation correction applied [Matsuda et al. 1995]. Voxel dimensions were 2.18 × 2.18 × 3.56 mm.

141

5.3.5 Regional perfusion ratios:

Uptake of 99mTc-ECD is approximately proportional to regional cerebral blood flow (rCBF)

[Matsuda et al. 1995] such that brain SPECT can be used to provide semi-quantitative measures of regional brain perfusion. Reconstructed images were co-registered to a SPECT template that was an average of 14 healthy, elderly control scans. A T1-weighted MRI with dimensions similar to the SPECT template was the source of 79 bilateral regions of interest (ROI) as previously described [Lobaugh et al. 2000]. In order to obtain ROI intensity values, we used a common transformation to move from the SPECT template space to MRI space. The cerebellum is frequently used to normalize SPECT counts in studies of dementia [Stamatakis et al. 2001].

However, crossed cerebellar diaschisis may lead to relative differences in perfusion between the left and right cerebellar hemispheres, and, if whole cerebellum is used as the reference region then this may overestimate rCBF in a particular ROI. We, therefore, applied the following rule: if there was more than a 5% difference in counts between left and right cerebellar hemispheres, we use the hemisphere that is more perfused as the reference region. If there is no difference then we use the whole cerebellum as the reference region. In this way, semi-quantitative perfusion ratios are derived and regional Z scores are calculated [Lobaugh et al. 2000].

5.3.6 Pathological analysis:

Neuropathological examination was carried out by one of the authors (J.B.). Paraffin-embedded sections were stained with haematoxylin and eosin, Luxol fast blue (LFB), Bielschowski and

Gallyas. Immunostains using commercial antibodies for tau (Dako, A0024), ubiquitin (Vector

Labs, ZPU576), and α-synuclein (Vector Labs) were performed. Immunostaining with

142 commercial antibodies for TDP43 (ProteinTech Group, Inc.) was performed, when

Frontotemporal Lobar Degeneration-Ubiquitin-positive, Tau-negative pathology (FTLD-U) was demonstrated.

5.3.7 Data analysis:

Statistical analysis of demographic, clinical, neuropsychological, MRI and ROI SPECT variables was performed using the Statistical Package for the Social Sciences (SPSS), version 16.

5.3.7.1 Demographic, clinical and neuropsychological measures:

Categorical demographic and clinical data were analyzed using chi-square or Fisher exact tests.

In comparing neuropsychological test results to the control sample, normalized z-scores were calculated. Normality of continuous demographic and neuropsychological data was determined by examining Q-Q probability plots. Parametric methods (e.g., independent samples t-test) were used if the data fit a normal distribution, otherwise non-parametric tests (e.g., Mann Whitney U test) were performed.

5.3.7.2 Region of interest (ROI) SPECT analysis:

Normality of ROI SPECT data was confirmed as described above. Independent sample t-tests between CBS and control groups were conducted to compare mean perfusion ratios of individual

ROIs within frontal, parietal, temporal, basal ganglia, and thalamic regions, areas previously shown to be affected in CBS. ROIs that were statistically significant on the t-test analysis were included in a multivariate, general linear model (GLM) analysis of covariance (ANCOVA).

143

5.3.7.3 Statistical Parametric Mapping SPECT analysis:

SPECT scans were decompressed, converted to Analyze 7.5 format, and each axial slice was visually inspected for image quality. Statistical Parametric Mapping version 5 (SPM5,

Wellcome Department of Imaging Neuroscience, University College London) was used to pre- process and analyze the scans. The images were spatially normalized to a standard SPECT template in Montreal Neurological Institute (MNI) space [Stamatakis et al. 2001]. This step re- sampled the voxel dimensions to 2 × 2 × 2 mm. The scans were then smoothed using an isotropic Gaussian kernel (12 mm FWHM). A thresholded mean voxel value was chosen for global calculation, and global normalization was achieved by proportional scaling to 50 mL/100 g/min. Voxel-by-voxel t-tests were performed to identify regions with differences in relative cerebral perfusion between groups. We reported significance using a voxel-wise p-value threshold (p < 0.05) corrected for multiple comparisons and an extent threshold of at least 20 contiguous voxels. Our correction methodologies included either controlling the family-wise error (FWE) rate [Worsley et al. 1996] or controlling the false discovery rate (FDR) [Genovese et al. 2002]. A whole brain mask was used to exclude extracranial voxels from the analysis. The maximal peak coordinates of the perfusion differences were converted to Talairach space using the Yale Non-linear MNI to Talairach Converter [Lacadie et al. 2008]

(http://www.bioimagesuite.org/Mni2Tal/index.html). These converted coordinates were translated into anatomical brain regions and Brodmann Areas (BAs) using Talairach Daemon

Client [Lancaster et al. 2000] (http://www.talairach.org/client.html).

144

5.4 Results

5.4.1 CBS cases vs. controls

5.4.1.1 Neuropsychological, behavioural and functional assessment

Figure 1. Normalized (z-) scores of neuropsychological measures in patients with CBS compared to control group. Z-score cut-off of -2.8 corresponds to a p-value of ≤ 0.0026 (Bonferroni-correction for multiple testing). WCST-per = Wisconsin Card Sort Test-perseverative errors (n=22); WCST-cat = Wisconsin Card Sort Test-categories (n=22); TMT = Trail Making Test (TMT-A, n=19; TMT-B, n=13); FAS = F-, A-, S-word phonemic fluency (n=21); Benton Judgement of Line Orientation (n=21); Rey Osterieth Complex Figure Copy (n=20); Digit span B = Backward (23); F = Forward (23); Semantic fluency (n=26); Boston Naming (n=22); WAB = Western Aphasia Battery (n=23); WAB-praxis (n=27); DVR = WMS-III-R delayed visual reproduction (n=19); CVLT = California Verbal Learning Test-long delay free recall (n=21); Raven‟s progressive matrices (n=22); NART = New Adult Reading Test (n=19); MDRS = Mattis Dementia Rating Scale (n=26); Clock Drawing Test (n=9); MMSE = Mini-Mental State Exam (n=31). Missing data is secondary to the inability of the patient to complete the test.

Patients were mildly demented at the time of their initial evaluation based on their mean MMSE score of 21.7 (Standard Error of the Mean = 1.2) and fell below the cut-off for dementia on the

Mattis Dementia Rating Scale (MDRS); mean score in CBS patients was 113.5 (4.1) /144

(MDRS cut-off for dementia in this age group = 123/144). Cognitive domains most impaired

145 were working memory, executive functions, praxis, visuospatial abilities and language tasks involving fluency and naming, with relative preservation of comprehension. Although delayed free recall on the CVLT was impaired, delayed cued recall was better (data not shown). Of the

23 CBS patients for which WAB data was available, the classification was as follows: 26.1%

(6/23) had no aphasia; 8.7% (2/23) were borderline aphasic; 47.8% (11/23) had an anomic aphasia; 8.7% (2/23) had a Broca‟s aphasia; 4.4% (1/23) had a conduction aphasia; and 4.4%

(1/23) had a Wernicke‟s aphasia.

Patients were moderately impaired on both basic and instrumental activities of daily living assessed using the Disability Assessment for Dementia. In terms of frequency of neuropsychiatric symptoms, 24/29 (82.8%) CBS patients had at least one neuropsychiatric symptom present (Supplementary Table 1). Neuropsychiatric symptoms are presented in order from most common to least common: apathy (58.6%), depressive symptoms (41.4%), abnormal appetite and eating behaviour (41.4%), irritability (34.5%), agitation (31.0%), anxiety (27.6%), aberrant night-time behaviour (24.1%), disinhibition (17.2%), aberrant motor behaviour (6.9%), and delusions (3.5%; Supplementary Table 1). No patients had hallucinations or euphoria. 13/30

(43.3%) had a CSDD score > 25% supportive of significant depressive symptomatology

(Supplementary Table 1), although none met DSM-IV diagnostic criteria for depression or had a history of clinically relevant depression before the neurodegenerative presentation.

5.4.1.2 MRI features

Table 1 provides case summaries of clinical, pathological and MRI features of the CBS patients.

Two patients did not have MRI examinations completed due to claustrophobia.

146

AFFECTED AGE OF TYPE OF HEMISPHERIC LOBAR WHITE MATTER SEX PATHOLOGY SIDE OF ONSET PRESENTATION ATROPHY PREDILECTION HYPERINTENSITY BODY 65 F Dementia FTLD-U/TDP43 L R O,P > Fr,Te R>L POST 63 F Dementia AGD/CBD/CAA R L Gen R=L ANT/POST 77 F Dementia N/A R L Fr,Te,P R=L ANT/POST 74 F Motor CBD L SYM Fr,Te,P R=L ANT 75 M Motor N/A R SYM Gen R=L ANT 58 F Dementia N/A L SYM O,P > Te R=L ANT 86 M Dementia N/A L SYM Fr,Te,P R=L ANT 61 F Dementia CBD R L Fr,Te,P L ANT 75 F Dementia N/A R L Fr,P R=L ANT 63 F Motor CBD L SYM Gen R=L ANT/POST 70 F Dementia N/A R SYM Fr,Te,P R=L ANT 57 F Dementia CBD R L Te,P,O > Fr L>R POST 59 F Dementia N/A L SYM P,O > Fr,Te L>R POST 57 M Dementia N/A L R P > Te,O R=L mild 62 F Dementia N/A L SYM P > Fr,Te R=L mild 54 M Dementia N/A L R Gen Absent 59 M Motor N/A L SYM P R>L POST 71 F Motor N/A L R Fr,Te,P R=L ANT/POST 76 M Motor N/A L SYM P > Fr,Te L>R POST 69 M Dementia N/A L R P,T > F R=L ANT 49 F Dementia N/A R SYM P > Fr,Te,O L>R ANT/POST 80 F Dementia N/A R L Fr,Te,P > O L>R POST 68 F Motor N/A R L Fr,Te,P > O R=L mild 55 F Dementia N/A R SYM P > Fr R=L mild 69 F Dementia CBD L R Fr,P > Te R=L mild 71 F Motor N/A R SYM Te,P R=L ANT/POST 46 M Dementia N/A R L Gen R=L mild 62 M Dementia PSP R SYM P Absent 58 M Dementia N/A R L P > Te Absent 67 M Dementia N/A R No MRI N/A N/A 65 M Motor N/A L No MRI N/A N/A Table 1. Case summaries of clinical, pathological, and MRI features of CBS patients. M = Male; F = Female; FTLD-U/TDP43 = Frontotemporal lobar degeneration-ubiquitin+/Tar DNA binding protein+; AGD = agyrophilic grain disease; CAA = cerebral amyloid angiopathy; CBD = Corticobasal degeneration; PSP = Progressive supranuclear palsy; L = Left; R = Right; SYM = Symmetrical; O = occipital; P = Parietal; Fr = frontal; Te = Temporal; Gen = Generalized; POST = posterior; ANT = anterior MRI atrophy pattern Asymmetric Symmetric Right Left Motor Left 6 (20.7%) 0 (0%) 8 (27.6%) side Right 0 (0%) 9 (31%) 6 (20.7%) Column Totals 15 (51.7%) 14 (48.3%) Table 2. MRI atrophy patterns in CBS cases stratified according to body side most affected by motor symptoms. All percentages are calculated based on total of 29 CBS patients who had MRI scans.

In terms of lobar predilection, parietal>temporal>frontal atrophy was most commonly seen with only eight (27.6%) patients having evidence for occipital involvement. Generalized lobar atrophy was observed in five patients; three with maximal involvement contralateral to the most affected side of the body, while two had symmetrical generalized atrophy.

The majority of patients (89.7%; 26/29) demonstrated subcortical T2/PD white matter hyperintensities (WMH) on MRI; 10.3% (3/29) of patients showed no WMH. Of the patients with WMH, 15.4% (4/26) had maximal WMH contralateral to the most affected side of the body corresponding to the region of maximal atrophy. 7.7% (2/26) of patients with symmetrical cortical atrophy had WMH contralateral to the most affected side of the body. 7.7% (2/26) of patients with symmetrical atrophy had WMH ipsilateral to the most affected side of the body.

69.2% (18/26) of patients had symmetrical WMH independent of the most affected side of the body and of cortical atrophy.

5.4.2 Early dementia vs. early motor presentations

5.4.2.1 Demographic and clinical characteristics

There were no statistically significant differences between the early motor and dementia groups in terms of gender, handedness, years of education, age of onset, and body side most affected

(Table 3).

Demographic variable CBS-dementia (n=22) CBS-motor (n=9) Gender 14 (63.6%) F 5 (55.6%) F 8 (36.4%) M 4 (44.4%) M Handedness 21 (95.5%) R 8 (88.9%) R 1 (4.5%)L 1 (11.1%) L Site of recruitment* 19 (86.4%) Cog 1 (11.1%) Cog 3 (13.6%) MD 8 (88.9%) MD Age of Onset 63.6  2.1 69.1  1.9 (mean  SEM years) Age at Investigation 66.8  2.1 72.6  2.0 (mean  SEM years) Duration of symptoms 3.2  0.4 3.4  0.7 (mean  SEM years) Years of Education 12.2  0.7 12.9  0.9 (mean  SEM years) Body side most affected 13 (59.1%) R 3 (33.3%) R 9 (40.9%) L 6 (66.6%) L Table 3. Demographic features of CBS groups presenting with early dementia versus early motor features. F = Female; M = Male; Cog = Cognitive Neurology Clinic; MD = Movement Disorders Clinic; R = Right; L = Left; *Fisher‟s Exact test, p<0.0005.

In terms of clinical characteristics (Figure 2), CBS patients presenting with early motor features were statistically more likely to have cortical sensory loss (defined as occurrence of astereognosis, agraphesthesia and/or sensory extinction) as compared to the early dementia group

(Fisher‟s Exact Test [2-tailed], p=0.005; Figure 2B). This association was driven by the higher occurrence of astereognosis in the CBS-M group (77.8% [7/9 cases]; Fisher‟s Exact Test [2- tailed], p=0.01) compared to CBS-D (22.7% [5/22]). There were no differences between the

149

CBS-M and -D groups in terms of agraphesthesia or extinction to double simultaneous stimuli.

Because sensory extinction is known to localize to the right parietal region, the sample was stratified into those presenting with their left side of the body most affected vs. those involving mainly their right body. Consistent with this localization, there was a trend for CBS-L patients to have higher rates of sensory extinction than CBS-R patients (CBS-L: 33.3% [5/15] vs. CBS-R:

6% [1/16 cases]; Fisher‟s Exact test, p=0.08). There were no significant differences between the

CBS-L and CBS-R groups in terms of occurrence of agraphesthesia or astereognosis. Of the 11 patients who died, time to death was not significantly different between the eight patients who presented with early dementia (80 ± 17 months) and the three patients with early motor features

(91 ± 9 months).

2A 2B *

Figure 2. Frequency of (A) extrapyramidal and (B) cortical features of CBS patients presenting with early dementia vs. early motor symptoms. CSL = Cortical sensory loss; *Fisher‟s Exact Test, p = 0.005

5.4.2.2 Neuropsychological, behavioural and functional evaluation

CBS patients presenting with early dementia had statistically significant lower scores on MMSE;

WAB aphasia quotient (AQ) and subscores including content, fluency, repetition, and

150 comprehension; Boston naming; forward digit span; and Benton line orientation than those presenting with early motor features (See Table 4). This indicates relative difficulties in general cognition, language function, and tasks involving sustained attention and visuospatial orientation and attention.

Psychometric Measures CBS-D (n) CBS-M (n) General cognition MMSE /30 [n=31]¥ 19.8 ± 1.5 (22) 26.6 ± 1.0 (9) Clock Drawing Test /10 [n=9] 6.8 ± 1.1 (6) 7.0 ± 1.7 (3) NART /127.8 [n=19] 104.4 ± 2.5 (11) 111.7 ± 2.6 (8) Raven‟s Progressive Matrices /36 [n=22]* 18.8 ± 1.3 (14) 27.3 ± 3.0 (8) MDRS /144 [n=26] 108.0 ± 5.3 (17) 123.9 ± 4.7 (9) Memory CVLT Long Delay Free Recall /16 [n=21] 5.9 ± 0.7 (13) 7.6 ± 1.2 (8) Delayed Visual Reproduction /41 [n=19] 8.7 ± 2.8 (12) 13.7 ±4.1 (7) Language WAB total /100 [n=23]* 82.7 ± 3.0 (17) 94.5 ± 1.9 (6) WAB content /10* 7.8 ± 0.4 9.1 ± 0.4 WAB fluency /10* 8.0 ± 0.5 9.3 ± 0.2 WAB comprehension /10¥ 8.9 ± 0.3 9.9 ± 0.1 WAB repetition /10¥ 8.5 ± 0.3 9.8 ± 0.1 WAB naming /10 8.1 ± 0.3 9.0 ± 0.4 Boston Naming /30 [n=22]* 22.2 ± 1.3 (14) 27.1 ± 0.8 (8) Semantic Fluency /20 [n=26] 8.3 ± 1.2 (18) 13.3 ± 2.8 (8) Praxis WAB praxis /60 [n=27] 51.7 ± 2.4 (18) 56.1 ± 0.9 (9) Attention & working memory Digit span - forward /12 [n=23] ¥ 5.9 ± 0.6 (15) 9.1 ± 0.7 (8) Digit span - backward /12 [n=23] 3.9 ± 0.8 (15) 5.8 ± 0.9 (8) Visuospatial abilities Rey Osterieth Complex Figure – Copy /36 [n=20] 13.8 ± 3.6 (12) 21.7 ± 4.9 (8) Benton Line Orientation /30 [n=21]* 9.5 ± 2.6 (13) 19.3 ± 4.2 (8) Executive functions Phonemic fluency (FAS) [n=21] 16.6 ± 2.8 (13) 23.1 ± 5.4 (8) Trail Making Test A (time in seconds) [n=19] 119.9 ± 14.3 (12) 89.6 ± 31.7 (7) Trail Making Test B (time in seconds) [n=13] 273.6 ± 63.3 (7) 145.5 ± 28.5 (6) WCST categories /6 [n=22] 1.6 ± 0.4 (13) 2.1 ± 0.4 (9) WCST perseverative errors [n=22] 12.5 ± 3.8 (13) 10.6 ± 4.7 (9) Table 4. Mean scores (± SEM) on neuropsychological measures in CBS patients presenting with early dementia vs. early motor symptoms. The number of patients (n) tested is listed next to individual measures. Missing data is secondary to the inability of the patient to complete the test. CBS-D = Early dementia; CBS-M = Early motor; MMSE = Folstein‟s Mini-Mental

151

State Exam; NART = National Adult Reading Test; MDRS = Mattis Dementia Rating Scale; CVLT = California Verbal Learning Test; WAB = Western Aphasia Battery; FAS = F-, A-, and S-phonemic fluency; WCST = Wisconsin Card Sort Test. Independent samples t-tests were used to compare MMSE, NART, Clock, MDRS, Boston naming, semantic fluency, visual reproduction, forward and backward digit span, CVLT, Benton, Trails A and B mean scores between groups. Mann Whitney U test was used to compare scores on FAS, WAB, Rey and WCST between groups. *p≤0.05; ≦≤0.005

The early dementia group was statistically more likely to have a language disturbance based on combined aphasia categories on the WAB as compared to those presenting with early motor features (Fisher‟s Exact Test [2-tailed], p=0.02). Specifically, 33.3% of the early motor group had an anomic aphasia, while the rest had no aphasia. Contrast this to 88.2% of the early dementia group having a language disturbance (borderline aphasia: 11.8%; anomic aphasia:

52.9%; Broca‟s aphasia: 11.8%; conduction aphasia: 5.9%; Wernicke‟s aphasia: 5.9%) while

11.8% had no aphasia (Figure 3). This difference was even more striking when aphasia classification was stratified according to side of maximal motor involvement. Specifically, 100%

(10/10) of patients presenting with their right side of the body most affected had evidence for an aphasic disturbance, whereas only 46% (6/13) of those with left sided motor symptoms were classified as having aphasia (Fisher‟s Exact Test [2-tailed], p = 0.007). Additionally, mean WAB

AQ scores were lower in those presenting with the right side of the body most affected (78.9 ±

3.8) compared to the left (91.0 ± 2.6) [Mann-Whitney U test (2-tailed), p = 0.005].

152

Figure 3. Frequency of CBS patients with early dementia vs. early motor presentation stratified according to category on the Western Aphasia Battery (WAB). None of the functional or behavioural measures were statistically different between the early dementia and early motor groups (Table 5).

153

Behavioural/Functional Measures CBS-D (n) CBS-M (n) Activities of daily living DAD (%) [n=30] 65.7 ± 7.0 (21) 75.8 ± 9.9 (9) DAD ADL (%) 77.0 ± 7.6 76.0 ± 9.1 DAD iADL (%) 58.6 ± 7.8 75.7 ± 10.5 Neuropsychiatric symptoms NPI total /144 [n=29] 14.4 ± 3.4 (20) 7.1 ± 2.7 (9) NPI apathy /12 4.1 ± 1.0 2.1 ± 1.3 NPI appetite and eating behaviour /12 2.3 ± 0.7 1.8 ± 1.2 NPI dysphoria/depression /12 1.7 ± 0.6 0.6 ± 0.2 NPI night-time behaviour /12 1.1 ± 0.5 1.7 ± 1.1 NPI irritability/lability /12 1.3 ± 0.5 0.4 ± 0.3 NPI agitation/aggression /12 1.1 ± 0.6 0.3 ± 0.2 NPI aberrant motor behaviour /12 0.8 ± 0.6 0.0 ± 0.0 NPI disinhibition /12 0.7 ± 0.3 0.0 ± 0.0 NPI anxiety /12 0.7 ± 0.3 0.2 ± 0.2 NPI delusions /12 0.3 ± 0.3 0.0 ± 0.0 NPI hallucinations /12 0.0 ± 0.0 0.0 ± 0.0 NPI euphoria /12 0.0 ± 0.0 0.0 ± 0.0 NPI caregiver distress /12 7.9 ± 1.9 4.2 ± 1.6 Cornell Depression Scale (%) [n=30] 24.3 ± 3.1 (21) 19.3 ± 4.9 (9) Table 5. Mean scores (± SEM) on behavioural and functional measures in the CBS group. The number of patients (n) tested is listed next to individual measures. Missing values are secondary to the inability of the caregiver to complete the test. CBS-D = Early dementia; CBS-M = Early motor; DAD = Disability Assessment for Dementia; ADL = Activities of daily living; iADL = Instrumental activities of daily living; NPI = Neuropsychiatric Inventory. There were no statistically significant differences between groups.

5.4.2.3 MRI features

There were no significant differences between the early dementia and motor groups in terms of symmetry/asymmetry of atrophy on MRI. Stratifying CBS patients with asymmetric MRI atrophy into those with and without aphasia using the WAB, there was a trend for aphasic patients to have left hemispheric atrophy compared to those without aphasia (Fisher‟s Exact Test, p=0.06; Refer to Table 6).

MRI atrophy pattern Asymmetric* Symmetric Right Left Aphasia Present (n=17) 2 6 7 Absent (n=6) 3 0 3 Table 6. MRI atrophy patterns in CBS cases stratified by the presence or absence of aphasia as determined by the WAB. Fisher‟s exact test, p=0.06*

154

5.4.2.4 SPM and ROI SPECT

Figure 4 and tables 7 and 8 demonstrate results of SPM and ROI SPECT analysis comparing perfusion in the CBS patients presenting with early dementia to those with early motor features.

In the early dementia versus early motor groups, cortical areas of relative hypoperfusion were identified in the left fusiform gyrus (uncorrected p<0.001); this result did not survive correction for multiple testing using FDR or FWE methods. However, employing ROI MANCOVA analysis, CBS patients presenting with early motor symptoms had relatively reduced perfusion in the right precentral gyrus and right paracentral lobule (supplementary motor area) as well as in the left middle posterior cingulate region compared to those with early dementia.

Figure 4. Statistical parametric maps overlaid on multi-slice brain MRI template showing decreased perfusion in left fusiform gyrus (uncorrected p<0.001) in CBS cases presenting with early dementia versus early motor features.

155

Group Anatomical locus Talairach No. of SPM t-score (Brodmann area) Coordinates voxels (p-value) X y Z CBS-d Occipitotemporal region - SPM vs. Left fusiform gyrus (19) -34 -76 -11 215 4.5 (p<0.001) CBS-m Table 7. Areas of relative hypoperfusion on SPECT in CBS patients presenting with early dementia (CBS-d) versus those presenting with early motor features (CBS-m). Data are shown for only SPM analysis since no areas of hypoperfusion were shown with ROI; Uncorrected p-value of p<0.001 used.

Group Anatomical locus No. of ROI F-score (Brodmann area) voxels (p-value)

CBS-m Frontal regions – ROI vs. Right paracentral lobule 741 7.0 (p<0.01) CBS-d Right precentral gyrus 2723 4.9 (p<0.04)

Limbic regions - ROI Left posterior cingulate - middle 512 5.1 (p<0.03) Table 8. Areas of relative hypoperfusion on SPECT in CBS patients presenting with early motor (CBS-m) versus those presenting with early dementia (CBS-d). Data are shown for ROI analyses only since SPM results failed to demonstrate any relative perfusion differences in this comparison. All p-values for the ROI analysis are derived from a multivariate ANOVA running under a general linear model.

5.4.3 Description of pathological series and relation to MRI findings

Of the 31 CBS patients in our series, 11 patients died from their neurodegeneration usually from malnutrition and/or aspiration pneumonia secondary to severe dysphagia. The average time to death was 82.7 (SEM 12.2; range 32 to 159) months or approximately 7 years. Pathological analysis was performed on 8 of the 11 patients who died. 5 patients met pathological criteria for

CBD, one had PSP, one had FTLD-U/TDP43 proteinopathy and one had combined dementia with agyrophilic grains, CBD and cerebral amyloid angiopathy (AGD/CBD/CAA). The average time to death in the pathologically-confirmed CBD group was 79.2 (22.5) months with a range of

32 to 159 months. The patient with pathologically-proven PSP died after approximately 94 months, while the time to death in the AGD/CBD/CAA and FTLD-U/TDP43 patients were 117

156 and 112 months, respectively. The CBS cases with FTLD-U/TDP43 (left motor symptoms),

AGD/CBD/CAA (right motor symptoms), or PSP (right motor symptoms) pathology presented with an early dementia syndrome. 60% (3/5) of the CBD cases presented with an early dementia syndrome; two of these cases had right-sided motor symptoms while one had left-sided involvement. 40% (2/5) of the CBD cases presented with early motor features and both had the left side prominently affected by motor symptoms.

There was a relatively good association between the severity and lateralization of cortical atrophy/subcortical white matter changes observed on MRI in vivo with that of the underlying pathology. Of the CBS cases with a pathological diagnosis of CBD, the pattern of cortical atrophy detected on MRI matched that detected by pathological investigation in 60% (3/5) of cases; in the two cases that did not match, MRI-detected atrophy was asymmetrical while the macroscopic brain pathology showed symmetrical atrophy. Also in 100% (5/5) of the CBD group, the severity of white matter changes on MRI correlated well with the severity of underlying Tau-positive threads and glial coils observed in the white matter with associated pallor and gliosis. The case with FTLD-U/TDP43 demonstrated marked right > left-sided cortical atrophy worse in the parieto-occipital region, but also involving the frontotemporal regions on MRI [Masellis et al. 2006]. There was also severe underlying white matter hyperintensities worse on the right in the posterior regions. These MRI findings are strongly correlated with pathological changes of underlying cortical and white matter atrophic and gliotic changes. The generalized left > right-sided atrophy seen on MRI in the AGD/CBD/CAA case was discordant with the symmetrical atrophic changes noted on pathology. However, MRI white matter hyperintensities did correlate with white matter neuropil threads and coiled bodies.

157

Finally, the mild symmetrical atrophy on MRI detected in the PSP case and absence of white matter changes correlated well with that of the cortical atrophy seen on pathology as well as the minimal white matter gliotic changes. None of the cases had evidence for clasmatodendrosis in the white matter.

5.5 Discussion

This study provides a comprehensive and integrated multi-modal assessment of a prospective cohort of CBS patients using clinical data, standardized neuropsychological, neuropsychiatric and functional measures, as well as brain MRI and SPECT data. Patients were ascertained from both university-based cognitive and movement disorders clinics overcoming the important limitation of selection bias in prior studies. Many prior studies did not stratify patients based on most affected motor side of the body as well as presenting syndrome – dementia or motor – and our study significantly adds to the literature that has studied lateralization in this syndrome.

Finally, our CBS cohort was followed prospectively over time with about 25% of our sample coming to autopsy, which was performed in 72% (8/11) who died during the course of the study.

We will now discuss highlights and novel findings of this study.

5.5.1 CBS presenting with early dementia vs. early motor features

There have been very few studies that directly compare the clinical and neuropsychological profiles of CBS presenting with early dementia (CBS-D) vs. early motor features (CBS-M)

[Kertesz et al. 2000b;McMonagle et al. 2006]. The current study represents the first attempt to

158 compare these CBS subtypes in terms of their relative brain perfusion on SPECT as a biomarker of brain dysfunction. Compared to previous studies, it also provides a more comprehensive, comparative neuropsychological assessment of these CBS subtypes that cover all domains of cognitive functioning.

A highly significant finding was that all CBS patients presenting with early motor features demonstrated cortical sensory loss on clinical examination in comparison to less than half of those presenting with early dementia. To our knowledge, this is the first study to demonstrate this finding. Although sensory extinction can be represented bilaterally in the parietal regions, it is more commonly associated with lesions involving the right (non-dominant) superior parietal lobe, specifically, areas 5 and 7 in the inferior part [Rizzo and Eslinger 2004;Ropper and Brown

2005]. Although we could not localize precisely to this region based on the nature of our data, the CBS patients with prominent symptoms involving the left side of the body had higher rates of sensory extinction. Agraphesthesia (ability to recognize figures drawn on the hand) has been associated with lesions of the left intraparietal sulcus [Rizzo and Eslinger 2004]. However, in our sample, we did not observe any association of left hemispheric/parietal atrophy with presence of agraphesthesia. This may be due to low power to detect a difference using cateogorical data

(presence/absence of left hemispheric atrophy on MRI) secondary to a relatively small sample size. Astereognosis (an inability to tactually perceive both texture [ahylognosis] and shape

[amorphognosis] of an object with the hands) localizes to parietal areas 1 (ahylognosis), and 2 through 5 (amorphognosis) and is bilaterally represented [Rizzo and Eslinger 2004]. Our findings support the bilateral localization of this clinical phenomenon in that there was no lateralization of MRI-rated atrophy observed.

159

Lateralization in our sample is further supported by a careful examination of language function within the CBS-D vs. -M groups. Specifically, patients with language dysfunction classified according to the WAB were more likely to have their right side of the body most affected, which was more commonly the case in the early dementia group. Similar to prior studies [Kertesz et al.

2000b;McMonagle et al. 2006], our CBS patients presenting with early dementia tended to perform more poorly on the WAB as exemplified by lower mean aphasia quotients and were more likely to be classified as having severe aphasic disturbances. Furthermore, the CBS-D group had significant impairments in naming on the Boston Naming Test compared to the CBS-

M group. This association was even stronger when examining for the presence of aphasic disturbance based on the most affected motor side of the body with 100% of right side afflicted patients demonstrating aphasia compared to less than half of those with their left body side affected. Compared to a trend towards this finding in McMonagle et al. [McMonagle et al.

2006], our study achieved high statistical significance. This difference may be due to patient selection biases – all of our patients had to have substantial asymmetric rigidity and/or apraxia at some point during the disease course to increase the specificity of the CBS diagnosis whereas 5 patients in McMonagle et al [McMonagle et al. 2006] had clinical findings that were symmetrical and “atypical” for CBS; our patients were recruited prospectively from both a cognitive or movement disorders clinic; or this discrepancy may be due to other unknown biases in the data due to small sample sizes across both patient series.

160

Although right-sided motor involvement was strongly associated with aphasia in this study, an examination of MRI data did not show such a predictable relationship with side of maximal atrophy. Patients with aphasia tended to have either symmetric or left hemispheric atrophy on

MRI compared to those without aphasia who had symmetric or right hemispheric atrophy. These results did not achieve statistical significance, but were based on a small sample size. This lack of an association of aphasia with side of atrophy is consistent with the study of McMonagle et al

[McMonagle et al. 2006]. Asymmetric atrophy on MRI was not found to predict the CBS-D vs.

CBS-M presentation. Few studies have directly compared these two subtypes of CBS utilizing

MRI. One MRI study demonstrated that pathologically-proven CBD patients (n = 11) had more cortical and subcortical grey matter atrophy on MRI if they presented with dementia symptoms

(CBD-D) compared to a higher degree of subcortical white matter atrophy in motor-onset CBD

(CBD-M) [Josephs et al. 2008]. Similar to our study, their analysis comparing CBD-M and

CBD-D cases was completed qualitatively by visual inspection even though they used voxel- based morphometric analysis when comparing CBD patients to controls [Josephs et al. 2008].

The lack of association between MRI-rated atrophy and pattern of CBS presentation is likely due to the fact that even though one hemisphere may be more selectively vulnerable at an earlier stage of disease, ultimately the pathology underlying CBS is bilateral and frequently does not always correlate perfectly with the clinical syndrome. Perfusion deficits may be more sensitive to detect asymmetry in CBS presenting with cognitive- versus motor-onset.

We hypothesized that the CBS-D group would show reduced perfusion in the left peri-Sylvian region compared to CBS-M given that aphasia was more common and severe in this group.

Conversely, we hypothesized that the CBS-M group would show reduced perfusion in the right

161 parietal region compared to the CBS-D group because of the high preponderance of cortical sensory loss. There were no significant areas of reduced perfusion between the CBS-D and CBS-

M groups identified that survived correction for multiple testing. This was likely on the basis of the small sample size in each CBS group (CBS-D, n = 22; CBS-M, n = 9) so there was insufficient power to detect any differences using the conservative SPM method. Even when the stringency of a whole brain SPM analysis was reduced to explore the data, the only area which showed reduced perfusion in CBS-D versus -M was the left fusiform gyrus. The fusiform gyri have been identified as important neural correlates of both facial and word recognition and perception [Rizzo and Eslinger 2004]. Damage to the right fusiform gyrus is associated with prosopagnosia, an inability to recognize faces [Rizzo and Eslinger 2004]. The left fusiform gyrus plays an important role in recognition and processing of visual word forms and as such is also known as the „visual word form area‟ important in processing strings of letters [Cohen et al.

2000]. However, a more recent study has also shown its importance in perception and memory of faces [Mei et al. 2010]. To our knowledge, there have been no prior studies demonstrating reduced perfusion or atrophy in this region in CBS nor has prosopagnosia or impaired processing of letter strings been observed as a feature of CBS. However, in a combined group of neurodegenerative disorders including a small subsample with CBS, empathy loss was associated with atrophy in the right fusiform gyrus among other frontotemporal regions [Rankin et al.

2006]. The authors suggest that the importance of this region in facial perception may be related to its association with empathy. Future studies will need to clarify whether or not this structure is indeed important in CBS, especially given that our result could be a false positive association.

162

Interestingly, in the ROI analysis, the supplementary motor area on the right showed reduced perfusion in CBS-M compared to CBS-D. This may be an important neural correlate of the motor disorder associated with CBS and the observation that it lateralizes to the right in CBS-M patients is consistent with the finding of an increased number of left-sided motor presenters in the CBS-M group. In the longitudinal study of McMonagle et al. [McMonagle et al. 2006], there was a shorter onset to the development of motor symptoms in patients with prominent right hemispheric atrophy or left-sided akinesia in support of our perfusion findings. Our finding is further substantiated by a voxel-based morphometry (VBM) MRI study demonstrating that pathologically-proven CBD patients presenting with a prominent extrapyramidal syndrome, as opposed to dementia, had atrophy compared to controls involving the superior premotor cortex extending into the posterior superior, middle and inferior frontal lobes [Josephs et al. 2008].

Grey matter involvement of the supplementary motor area and parietal lobes was also observed

[Josephs et al. 2008]. Although the atrophy pattern was bilateral, it was slightly more pronounced in the right hemisphere [Josephs et al. 2008]. A limitation of this study was that it did not document the most affected side of the body in their CBS sample [Josephs et al. 2008].

Consistent with our patients meeting DSM-IV criteria for a diagnosis of dementia, MMSE scores were, as expected, found to be significantly lower in the CBS-D versus CBS-M group. The CBS-

D group performed significantly worse on the forward digit span indicating a lower primary attention span suggestive of fronto-subcortical dysfunction. The CBS-D group also showed significantly more deficits on the Benton judgement of line orientation indicative of right parieto-occipital dysfunction. These tests do not rely on intact motor function so the differences are likely not related to the degree of motor impairment or apraxia.

163

To our knowledge, no studies have compared neuropsychiatric features in CBS-D versus CBS-M case series using the Neuropsychiatric Inventory. Although our results did not achieve statistical significance, the CBS-D group had higher rates of apathy, irritability, and depressive symptoms which typically involve the limbic-prefrontal circuit. Abnormal appetite and eating behaviours were also seen in CBS-D versus CBS-M suggestive of more temporal lobe involvement in this group. Using the Frontal Behavioural Inventory (FBI), a validated, caregiver-administered rating scale of frontotemporal behavioural symptoms, Kertesz et al. [Kertesz et al. 2000b] demonstrated variable, but higher scores on the FBI in cognitive-onset CBS patients compared to motor-onset CBS consistent with this disorder having symptoms occurring along the spectrum of

FTLD.

5.5.2 Pathology

CBS is not only a clinically heterogeneous disorder, but it is also highly variable in terms of the underlying pathologies that can cause the syndrome. It is the location and burden of the pathology that produces the clinical syndrome with likely a lesser contribution coming from the specific pathological changes [Lang 2003]. As such, the clinical syndrome of CBS does not always predict the specific underlying pathology of CBD [Ling et al. 2010]. Wadia & Lang

[Wadia and Lang 2007] reviewed several studies (total of 83 cases) demonstrating that the CBS predicts CBD pathology approximately 55% of the time. The second most common pathology was PSP (21%) followed by Pick‟s disease (7%) [Wadia and Lang 2007]. McMonagle et al.

[McMonagle et al. 2006] followed 19 CBS patients prospectively until autopsy. Specifically,

164 they found that a clinical diagnosis of CBS predicted CBD pathology in 58% of the cases, and predicted underlying Tau histology in 84% of cases. Other pathologies included Alzheimer‟s disease, FTLD-U pathology, and Gerstmann Straussler Scheinker disease [McMonagle et al.

2006]. Our study also followed patients prospectively until autopsy and found that the CBS predicted CBD histology in 63% of cases. The rates of CBD pathology in our CBS sample are similar to prior studies [Boeve et al. 1999;McMonagle et al. 2006]. Other Tau based pathologies

(25%) included mixed CBD with agyrophilic grains and PSP. Although the majority of our cases were sporadic, one of our patients with CBS had a strong family history of CBS and ended up having FTLD-U/TDP43 pathology due to PGRN mutation [Masellis et al. 2006]. With the identification of mutations in PGRN as a major cause of FTLD spectrum disorders and in particular because the FTLD-U/TDP43 pathology often affects the parietal lobes, CBS related to

PGRN mutation has been more commonly identified in recent years [Benussi et al. 2008;Benussi et al. 2009;Gabryelewicz et al. 2010;Gass et al. 2006;Ghetti et al. 2008;Guerreiro et al.

2008;Kelley et al. 2009;Le, I et al. 2008;Lopez de et al. 2008;Moreno et al. 2009;Rademakers et al. 2007;Rohrer et al. 2009;Spina et al. 2007;Yu et al. 2010].

We observed an association between atrophy patterns and white matter hyperintensities found on

MRI with findings on neuropathological examination. For the most part, the white matter changes on MRI correlated with underlying pathological white matter atrophy and gliosis. In the two CBD cases, who did not show corresponding asymmetry on the pathological examination, we suspect that by the time of death, the asymmetrical cortical atrophy observed on MRI had progressed to symmetrical involvement.

165

5.5.3 MRI investigation

Assuming that 1) the presenting lateralized clinical syndrome correlates most strongly with the area of maximal brain pathology and 2) brain atrophy on MRI is a biomarker of the underlying burden of pathology, then why do only half of our cases demonstrate asymmetrical brain atrophy contralateral to the most affected body side? One possibility is that the heterogeneous nature of pathology underlying CBS may cause varying degrees of atrophy and of hemispheric/lobar asymmetry. This may be even the case within subtypes of the same pathological substrate; a study that examined MRI atrophy patterns in patients with diagnoses occurring along the FTLD spectrum stratified according to type 1, 2, or 3 FTLD-U/TDP43 pathological subtypes demonstrated that type 1 and 3 pathology were associated with asymmetrical atrophy whereas type 2 pathology was associated with symmetrical atrophy [Rohrer et al. 2010a]. Similarly,

Whitwell et al. [Whitwell et al. 2010] stratified patients presenting with CBS based on their underlying pathological diagnosis and compared the subgroups based on their MRI patterns of atrophy. Although they did not comment on this specifically, it can be extrapolated from their paper that CBS patients with FTLD-U/TDP43 and AD pathology tended to have more asymmetrical MRI atrophy whereas those with CBD pathology had more symmetrical atrophy patterns [Whitwell et al. 2010].

Another possible explanation to account for the lack of asymmetrical atrophy in half of our CBS patients is that assessment of grey matter loss on MRI may not be a sensitive enough measure.

As such, SPECT perfusion or FDG-PET hypometabolism may be more sensitive in detecting

166 hemispheric asymmetries in CBS compared to MRI. Consistent with this, Mendez et al [Mendez et al. 2007] found that the use of SPECT/PET increased the sensitivity of establishing the correct diagnosis of FTD (90.5%) compared to MRI atrophy patterns (63.5%). We have also recently shown in a case that initially presented with PNFA and later evolved to CBS with underlying

FTLD-U/TDP43 pathology due to a novel PGRN mutation that longitudinal SPECT perfusion loss in the less affected hemisphere occurred before atrophy had progressed on that side

[Gabryelewicz et al. 2010]. Similarly, in two cases of very early stage FTD followed longitudinally, neuroimaging can be initially normal, but when perfusion abnormalities and atrophy are eventually shown on SPECT and MRI, respectively, the perfusion abnormalities are more extensive than the atrophy patterns [Gregory et al. 1999].

Subcortical white matter disease may also be contributing to some of the clinically lateralized dysfunction and should also be considered as a potential biomarker underlying asymmetry in

CBS. Examining our MRI data, close to 90% of CBS cases demonstrated hyperintensities in the white matter on T2/PD sequences, and, in about 23%, these were localized contralateral to the most affected side of the body correlating with side of maximal atrophy (15%) or occurred contralateral to the motor deficits independent of atrophy in symmetrical cases (8%). This data is descriptive, but supports the idea that some white matter hyperintensities do not just represent age-associated microangiopathy [Levy-Cooperman et al. 2008b], but instead may be the result of white matter glial damage related to CBD [Tan et al. 2005;Forman et al. 2002]. Our MRI- pathological correlative data confirm this is indeed the case in CBS due to a variety of pathologies. Recent diffusion tensor imaging (DTI) studies examining the integrity of white matter tracts adds stronger evidence to this argument. Specifically, two studies have shown

167 reduced fractional anisotropy and increased mean diffusivity in CBS occurring contralateral to the most affected side of the body compared to controls, suggesting that damage to the integrity of white matter tracts may account for some of the contralateral clinical findings [Boelmans et al.

2009;Bozzali et al. 2008]. Another DTI study found that mean diffusivity was elevated in the motor thalamus in CBD ipsilateral to the most affected hemisphere (i.e., contralateral to most affected side of the body), while mean diffusivity was elevated bilaterally in anterior and medial thalamic areas in PSP [Erbetta et al. 2009].

5.5.4 Limitations

In terms of limitations, although our sample size would be considered reasonably large given the rarity of the CBS phenotype, from a statistical perspective, the sample was indeed small with a low power to detect differences especially with the use of non-parametric measures as well as with multivariate analyses. This was particularly evident in the SPM analyses that attempted to correct for multiple testing. Other limitations include the fact that not all patients were able to complete the neuropsychological tests given that their degree of dementia or alternatively degree of motor disability may have precluded this. Along the same lines, it is difficult to know if impaired performance on tasks that involve writing and drawing were affected by the cognitive disruption, motor disability or both. As such, our neuropsychological data is likely biased in favour of those with milder forms of CBS and this will further reduce the power of the analysis.

Another limitation is that we employed a qualitative instead of a quantitative analysis of the MRI data since there were only 21 patients who completed MRIs that were not degraded by motion

168 artifact. Therefore, only about 2/3 of the sample had useable MRI data for a quantitative analysis, which is currently in progress in an extended sample.

5.5.5 Conclusions

The current study provides a cross-sectional examination of neuropsychological, MRI and

SPECT features of a prospectively ascertained sample of CBS patients around the time of their initial diagnosis with a subset followed until autopsy. It highlights the importance of having a phenotypically well-characterized sample of patients diagnosed with CBS and provides new insights into the neuropsychological and neuroimaging correlates of lateralized brain dysfunction in the syndrome. Future studies will include analyzing neuropsychological, SPECT and MRI changes longitudinally as the disease progresses with correlation to underlying pathology. Only this kind of study will yield a true incidence of CBD and related pathological subtypes of the syndrome. Finally, we believe that one of the issues that makes identification and replication of genomic risk factors for neurodegenerative syndromes challenging is the significant heterogeneity across these conditions. It will be increasingly necessary that groups share their data in order to conduct larger studies, and, more importantly, efforts are made to develop novel endophenotypes of these syndromes that will facilitate the identification of genomic and epigenomic risk factors for CBS [Masellis et al. 2010].

169

Behavioural/Functional Measures Scores Activities of daily living DAD (%) [n=30] 68.8 ± 5.7 DAD ADL (%) 76.7 ± 5.9 DAD iADL (%) 63.9 ± 6.4 Neuropsychiatric symptoms NPI total /144 [n=29] 12.1 ± 2.5 NPI apathy /12 3.5 ± 0.8 NPI appetite and eating behaviour /12 2.1 ± 0.6 NPI dysphoria/depression /12 1.3 ± 0.5 NPI night-time behaviour /12 1.2 ± 0.5 NPI irritability/lability /12 1.0 ± 0.4 NPI agitation/aggression /12 0.9 ± 0.4 NPI aberrant motor behaviour /12 0.6 ± 0.4 NPI disinhibition /12 0.5 ± 0.2 NPI anxiety /12 0.5 ± 0.2 NPI delusions /12 0.2 ± 0.2 NPI hallucinations /12 0.0 ± 0.0 NPI euphoria /12 0.0 ± 0.0 NPI caregiver distress /12 6.7 ± 7.5 Cornell Depression Scale (CSDD; %) [n=30] 22.8 ± 2.6 Supplementary Table 1. Mean scores (± SEM) on behavioural and functional measures in the CBS group. The number of patients (n) tested is listed next to individual measures. Missing values are secondary to the inability of the patient or caregiver to complete the test. DAD = Disability Assessment for Dementia; ADL = Activities of daily living; iADL = Instrumental activities of daily living; NPI = Neuropsychiatric Inventory. CSDD = Cornell Scale for Depression in Dementia.

170

6.0 Summary and General Discussion

171

6.0 Summary and General Discussion

This thesis has examined several aspects of CBS through the comprehensive study of 31 patients who met criteria for the clinical syndrome [Boeve et al. 2003]. Multi-modal assessments were used to characterize the patients including a detailed neurological examination by a movement disorder and/ or cognitive neurologist with experience in diagnosing the condition, comprehensive neuropsychological and neuropsychiatric testing, qualitative visual assessment of

MRI scans, and semi-quantitative assessment of SPECT perfusion images using both region of interest and voxel-wise approaches to data analysis. Additionally, a subset of eight patients underwent neuropathological examination in order to identify the underlying pathological substrate of the syndrome and also to correlate imaging features with the burden of observed pathology. Finally, two families that segregated CBS and related FTD spectrum disorders due to

PRGN mutation are discussed in terms of their heterogeneity in clinical and neuroimaging findings. This thesis provides many new insights into aspects of CBS from both the genetic and brain-behaviour correlative perspective and also confirms many findings of previous studies. The overall thesis findings will now be discussed in the context of prior literature. Limitations and future recommendations will be also be reviewed.

6.1 Representative sample

6.1.1 Demographic features

The mean age of onset in our CBS sample was approximately 65 years, and 61% of our sample was female. Our basic demographics compare well to those of previously published case series of CBS that have shown a mean age of onset of approximately 63 [Wenning et al. 1998] and 61

172 years [Murray et al. 2007] with a predominance of affected females in several studies

[Mahapatra et al. 2004]. Differences in age of onset between our study and others might reflect the fact that these studies included only pathologically confirmed cases of CBD, some of whom did not meet clinical criteria for CBS, while our study included clinically-diagnosed cases with a small proportion of pathologically-confirmed ones. When we examined the age of onset in our 5 pathologically-confirmed cases of CBD, it was 64.8 (standard error of the mean 3.0) years similar to that found in our entire patient sample.

6.1.2 Clinical and neuropsychological features

In terms of clinical features, about half of the patients presented with right-sided predominant symptoms, while the left was most severely affected in the other half. This distribution of asymmetry is consistent with other large studies [Huey et al. 2009a;Riley et al. 1990;Rinne et al.

1994;Shelley et al. 2009]. Overall, there does not seem to be a predilection for one hemisphere over another, although this has been seen perhaps as an artifact in smaller samples by chance alone [Chang et al. 2007]. At the time of their initial presentation, on average, about three years into the course of the disease, the most common symptoms/signs were asymmetric rigidity and apraxia, which eventually occurred in all patients. This was very similar to several large studies

[Kompoliti et al. 1998;Riley and Lang 2000;Rinne et al. 1994;Wenning et al. 1998].

Patients presenting with early dementia before the onset of motor features was also relatively frequent affecting approximately 71% of the sample. The frequent occurrence of an early dementia presentation is similar to previous studies [Bergeron et al. 1998;Grimes et al.

173

1999b;Mathuranath et al. 2000], although it may have been biased in our study because close to

2/3 of patients were recruited from a cognitive neurology clinic. However, there were some cases with early dementia who presented to a movement disorders clinic initially. The presence of a language or speech disturbance was also very common in our study. Cortical sensory loss manifested by extinction to double blind tactile stimuli, agraphesthesia and/or astereognosis was also relatively frequent followed by dystonia. Contrary to previous studies, alien limb phenomenon in our cohort was uncommon, although limb levitation was relatively common.

This likely reflects the fact that many prior studies did not distinguish between these two phenomena, as there is debate as to the true definition of alien-limb phenomenon (reviewed in

Boeve et al., 2003 Lang). Apart from the differences highlighted above, our cohort presented with similar clinical features as previous cohorts [Bergeron et al. 1998;Grimes et al.

1999b;Mathuranath et al. 2000] suggesting our sample is representative.

As expected, compared to the normal control group, the CBS patients were significantly impaired across all domains of cognitive functioning and were mildly demented based on MMSE and MDRS scores. Most pronounced deficits were noted on tests of working memory, executive functions, praxis, and visuospatial abilities. This cognitive profile is quite typical of prior studies in CBS, which have shown frontal subcortical and visuospatial dysfunction as well as significant apraxia, the latter being the most common finding [Graham et al. 2003b].

Another study compared several tasks of executive function using the Delis-Kaplan Executive

Function System (D-KEFS) in CBS to an FTD group [Huey et al. 2009a]. This study found that

174 although both groups exhibited prominent executive dysfunction, performance on most executive tests tended to be worse in FTD except for tasks such as Trail Making and timed measures of the

Tower test. These tests require intact motor and visuospatial function that are more impaired in

CBS than in FTD [Huey et al. 2009a].

Episodic memory disturbance has also been documented, although this tends to be variable across individual patients and based on severity of illness [Graham et al. 2003b]. It is thought to reflect dysfunction of frontal subcortical circuits, rather than primary hippocampal involvement, and, as such, episodic memory function in CBS also tends to be less affected than that seen in

AD [Graham et al. 2003b]. Two other studies have also confirmed relative preservation of episodic memory function in CBS [Huey et al. 2009a;Murray et al. 2007]. Although we did not compare episodic memory performance in our cohort directly with that of an AD cohort, performance on delayed cued recall on the CVLT was better than delayed free recall supporting the premise that it may be associated with poor use of strategic processes in encoding and retrieval (i.e., frontal-subcortical dysfunction) as proposed by Pillon and Dubois [Pillon and

Dubois 2000].

Aphasia was commonly associated with CBS in our sample. Based on purely a clinical assessment, 77% of patients had an observed language disturbance. This finding was strongly supported using the WAB, a formal rating instrument of language function, as approximately

74% of the sample was identified as having an aphasic disturbance at the time of their initial neuropsychological testing. Based on the data available, we were not able to determine how

175 many individuals had an associated apraxia of speech, which along with progressive non-fluent aphasia, have been proposed to be clinical markers of both CBD and PSP pathology [Josephs and

Duffy 2008]. The most common language disturbances seen in our sample were anomic and non- fluent aphasia subtypes, which are consistent with findings of previous studies [Ferrer et al.

2003;Frattali et al. 2000;Graham et al. 2003a;Kertesz et al. 2000b;McMonagle et al.

2006;Murray et al. 2007]. One patient had a Wernicke‟s aphasia, which has previously been observed [McMonagle et al. 2006], while another patient had a conduction aphasia with reasonably intact comprehension and fluency, but impaired repetition. Graham et al. [Graham et al. 2003a] found that phonologic processing was impaired in a series of 10 unselected CBS patients. Only two of their patients had a full syndrome of progressive non-fluent aphasia. In a review of the literature that included 399 patients with CBS, 34% had aphasia, and, of 39 patients with sufficient language characterization to allow for stratification into different aphasic groups, 56% of these patients had a non-fluent presentation [Graham et al. 2003b]. They proposed that patients with early troubles in phonologic processing may represent part of the same spectrum with PNFA being at the more severe extreme.

6.1.3 Neuropsychiatric features

The majority of CBS patients in our sample experienced neuropsychiatric symptoms. Apathy was both the most frequent and severe symptom in CBS patients followed by abnormal appetite/eating behaviour and depressive symptoms. Litvan et al [Litvan et al. 1998] used the

NPI 10-item version, which does not assess for abnormal appetite/eating behaviour or aberrant night-time behaviour, to compare behavioural symptoms in 15 CBS and 34 PSP patients. In

176 contrast to our results, they found that depressive symptoms in CBS were more common than apathy, the latter being more severe in PSP [Litvan et al. 1998]. Their CBS sample was smaller than ours, which may have accounted for the difference seen between our sample and theirs.

Alternatively, it may reflect differences between the two groups in the caregivers‟ interpretation of observed patient signs. Irrespective of these inter-group differences, both depression and apathy have been shown to localize to overlapping regions of the limbic-prefrontal circuit, which are affected in CBS.

To our knowledge, there have been no studies which directly correlate apathy or depressive symptoms to specific brain areas in CBS. Dorsolateral prefrontal and anterior cingulate regions have been shown to be hypometabolic in FDG-PET studies of primary depression [Liotti and

Mayberg 2001] and hypoperfused in ECD-SPECT studies of depressive symptoms associated with AD [Levy-Cooperman et al. 2008a]. With respect to apathy, MRI studies have demonstrated that atrophy in the dorsolateral prefrontal cortex and anterior cingulate gyrus was associated with apathy in FTD [Zamboni et al. 2008]. Similarly, in Alzheimer‟s disease, apathy was associated with reduced blood flow on SPECT in the anterior cingulate gyrus and orbitofrontal regions [Lanctot et al. 2007]. Therefore, it is likely that these regions also play a role in mediating apathy and depressive symptoms in CBS.

Similar to our results, Litvan et al. [Litvan et al. 1998] also found that irritability and agitation were also relatively frequent in CBS. Symptoms of anxiety and disinhibition were more frequent in our CBS patients than theirs; this difference likely occurred by chance due to small sample

177 sizes. A retrospective review of 36 pathologically-proven cases of CBD found that eight of these patients had well-documented neuropsychiatric problems including: behavioural dyscontrol, depression, compulsive behaviour, irritability and disinhibition [Geda et al. 2007]. The majority of these eight patients had clinical diagnoses occurring along the spectrum of FTD, while only two patients clinically had CBS [Geda et al. 2007]. The symptoms were identified retrospectively and there was no use of formal psychometric measures to detect symptoms. Both of these factors likely account for the lower occurrence of neuropsychiatric symptoms in this study from the Mayo clinic compared to ours. However, we cannot exclude that the specificity of identifying true CBD pathology may also be contributing to this discrepant finding between samples. Other studies have confirmed the occurrence of frontal behaviours in CBS [Borroni et al. 2009;Kertesz et al. 2000b]. None of our prospectively ascertained CBS patients experienced visual hallucinations similar to findings of a retrospective review of 36 pathologically-proven cases of CBD [Geda et al. 2007] and extending on the findings of a prospective study of 11 patients with CBS [Cooper and Josephs 2009] in a larger sample.

6. 2 Apraxia in CBS

As discussed in chapter 4, there have been inconsistent findings with respect to subtypes of limb apraxia observed in CBS. Ideomotor apraxia is the most commonly documented apraxia with all patients developing this at some point during the course of the disease followed by limb-kinetic apraxia, although many studies did not examine specifically for this subtype. However, the occurrence of conceptual/ ideational apraxia appears to be highly variable with several studies not demonstrating this phenomenon, while others demonstrated it in 30 to 60% of the CBS

178 patients studied [Kertesz et al. 2000b;Leiguarda et al. 1994;Spatt et al. 2002]. There are several reasons why conceptual/ ideational apraxia appears to be so variable in CBS and important reasons include the way that this subtype is defined and also the fact that different rating instruments have been used to assess for this phenomenon. In addition, one predicts that there would be inconsistency in our ability to map conceptual/ ideational apraxia to specific brain regions based on this variable nosology. However, other difficulties with characterizing this type of apraxia may be even more fundamental in nature as discussed below.

Phenomenological confusion obviously exists because of the fact that „artificial‟ neuropsychological constructs, such as conceptual/ ideational apraxia, have been synthesized by neuropsychologists, cognitive neurologists and behavioural scientists somewhat independently in order explain observed behaviours. The reality is that praxis, especially relating to tool use, has slowly evolved over time in response to evolutionary forces to which humans (and their primitive ancestors) have been subjected. Therefore, definitions that we create and simplistic models of clinical localization to discrete brain regions are likely to be inadequate in the study of brain-behaviour relationships [Masellis et al. 2010], as emerging functional connectivity methods are now revealing [Greicius et al. 2004].

Notwithstanding these issues, many questions remain unanswered with respect to limb praxis in general and apraxia in neurodegenerative disease: Why does praxis appear to reside mainly in the dominant hemisphere? What evolutionary forces (environmental factors) combined with individual heritability (genes) caused the phenomenon to localize there? What disease specific

179 factors cause apraxia to manifest? Questions relating to the latter include: what specific hemispheric, cortical and subcortical lobar involvement predisposes to the different subtypes of apraxia and what are the specific effects of the underlying neurodegenerative pathology in determining apraxia? Answers to these questions will have to come from well-designed prospective and longitudinal neuroimaging and neuropsychological studies of corticobasal syndrome followed to death with subsequent histopathological, genomic and epigenomic analyses.

6.3 Comment on the neuroimaging methods

The current study is unique in that SPECT imaging was analyzed using two different approaches.

This included the unbiased, „top-down‟ SPM method, which is more stringent, in addition to the hypothesis-driven, „bottom-up‟ ROI method, which examines mean perfusion differences referenced to the cerebellum across select brain regions hypothesized to be involved in the pathophysiology of disease. Although both methods often identify the same regions affected in the between group comparisons, occasionally there was no overlap seen. This is because the methods look at the data in different ways. Statistical parametric maps are image processes based on voxel values relating to intensity in the case of SPECT, that, under the null hypothesis, fall under a known distribution such as the Student‟s t- or F distributions

(http://www.fil.ion.ucl.ac.uk/spm/doc/intro/). Multivariate analysis of covariance is performed on a voxel-by-voxel basis using the general linear model and Gaussian Random Field theory to make inferences about the spatially extended data (http://www.fil.ion.ucl.ac.uk/spm/doc/intro/).

Voxels of altered intensity that cluster together and survive correction for multiple testing

180 identify areas of altered perfusion between or within groups depending on the type of association being performed. Contrast this to the ROI method that looks at the mean perfusion in a defined region of interest (ROI) referenced to the cerebellum. The number of voxels within that region is pre-defined and the ROIs are based on a template map that is usually traced on a structural MRI.

The use of the ROI method may facilitate the identification of potentially „false negative‟ areas

(type II errors) of reduced perfusion missed by the stringent SPM analysis, while at the same time the SPM method may either identify important smaller regions of reduced perfusion

„washed out‟ in the larger ROI or alternatively emphasize the most strongly hypoperfused regions in the CBS group. In this way, the methods are used in synergy to capture the most salient regions of reduced perfusion in CBS.

As described in the introduction, two main functional neuroimaging techniques, PET and

SPECT, have been used to characterize CBS from the perspective of alterations in cerebral metabolism and perfusion, respectively. A brief discussion of these two nuclear medicine imaging modalities will now be provided in order to contrast strengths and weaknesses of both techniques. The literature will also be reviewed in terms of how they compare with respect to ability to assist in the accurate diagnosis of dementia. The focus of the discussion will be on

FDG-PET compared to HMPAO- or ECD-SPECT.

The basic underlying principles of PET and SPECT neuroimaging are similar: a radionuclide- labeled tracer is given intravenously and is taken up by the brain and based on its kinetic properties (absorption, distribution and metabolism) combined with the decay of the radionuclide

181 and the detection of the latter, images of the brain can be obtained that demonstrate relative distribution of the tracer in different cerebral regions. For the most part, cerebral glucose metabolism parallels cerebral perfusion, but occasionally this relationship breaks down in the presence of cerebrovascular disease [Silverman 2004]. The most commonly used PET tracer is

18F-fluorodeoxyglucose (FDG), which serves as a marker of cerebral glucose metabolism [Bailey et al. 2005;Frackowiak and Friston 1994]. The radioactively-labeled 18F isotope is synthesized in a cyclotron by accelerating protons into the nuclei of fluorine atoms. This results in fluorine with an extra proton in the nucleus producing an unstable isotope, 18F, with a relatively short half-life. The 18F isotope is then incorporated into the deoxyglucose molecule forming FDG.

FDG is then administered to the subject intravenously and gets distributed regionally in the brain reflecting cerebral glucose metabolism. After a delay to ensure the appropriate uptake of FDG in the brain, the subject is placed in a PET scanner, which is comprised of arrays of gamma ray detectors that encircle the subject‟s head. As the isotope undergoes positron emission decay, the emitted positron travels a very short distance usually in the range of millimeters dissipating energy until it encounters an electron. The interaction between the positron and the electron annihilate each other and results in two gamma photons that travel in opposite directions from each other. The gamma detectors that encircle the subject‟s head are set up in such a way as to only detect coincident gamma photons, that are, ones detected simultaneously by two detectors oriented directly across from each other. The scanner, therefore, detects the site of the annihilation event and the distance between this and the emitting nucleus limits the spatial resolution of the technique. During the scanning process, multiple detections are obtained all oriented along lines each in one plane or slice and from these reconstructed images of the three- dimensional brain can be derived.

182

Brain SPECT has some important fundamental differences that result in practical advantages and important limitations of this technique compared to PET [Rahmim and Zaidi 2008]. SPECT also uses radionuclide tracers to study CBS and other dementias with the most common ones being,

Tc99-HMPAO and -ECD, and these are taken up at a rate that is proportional to cerebral blood flow [Matsuda et al. 1995]. Therefore, SPECT provides a measure of regional cerebral perfusion to the brain as opposed to glucose metabolism. SPECT tracers also emit gamma radiation.

However, the gamma photon is detected directly by a camera comprised of a series of physical collimators lying over detection crystals and photomultiplier arrays. The collimators reject photons that are not within a small angular range thereby facilitating the localization of the origin of the gamma ray [Rahmim and Zaidi 2008]. The gamma camera usually rotates 360 degrees around the patient‟s head allowing for two dimensional images that can then be reconstructed into a three-dimensional view of the brain. Because the camera only detects one gamma photon for every emission event, the spatial resolution of SPECT is much lower than that of PET and this is a relative weakness of the technique [Rahmim and Zaidi 2008]. On the other hand, the main advantage of this technique has to do with cost and availability. This is because the radionuclide tracers typically used are more stable in terms of their gamma decay and can therefore be synthesized off-site and transported to the imaging centre. In other words, there is no need for a cyclotron on site to synthesize the compounds. In addition, the cameras used are typically less expensive than the PET scanner technology. As a result of these factors, SPECT imaging can be acquired more quickly, is significantly cheaper and is more widely available

[Colloby and O'Brien 2004].

183

There have been a few studies that have compared SPECT and PET in the same group of patients with Alzheimer‟s disease in order to determine the ability of these investigations to improve diagnostic accuracy [Silverman 2004]. In general, the degree of hypometabolism detected by

FDG-PET is usually greater than the magnitude of hypoperfusion abnormalities seen with perfusion SPECT although the regions of deficit observed by both methods are similar for AD

[Silverman 2004]. Side-by-side FDG-PET and perfusion SPECT studies of AD have demonstrated higher sensitivity and diagnostic accuracy of FDG-PET [Silverman 2004]. Studies using high-resolution SPECT cameras have shown that perfusion SPECT has 15 to 20% reduced accuracy in the diagnosis of AD compared to FDG-PET [Mielke and Heiss 1998]. An important study that compared 26 patients with AD to six healthy controls using both HMPAO-SPECT and

FDG-PET data in a voxel-wise analysis found that the correlation across the whole brain using both methods achieved statistical significance, but the strength of the association was weak

(average correlation coefficient [r] across all patients = 0.43) [Herholz et al. 2002]. However, this correlation improved substantially when the analysis was restricted to clusters of abnormal voxels in the temporoparietal and the posterior cingulate association cortices (r=0.90) [Herholz et al. 2002]. Despite this improved correlation, the tracer uptake reductions using FDG-PET were substantially more pronounced than that observed with HMPAO-SPECT indicating the former to be more sensitive in detecting abnormalities [Herholz et al. 2002]. In addition, although tracer uptake reductions correlated with severity of the dementia using both methods, the correlation was stronger for PET compared to SPECT [Herholz et al. 2002]. Finally, this study also showed that distinction of AD patients from controls was better over a wider range of z-thresholds for

FDG-PET than HMPAO-SPECT indicating increased sensitivity of the former [Herholz et al.

2002]. A more recent study compared the sensitivities of mesiotemporal atrophy on MRI,

184 reduced perfusion/hypometabolism in temporoparietal and posterior cingulate cortices on ECD-

SPECT/FDG-PET, respectively, and CSF biomarkers of beta-amyloid 1-42, total tau and phosphorylated tau in 207 AD patients of varying severity [Morinaga et al. 2010].

Mesiotemporal atrophy on MRI was identified in 77.4%, reduced perfusion and metabolism in the pre-defined regions in 81.6% (ECD-SPECT) and 93.1% (FDG-PET), and the typical CSF profile for AD (that is, reduced beta-amyloid and increased total and phosphorylated tau) in 94% of all the AD patients [Morinaga et al. 2010]. Using the Clinical Dementia Rating (CDR) scale, used to assess severity of dementia with higher scores being more severe, they observed that all investigations were sensitive at a CDR of 2, whereas at a CDR of 1 only the FDG-PET and CSF biomarkers showed high sensitivity [Morinaga et al. 2010]. Finally at the mildest stages of disease (CDR of 0.5), usually corresponding to cases of amnestic mild cognitive impairment, only CSF biomarkers showed high sensitivity [Morinaga et al. 2010]. Limitations of this study were that there was no control group so a discriminant function analysis could not be done, and also that cases were not pathologically confirmed.

6.4 Can CBS serve as a model of etiology for common sporadic disorders?

Although CBS is a rare syndrome, it can serve as a good model of complex disease due to a variety of observations. As discussed throughout this thesis, there is substantial pathological heterogeneity that can produce the syndrome and this can make it difficult to identify genetic and environmental factors that increase risk for the disease in clinically diagnosed cases. As shown by recent studies in FTD and PSP, increasing the homogeneity of the sample by inclusion of only one specific pathological subtype, for example, recent genome wide association studies (GWAS)

185 of FTLD-U/ TDP43 and PSP pathological cases, makes it easier to identify genetic risk factors for the disease [Hoglinger et al. 2011;Van Deerlin et al. 2010]. In CBS, the pathological heterogeneity can account for the multiple disease susceptibility loci observed (i.e., genetic heterogeneity), for example, MAPT and PGRN mutations. However, within patients having a pathological diagnosis of CBD, there are likely other disease susceptibility loci that have not yet been discovered that can produce the typical pathology. The reason for this may be that different pathways in the processing of the MAPT gene and the tau protein might yield the final common pathology of CBD and therefore a systems biology approach will be helpful in sorting out the genetic heterogeneity of the disease [Noorbakhsh et al. 2009]. Genetic modifiers may also contribute to the heterogeneity of the CBS as demonstrated by a study identifying that common variants in the TMEM106B gene can increase risk for FTLD-U/ TDP43 pathology even within those harbouring PGRN mutations [Van Deerlin et al. 2010]. Clinical phenotypic variability is due mainly to the area of the brain most affected by the underlying pathology and this may also lead to misdiagnosis, which will further confound genetic studies of CBS. Finally, using PRGN mutation as an example, the age of onset of individuals with FTD and/ or CBS due to PGRN mutation is highly variable with some not developing disease until their late 80s [Kelley et al.

2009] and this may result in apparent incomplete penetrance of the disease-causing gene mutation. All of these features of CBS can be observed in complex, non-Mendelian diseases.

Common neurodegenerative diseases, such as Alzheimer‟s and Parkinson‟s disease, are generally considered sporadic disorders although rarely Mendelian segregation within families is observed due to identified disease-causing mutations. This is similar to that observed in CBS and given the rarity of CBS as a clinical entity, one can assume that there are even more numerous pathways

186 leading to AD and PD pathology and that these routes are even more complex in nature.

Recently, through the application of GWAS in very large, multicentre case-control cohorts, several genes for many of the common neurodegenerative disorders have been identified, each of which confers an incremental risk to individuals possessing the gene variants [2011;Do et al.

2011;Hollingworth et al. 2011;Naj et al. 2011;Nalls et al. 2011]. The genes include ones already known to be involved in the underlying pathology of these disorders, for example, apolipoprotein

E (APOE) for Alzheimer‟s disease and alpha-synuclein (SNCA) for Parkinson‟s disease, as well as many novel genes. The exact role that these genes play and the regional effects that their variants have on the brain and secondarily on neurological and cognitive functions are mostly unknown. Based on what we have learned about CBS, it is hypothesized that certain variants, alone or in combination, strongly increase or decrease risk for these common neurodegenerations in smaller subsets of the large GWAS samples rather than conferring a small risk to the entire patient sample, that is, genetic heterogeneity. Other mechanisms of disease within these identified candidate genes flagged by the GWAS, such as epigenomic factors, for example, DNA methylation, must also be considered. In order to sort out the complexities of the neurodegenerations in the post-genomics era, novel approaches such as “reverse phenotyping” will be required in order to understand the regional brain effects that these genomic and epigenomic factors have [Joober 2011;Schulze and McMahon 2004].

187

7.0 Conclusions and future directions

188

7.0 Conclusions and future directions

Corticobasal syndrome (CBS) is a unique cognitive and motor disorder. Better understanding of this rare clinical entity with respect to etiology, clinical and neuroimaging features will provide new insights that should improve our research approach to other more common, complex neurodegenerative disorders, such as AD and PD. This thesis work has demonstrated genetic and pathological heterogeneity underlying the CBS with discovery that PGRN mutation – producing

FTLD-U/ TDP43 pathology – can manifest as a fairly classical picture of the syndrome from the clinical, cognitive, and neuroimaging perspective. It underscores the importance of obtaining brain tissue for histopathological study as identification of the underlying pathology will facilitate subgrouping of clinical cases based on pathology so that heterogeneity is reduced in the attempts to discover underlying genetic factors.

Furthermore, the well-characterized clinical sample of CBS patients used in this study has allowed us to expand our understanding of the different phenotypic presentations of the CBS and associated clinical, cognitive and neuroimaging features. Those presenting with early dementia tended to have the right side of the body most affected and also had more significant language dysfunction indicating that the burden of underlying pathology was most pronounced in the left hemisphere although the SPECT and MRI studies did not confirm this. Apart from issues related to reduced power, we hypothesize that the reason leftwards asymmetry on MRI and SPECT was not seen may be because, at the stage that the patients were studied, the microscopic pathology that might have been asymmetrically more pronounced on the left did not appear to translate into changes that could be detected in vivo using neuroimaging. The contrary was seen in the early motor group who tended to have the left side of the body most affected with significant reductions in perfusion involving the right supplementary and premotor areas. 189

We have also shown that even within the same family, there can be substantial heterogeneity in clinical presentation due to the same PGRN mutation. The reason for this is that although the predicted pathology in the brain will be the same in the affected family members, the hemispheric and lobar localization of the pathology and its severity will determine the observed clinical syndrome. Therefore, studies that focus on attempting to discern early imaging and other biomarkers of disease that can predict a specific underlying pathological type in vivo will be challenged by this heterogeneity in clinical presentation. We propose that future studies attempting to look for early biological risk factors of CBS will need to examine subgroups of

CBS patients that are classified according to genetic or pathological type, as well as classified according to in vivo hemispheric and lobar localization on neuroimaging, or even cognitive endophenotypes, comparing “OMIC” measures across these subgroups. Since the sample size within each subgroup will likely be small, multicentre studies will be necessary in order to enhance the sample and thereby power of this approach.

Brain-behavioural correlative studies in CBS are an important line of research that can further enhance our understanding of the phenomenon observed in the syndrome. This is exemplified by our study examining severity of ideomotor apraxia and its association with hypoperfusion in the left inferior parietal lobule. The inconsistencies across studies in terms of localization of limb apraxia in CBS stem largely from differences in the tools used to assess this phenomenon and also from the different nosologies used to describe the subtypes of apraxia to date. Investigators working in this field will have to decide upon accepted diagnostic tools and definitions such that

190 studies are performed in a consistent fashion. We anticipate that this will resolve many of the inconsistencies in the apraxia literature.

As our CBS sample size increases, we anticipate completing multi-modal neuroimaging analyses including MRI brain with assessment of both grey matter atrophy and quantification of underlying white matter disease using our on-site developed program, Lesion Explorer [Ramirez et al. 2011], diffusion tensor imaging in a subset of patients, combined with brain SPECT perfusion. We also plan on correlating this imaging dataset with a comprehensive assessment of apraxia using a conceptual model in order to better identify the neuroanatomical correlates of the different apraxia subtypes [Stamenova et al. 2011]. Finally, in a subset of the CBS cases, longitudinal neuropsychological and neuroimaging data is available and further analysis of this data will allow us to track the progression of the disease in terms of the most affected brain regions and how these correlate with the neuropsychological and pathological measures over time. Only this kind of longitudinal study will allow us to develop better in vivo biomarkers of underlying pathology.

191

8.0 References

192

8.0 References

A two-stage meta-analysis identifies several new loci for Parkinson's disease. PLoS Genet 2011; 7: e1002142.

Ahmed Z, Mackenzie IR, Hutton ML, Dickson DW. Progranulin in frontotemporal lobar degeneration and neuroinflammation. J Neuroinflammation 2007; 4: 7.

Alexopoulos GS, Abrams RC, Young RC, Shamoian CA. Cornell Scale for Depression in Dementia. Biol Psychiatry 1988; 23: 271-284.

Andreadis A, Brown WM, Kosik KS. Structure and novel exons of the human tau gene. Biochemistry 1992; 31: 10626-10633.

Bailey DL, Townsend DW, Valk PE, Maisey MN. Positron Emission Tomography: Basic Sciences. Secaucus, NJ: Springer-Verlag; 2005.

Bak TH, Caine D, Hearn VC, Hodges JR. Visuospatial functions in atypical parkinsonian syndromes. J Neurol Neurosurg Psychiatry 2006; 77: 454-456.

Baker KE, Parker R. Nonsense-mediated mRNA decay: terminating erroneous gene expression. Curr Opin Cell Biol 2004; 16: 293-299.

Baker M, Mackenzie IR, Pickering-Brown SM et al. Mutations in progranulin cause tau-negative frontotemporal dementia linked to chromosome 17. Nature 2006.

Ball JA, Lantos PL, Jackson M, Marsden CD, Scadding JW, Rossor MN. Alien hand sign in association with Alzheimer's histopathology. J Neurol Neurosurg Psychiatry 1993; 56: 1020- 1023.

Beck J, Rohrer JD, Campbell T et al. A distinct clinical, neuropsychological and radiological phenotype is associated with progranulin gene mutations in a large UK series. Brain 2008; 131: 706-720.

Benussi L, Binetti G, Sina E et al. A novel deletion in progranulin gene is associated with FTDP- 17 and CBS. Neurobiol Aging 2008; 29: 427-435.

Benussi L, Ghidoni R, Pegoiani E, Moretti DV, Zanetti O, Binetti G. Progranulin Leu271LeufsX10 is one of the most common FTLD and CBS associated mutations worldwide. Neurobiol Dis 2009; 33: 379-385.

Bergeron C, Davis A, Lang AE. Corticobasal ganglionic degeneration and progressive supranuclear palsy presenting with cognitive decline. Brain Pathol 1998; 8: 355-365.

Berget SM. Exon recognition in vertebrate splicing. J Biol Chem 1995; 270: 2411-2414.

193

Blair JR, Spreen O. Predicting premorbid IQ: A revision of the National Adult Reading Test. Clinical Neuropsychologist 1989; 3: 129-136.

Blin J, Vidailhet MJ, Pillon B, Dubois B, Feve JR, Agid Y. Corticobasal degeneration: decreased and asymmetrical glucose consumption as studied with PET. Mov Disord 1992; 7: 348-354.

Boelmans K, Bodammer NC, Suchorska B et al. Diffusion tensor imaging of the corpus callosum differentiates corticobasal syndrome from Parkinson's disease. Parkinsonism Relat Disord 2010; 16: 498-502.

Boelmans K, Kaufmann J, Bodammer N et al. Involvement of motor pathways in corticobasal syndrome detected by diffusion tensor tractography. Mov Disord 2009; 24: 168-175.

Boeve BF, Hutton M. Refining frontotemporal dementia with parkinsonism linked to chromosome 17: introducing FTDP-17 (MAPT) and FTDP-17 (PGRN). Arch Neurol 2008; 65: 460-464.

Boeve BF, Lang AE, Litvan I. Corticobasal degeneration and its relationship to progressive supranuclear palsy and frontotemporal dementia. Ann Neurol 2003; 54 Suppl 5: S15-S19.

Boeve BF, Maraganore DM, Parisi JE et al. Pathologic heterogeneity in clinically diagnosed corticobasal degeneration. Neurology 1999; 53: 795-800.

Boeve BF, Maraganore DM, Parisi JE et al. Corticobasal degeneration and frontotemporal dementia presentations in a kindred with nonspecific histopathology. Dement Geriatr Cogn Disord 2002; 13: 80-90.

Borroni B, Alberici A, Agosti C, Cosseddu M, Padovani A. Pattern of behavioral disturbances in corticobasal degeneration syndrome and progressive supranuclear palsy. Int Psychogeriatr 2009; 21: 463-468.

Borroni B, Archetti S, Alberici A et al. Progranulin genetic variations in frontotemporal lobar degeneration: evidence for low mutation frequency in an Italian clinical series. Neurogenetics 2008a; 9: 197-205.

Borroni B, Garibotto V, Agosti C et al. White matter changes in corticobasal degeneration syndrome and correlation with limb apraxia. Arch Neurol 2008b; 65: 796-801.

Boxer AL, Geschwind MD, Belfor N et al. Patterns of brain atrophy that differentiate corticobasal degeneration syndrome from progressive supranuclear palsy. Arch Neurol 2006; 63: 81-86.

Bozzali M, Cercignani M, Baglio F et al. Voxel-wise analysis of diffusion tensor MRI improves the confidence of diagnosis of corticobasal degeneration non-invasively. Parkinsonism Relat Disord 2008; 14: 436-439.

Brouwers N, Sleegers K, Engelborghs S et al. Genetic variability in progranulin contributes to risk for clinically diagnosed Alzheimer disease. Neurology 2008; 71: 656-664.

194

Brown J, Lantos PL, Roques P, Fidani L, Rossor MN. Familial dementia with swollen achromatic neurons and corticobasal inclusion bodies: a clinical and pathological study. J Neurol Sci 1996; 135: 21-30.

Brown J, Lantos PL, Rossor MN. Familial dementia lacking specific pathological features presenting with clinical features of corticobasal degeneration. J Neurol Neurosurg Psychiatry 1998; 65: 600-603.

Bryden MP. Measuring handedness with questionnaires. Neuropsychologia 1977; 15: 617-624.

Bugiani O, Murrell JR, Giaccone G et al. Frontotemporal dementia and corticobasal degeneration in a family with a P301S mutation in tau. J Neuropathol Exp Neurol 1999; 58: 667- 677.

Buxbaum LJ, Kyle K, Grossman M, Coslett HB. Left inferior parietal representations for skilled hand-object interactions: evidence from stroke and corticobasal degeneration. Cortex 2007; 43: 411-423.

Caffrey TM, Wade-Martins R. Functional MAPT haplotypes: bridging the gap between genotype and neuropathology. Neurobiol Dis 2007; 27: 1-10.

Cairns NJ, Bigio EH, Mackenzie IR et al. Neuropathologic diagnostic and nosologic criteria for frontotemporal lobar degeneration: consensus of the Consortium for Frontotemporal Lobar Degeneration. Acta Neuropathol 2007; 114: 5-22.

Cannon A, Baker M, Boeve B et al. CHMP2B mutations are not a common cause of frontotemporal lobar degeneration. Neurosci Lett 2006; 398: 83-84.

Casseron W, Azulay JP, Guedj E, Gastaut JL, Pouget J. Familial autosomal dominant corticobasal degeneration with the P301S mutation in the tau gene: an example of phenotype variability. J Neurol 2005; 252: 1546-1548.

Chainay H, Humphreys GW. Ideomotor and ideational apraxia in corticobasal degeneration: a case study. Neurocase 2003; 9: 177-186.

Chang KH, Chuang CC, Lyu RK et al. Clinical characteristics of corticobasal syndrome amongst Chinese in Taiwan. Parkinsonism Relat Disord 2007; 13: 219-223.

Chen-Plotkin AS, Yuan W, Anderson C et al. Corticobasal syndrome and primary progressive aphasia as manifestations of LRRK2 gene mutations. Neurology 2008; 70: 521-527.

Cilia R, Rossi C, Frosini D et al. Dopamine Transporter SPECT Imaging in Corticobasal Syndrome. PLoS One 2011; 6: e18301.

Cohen L, Dehaene S, Naccache L et al. The visual word form area: spatial and temporal characterization of an initial stage of reading in normal subjects and posterior split-brain patients. Brain 2000; 123 ( Pt 2): 291-307.

195

Colloby S, O'Brien J. Functional imaging in Parkinson's disease and dementia with Lewy bodies. J Geriatr Psychiatry Neurol 2004; 17: 158-163.

Cooper AD, Josephs KA. Photophobia, visual hallucinations, and REM sleep behavior disorder in progressive supranuclear palsy and corticobasal degeneration: a prospective study. Parkinsonism Relat Disord 2009; 15: 59-61.

Cordivari C, Misra VP, Catania S, Lees AJ. Treatment of dystonic clenched fist with botulinum toxin. Mov Disord 2001; 16: 907-913.

Coulier IM, de Vries JJ, Leenders KL. Is FDG-PET a useful tool in clinical practice for diagnosing corticobasal ganglionic degeneration? Mov Disord 2003; 18: 1175-1178.

Crowe RR. Candidate genes in psychiatry: an epidemiological perspective. Am J Med Genet 1993; 48: 74-77.

Cruts M, Gijselinck I, van der ZJ et al. Null mutations in progranulin cause ubiquitin-positive frontotemporal dementia linked to chromosome 17q21. Nature 2006.

Cummings JL. The Neuropsychiatric Inventory: assessing psychopathology in dementia patients. Neurology 1997; 48: S10-S16.

Darvesh S, Leach L, Black SE, Kaplan E, Freedman M. The behavioural neurology assessment. Can J Neurol Sci 2005; 32: 167-177.

Delis DC, Kramer JH, Kaplan E, Ober BA. California Verbal Learning Test: Adult Version. San Antonio: Psychological Corporation; 1987.

Denny-Brown D, Meyer JS, Horenstein S. The significance of perceptual rivalry resulting from parietal lesion. Brain 1952; 75: 433-471.

Desgranges B, Baron JC, de lS, V et al. The neural substrates of memory systems impairment in Alzheimer's disease. A PET study of resting brain glucose utilization. Brain 1998; 121 ( Pt 4): 611-631.

Di Maria E., Tabaton M, Vigo T et al. Corticobasal degeneration shares a common genetic background with progressive supranuclear palsy. Ann Neurol 2000; 47: 374-377.

Dickson DW, Bergeron C, Chin SS et al. Office of Rare Diseases neuropathologic criteria for corticobasal degeneration. J Neuropathol Exp Neurol 2002; 61: 935-946.

Do CB, Tung JY, Dorfman E et al. Web-based genome-wide association study identifies two novel loci and a substantial genetic component for Parkinson's disease. PLoS Genet 2011; 7: e1002141.

Eidelberg D, Dhawan V, Moeller JR et al. The metabolic landscape of cortico-basal ganglionic degeneration: regional asymmetries studied with positron emission tomography. J Neurol Neurosurg Psychiatry 1991; 54: 856-862.

196

Erbetta A, Mandelli ML, Savoiardo M et al. Diffusion tensor imaging shows different topographic involvement of the thalamus in progressive supranuclear palsy and corticobasal degeneration. AJNR Am J Neuroradiol 2009; 30: 1482-1487.

Ferrer I, Hernandez I, Boada M et al. Primary progressive aphasia as the initial manifestation of corticobasal degeneration and unusual tauopathies. Acta Neuropathol 2003; 106: 419-435.

Finch N, Baker M, Crook R et al. Plasma progranulin levels predict progranulin mutation status in frontotemporal dementia patients and asymptomatic family members. Brain 2009; 132: 583- 591.

Folstein MF, Folstein SE, McHugh PR. "Mini-mental state". A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: 189-198.

Forman MS, Zhukareva V, Bergeron C et al. Signature tau neuropathology in gray and white matter of corticobasal degeneration. Am J Pathol 2002; 160: 2045-2053.

Frackowiak RS, Friston KJ. Functional neuroanatomy of the human brain: positron emission tomography--a new neuroanatomical technique. J Anat 1994; 184 ( Pt 2): 211-225.

Frattali CM, Grafman J, Patronas N, Makhlouf F, Litvan I. Language disturbances in corticobasal degeneration. Neurology 2000; 54: 990-992.

Frisoni GB, Pizzolato G, Zanetti O, Bianchetti A, Chierichetti F, Trabucchi M. Corticobasal degeneration: neuropsychological assessment and dopamine D2 receptor SPECT analysis. Eur Neurol 1995; 35: 50-54.

Gabryelewicz T, Masellis M, Berdynski M et al. Intra-familial clinical heterogeneity due to FTLD-U with TDP-43 proteinopathy caused by a novel deletion in progranulin gene (PGRN). J Alzheimers Dis 2010; 22: 1123-1133.

Gallien P, Verin M, Rancurel G, De Marco O, Eden G. First familial cases of corticobasal degeneration. 1998. p. A428.

Garraux G, Salmon E, Peigneux P et al. Voxel-based distribution of metabolic impairment in corticobasal degeneration. Mov Disord 2000; 15: 894-904.

Gass J, Cannon A, Mackenzie IR et al. Mutations in progranulin are a major cause of ubiquitin- positive frontotemporal lobar degeneration. Hum Mol Genet 2006; 15: 2988-3001.

Geda YE, Boeve BF, Negash S et al. Neuropsychiatric features in 36 pathologically confirmed cases of corticobasal degeneration. J Neuropsychiatry Clin Neurosci 2007; 19: 77-80.

Gelinas I, Gauthier L, McIntyre M, Gauthier S. Development of a functional measure for persons with Alzheimer's disease: the disability assessment for dementia. Am J Occup Ther 1999; 53: 471-481.

197

Genovese CR, Lazar NA, Nichols T. Thresholding of statistical maps in functional neuroimaging using the false discovery rate. Neuroimage 2002; 15: 870-878.

Geschwind N. The apraxias: neural mechanisms of disorders of learned movement. Am Sci 1975; 63: 188-195.

Ghetti B, Spina S, Murrell JR et al. In vivo and postmortem clinicoanatomical correlations in frontotemporal dementia and parkinsonism linked to chromosome 17. Neurodegener Dis 2008; 5: 215-217.

Gibb WR, Luthert PJ, Marsden CD. Corticobasal degeneration. Brain 1989; 112 ( Pt 5): 1171- 1192.

Gladsjo JA, Schuman CC, Evans JD, Peavy GM, Miller SW, Heaton RK. Norms for letter and category fluency: demographic corrections for age, education, and ethnicity. Assessment 1999; 6: 147-178.

Goedert M, Ghetti B, Spillantini MG. Tau gene mutations in frontotemporal dementia and parkinsonism linked to chromosome 17 (FTDP-17). Their relevance for understanding the neurogenerative process. Ann N Y Acad Sci 2000; 920: 74-83.

Goedert M, Spillantini MG, Potier MC, Ulrich J, Crowther RA. Cloning and sequencing of the cDNA encoding an isoform of microtubule-associated protein tau containing four tandem repeats: differential expression of tau protein mRNAs in human brain. EMBO J 1989; 8: 393- 399.

Goedert M, Wischik CM, Crowther RA, Walker JE, Klug A. Cloning and sequencing of the cDNA encoding a core protein of the paired helical filament of Alzheimer disease: identification as the microtubule-associated protein tau. Proc Natl Acad Sci U S A 1988; 85: 4051-4055.

Goldenberg G. Apraxia and the parietal lobes. Neuropsychologia 2009; 47: 1449-1459.

Goldenberg G, Spatt J. The neural basis of tool use. Brain 2009; 132: 1645-1655.

Goodale MA, Milner AD. Separate visual pathways for perception and action. Trends Neurosci 1992; 15: 20-25.

Graham NL, Bak T, Patterson K, Hodges JR. Language function and dysfunction in corticobasal degeneration. Neurology 2003a; 61: 493-499.

Graham NL, Bak TH, Hodges JR. Corticobasal degeneration as a cognitive disorder. Mov Disord 2003b; 18: 1224-1232.

Graham NL, Zeman A, Young AW, Patterson K, Hodges JR. Dyspraxia in a patient with corticobasal degeneration: the role of visual and tactile inputs to action. J Neurol Neurosurg Psychiatry 1999; 67: 334-344.

198

Gregory CA, Serra-Mestres J, Hodges JR. Early diagnosis of the frontal variant of frontotemporal dementia: how sensitive are standard neuroimaging and neuropsychologic tests? Neuropsychiatry Neuropsychol Behav Neurol 1999; 12: 128-135.

Greicius MD, Srivastava G, Reiss AL, Menon V. Default-mode network activity distinguishes Alzheimer's disease from healthy aging: evidence from functional MRI. Proc Natl Acad Sci U S A 2004; 101: 4637-4642.

Grimes DA, Bergeron CB, Lang AE. Motor neuron disease-inclusion dementia presenting as cortical-basal ganglionic degeneration. Mov Disord 1999a; 14: 674-680.

Grimes DA, Lang AE, Bergeron CB. Dementia as the most common presentation of cortical- basal ganglionic degeneration. Neurology 1999b; 53: 1969-1974.

Groschel K, Hauser TK, Luft A et al. Magnetic resonance imaging-based volumetry differentiates progressive supranuclear palsy from corticobasal degeneration. Neuroimage 2004; 21: 714-724.

Gross RG, Grossman M. Update on apraxia. Curr Neurol Neurosci Rep 2008; 8: 490-496.

Grossman M, McMillan C, Moore P et al. What's in a name: voxel-based morphometric analyses of MRI and naming difficulty in Alzheimer's disease, frontotemporal dementia and corticobasal degeneration. Brain 2004; 127: 628-649.

Guerreiro RJ, Santana I, Bras JM et al. Novel progranulin mutation: screening for PGRN mutations in a Portuguese series of FTD/CBS cases. Mov Disord 2008; 23: 1269-1273.

Hassan A, Whitwell JL, Boeve BF et al. Symmetric corticobasal degeneration (S-CBD). Parkinsonism Relat Disord 2010; 16: 208-214.

Heaton RK. Wisconsin Card Sorting Manual. Odessa: Psychological Assessment Resources, Inc.; 1981.

Heilman KM. Apraxia. In: Heilman KM, Valenstein E, editors. Clinical Neuropsychology. New York: Oxford University Press; 1979.

Heilman KM, Rothi LJ, Valenstein E. Two forms of ideomotor apraxia. Neurology 1982; 32: 342-346.

Heilman KM, Rothi LJG. Apraxia. In: Heilman KM, Valenstein E, editors. Clinical Neuropsychology. New York, Oxford: Oxford University Press; 1993. p. 141-64.

Herholz K, Schopphoff H, Schmidt M et al. Direct comparison of spatially normalized PET and SPECT scans in Alzheimer's disease. J Nucl Med 2002; 43: 21-26.

Hodges JR, Spatt J, Patterson K. "What" and "how": evidence for the dissociation of object knowledge and mechanical problem-solving skills in the human brain. Proc Natl Acad Sci U S A 1999; 96: 9444-9448.

199

Hoglinger GU, Melhem NM, Dickson DW et al. Identification of common variants influencing risk of the tauopathy progressive supranuclear palsy. Nat Genet 2011; 43: 699-705.

Hollingworth P, Harold D, Sims R et al. Common variants at ABCA7, MS4A6A/MS4A4E, EPHA1, CD33 and CD2AP are associated with Alzheimer's disease. Nat Genet 2011; 43: 429- 435.

Hosaka K, Ishii K, Sakamoto S et al. Voxel-based comparison of regional cerebral glucose metabolism between PSP and corticobasal degeneration. J Neurol Sci 2002; 199: 67-71.

Hossain AK, Murata Y, Zhang L et al. Brain perfusion SPECT in patients with corticobasal degeneration: analysis using statistical parametric mapping. Mov Disord 2003; 18: 697-703.

Houlden H, Baker M, Morris HR et al. Corticobasal degeneration and progressive supranuclear palsy share a common tau haplotype. Neurology 2001; 56: 1702-1706.

Huey ED, Goveia EN, Paviol S et al. Executive dysfunction in frontotemporal dementia and corticobasal syndrome. Neurology 2009a; 72: 453-459.

Huey ED, Pardini M, Cavanagh A et al. Association of ideomotor apraxia with frontal gray matter volume loss in corticobasal syndrome. Arch Neurol 2009b; 66: 1274-1280.

Hutton M. Missense and splice site mutations in tau associated with FTDP-17: multiple pathogenic mechanisms. Neurology 2001; 56: S21-S25.

Hutton M, Lendon CL, Rizzu P et al. Association of missense and 5'-splice-site mutations in tau with the inherited dementia FTDP-17. Nature 1998; 393: 702-705.

Ince PG, Morris JC. Demystifying lobar degenerations: tauopathies vs Gehrigopathies. Neurology 2006; 66: 8-9.

Jacobs DH, Adair JC, Macauley B, Gold M, Gonzalez Rothi LJ, Heilman KM. Apraxia in corticobasal degeneration. Brain Cogn 1999; 40: 336-354.

Joober R. The 1000 Genomes Project: deep genomic sequencing waiting for deep psychiatric phenotyping. J Psychiatry Neurosci 2011; 36: 147-149.

Josephs KA, Ahmed Z, Katsuse O et al. Neuropathologic features of frontotemporal lobar degeneration with ubiquitin-positive inclusions with progranulin gene (PGRN) mutations. J Neuropathol Exp Neurol 2007; 66: 142-151.

Josephs KA, Duffy JR. Apraxia of speech and nonfluent aphasia: a new clinical marker for corticobasal degeneration and progressive supranuclear palsy. Curr Opin Neurol 2008; 21: 688- 692.

Josephs KA, Petersen RC, Knopman DS et al. Clinicopathologic analysis of frontotemporal and corticobasal degenerations and PSP. Neurology 2006; 66: 41-48.

200

Josephs KA, Tang-Wai DF, Edland SD et al. Correlation between antemortem magnetic resonance imaging findings and pathologically confirmed corticobasal degeneration. Arch Neurol 2004; 61: 1881-1884.

Josephs KA, Whitwell JL, Dickson DW et al. Voxel-based morphometry in autopsy proven PSP and CBD. Neurobiol Aging 2008; 29: 280-289.

Juh R, Pae CU, Kim TS, Lee CU, Choe B, Suh T. Cerebral glucose metabolism in corticobasal degeneration comparison with progressive supranuclear palsy using statistical mapping analysis. Neurosci Lett 2005; 383: 22-27.

Kas A, de Souza LC, Samri D et al. Neural correlates of cognitive impairment in posterior cortical atrophy. Brain 2011; 134: 1464-1478.

Kelley BJ, Haidar W, Boeve BF et al. Prominent phenotypic variability associated with mutations in Progranulin. Neurobiol Aging 2009; 30: 739-751.

Kertesz A. Pick Complex: an integrative approach to frontotemporal dementia: primary progressive aphasia, corticobasal degeneration, and progressive supranuclear palsy. Neurologist 2003; 9: 311-317.

Kertesz A. Frontotemporal dementia: one disease, or many?: probably one, possibly two. Alzheimer Dis Assoc Disord 2005; 19 Suppl 1: S19-S24.

Kertesz A. Western Aphasia Battery, Revised: Examiner's Manual. San Antonio: Harcourt Assessment, Inc.; 2007.

Kertesz A, Ferro JM. Lesion size and location in ideomotor apraxia. Brain 1984; 107 ( Pt 3): 921-933.

Kertesz A, Kawarai T, Rogaeva E et al. Familial frontotemporal dementia with ubiquitin- positive, tau-negative inclusions. Neurology 2000a; 54: 818-827.

Kertesz A, Martinez-Lage P, Davidson W, Munoz DG. The corticobasal degeneration syndrome overlaps progressive aphasia and frontotemporal dementia. Neurology 2000b; 55: 1368-1375.

Kertesz A, McMonagle P, Blair M, Davidson W, Munoz DG. The evolution and pathology of frontotemporal dementia. Brain 2005; 128: 1996-2005.

Kertesz A, Poole E. The aphasia quotient: the taxonomic approach to measurement of aphasic disability. Can J Neurol Sci 1974; 1: 7-16.

Kidd KK. Associations of disease with genetic markers: deja vu all over again. Am J Med Genet 1993; 48: 71-73.

Kim YD, Kim JS, Lee ES, Yang DW, Lee KS, Kim YI. Progressive "vascular" corticobasal syndrome due to bilateral ischemic hemispheric lesions. Intern Med 2009; 48: 1699-1702.

201

Klaffke S, Kuhn AA, Plotkin M et al. Dopamine transporters, D2 receptors, and glucose metabolism in corticobasal degeneration. Mov Disord 2006; 21: 1724-1727.

Kleiner-Fisman G, Bergeron C, Lang AE. Presentation of Creutzfeldt-Jakob disease as acute corticobasal degeneration syndrome. Mov Disord 2004; 19: 948-949.

Kleiner-Fisman G, Black SE, Lang AE. Neurodegenerative disease and the evolution of art: the effects of presumed corticobasal degeneration in a professional artist. Mov Disord 2003; 18: 294- 302.

Kleist K. Apraxie. Jarbuch Psychiatr Neurol 1907; 28: 46-112.

Knibb JA, Xuereb JH, Patterson K, Hodges JR. Clinical and pathological characterization of progressive aphasia. Ann Neurol 2006; 59: 156-165.

Kompoliti K, Goetz CG, Boeve BF et al. Clinical presentation and pharmacological therapy in corticobasal degeneration. Arch Neurol 1998; 55: 957-961.

Kovacevic N, Lobaugh NJ, Bronskill MJ, Levine B, Feinstein A, Black SE. A robust method for extraction and automatic segmentation of brain images. Neuroimage 2002; 17: 1087-1100.

Koyama M, Yagishita A, Nakata Y, Hayashi M, Bandoh M, Mizutani T. Imaging of corticobasal degeneration syndrome. Neuroradiology 2007; 49: 905-912.

Kreisler A, Defebvre L, Lecouffe P et al. Corticobasal degeneration and Parkinson's disease assessed by HmPaO SPECT: the utility of factorial discriminant analysis. Mov Disord 2005; 20: 1431-1438.

Kroliczak G, Frey SH. A common network in the left cerebral hemisphere represents planning of tool use pantomimes and familiar intransitive gestures at the hand-independent level. Cereb Cortex 2009; 19: 2396-2410.

Lacadie CM, Fulbright RK, Rajeevan N, Constable RT, Papademetris X. More accurate Talairach coordinates for neuroimaging using non-linear registration. Neuroimage 2008; 42: 717-725.

Lai SC, Weng YH, Yen TC et al. Imaging early-stage corticobasal degeneration with [99mTc]TRODAT-1 SPET. Nucl Med Commun 2004; 25: 339-345.

Lancaster JL, Woldorff MG, Parsons LM et al. Automated Talairach atlas labels for functional brain mapping. Hum Brain Mapp 2000; 10: 120-131.

Lanctot KL, Moosa S, Herrmann N et al. A SPECT study of apathy in Alzheimer's disease. Dement Geriatr Cogn Disord 2007; 24: 65-72.

Lander ES, Schork NJ. Genetic dissection of complex traits. Science 1994; 265: 2037-2048.

202

Lang AE. Corticobasal degeneration: selected developments. Mov Disord 2003; 18 Suppl 6: S51-S56.

Lang AE, Bergeron C, Pollanen MS, Ashby P. Parietal Pick's disease mimicking cortical-basal ganglionic degeneration. Neurology 1994; 44: 1436-1440.

Laureys S, Salmon E, Garraux G et al. Fluorodopa uptake and glucose metabolism in early stages of corticobasal degeneration. J Neurol 1999; 246: 1151-1158.

Le B, I, Camuzat A, Hannequin D et al. Phenotype variability in progranulin mutation carriers: a clinical, neuropsychological, imaging and genetic study. Brain 2008; 131: 732-746.

Lee SE, Rabinovici GD, Mayo MC et al. Clinicopathological correlations in corticobasal degeneration. Ann Neurol 2011; 70: 327-340.

Leiguarda R, Lees AJ, Merello M, Starkstein S, Marsden CD. The nature of apraxia in corticobasal degeneration. J Neurol Neurosurg Psychiatry 1994; 57: 455-459.

Leiguarda RC, Marsden CD. Limb apraxias: higher-order disorders of sensorimotor integration. Brain 2000; 123 ( Pt 5): 860-879.

Leiguarda RC, Merello M, Nouzeilles MI, Balej J, Rivero A, Nogues M. Limb-kinetic apraxia in corticobasal degeneration: clinical and kinematic features. Mov Disord 2003; 18: 49-59.

Levy-Cooperman N, Burhan AM, Rafi-Tari S et al. Frontal lobe hypoperfusion and depressive symptoms in Alzheimer disease. J Psychiatry Neurosci 2008a; 33: 218-226.

Levy-Cooperman N, Ramirez J, Lobaugh NJ, Black SE. Misclassified tissue volumes in Alzheimer disease patients with white matter hyperintensities: importance of lesion segmentation procedures for volumetric analysis. Stroke 2008b; 39: 1134-1141.

Lezak MD. Neuropsychological Assessment. New York: Oxford University Press; 1983.

Liepmann H. Apraxie. In: Brugsch H, editor. Ergebnisse der gesamten Medizin. Wien, Berlin: Urban & Schwarzenberg; 1920. p. 516-43.

Ling H, O'Sullivan SS, Holton JL et al. Does corticobasal degeneration exist? A clinicopathological re-evaluation. Brain 2010; 133: 2045-2057.

Liotti M, Mayberg HS. The role of functional neuroimaging in the neuropsychology of depression. J Clin Exp Neuropsychol 2001; 23: 121-136.

Lipton AM, White CL, III, Bigio EH. Frontotemporal lobar degeneration with motor neuron disease-type inclusions predominates in 76 cases of frontotemporal degeneration. Acta Neuropathol (Berl) 2004; 108: 379-385.

Litvan I, Agid Y, Goetz C et al. Accuracy of the clinical diagnosis of corticobasal degeneration: a clinicopathologic study. Neurology 1997; 48: 119-125.

203

Litvan I, Bhatia KP, Burn DJ et al. Movement Disorders Society Scientific Issues Committee report: SIC Task Force appraisal of clinical diagnostic criteria for Parkinsonian disorders. Mov Disord 2003; 18: 467-486.

Litvan I, Cummings JL, Mega M. Neuropsychiatric features of corticobasal degeneration. J Neurol Neurosurg Psychiatry 1998; 65: 717-721.

Llado A, Sanchez-Valle R, Rene R et al. Late-onset frontotemporal dementia associated with a novel PGRN mutation. J Neural Transm 2007; 114: 1051-1054.

Lobaugh NJ, Caldwell CB, Black SE, Leibovitch FS, Swartz RH. Three brain SPECT region-of- interest templates in elderly people: normative values, hemispheric asymmetries, and a comparison of single- and multihead cameras. J Nucl Med 2000; 41: 45-56.

Lopez de MA, Alzualde A, Gorostidi A et al. Mutations in progranulin gene: clinical, pathological, and ribonucleic acid expression findings. Biol Psychiatry 2008; 63: 946-952.

Mackenzie IR, Baker M, West G et al. A family with tau-negative frontotemporal dementia and neuronal intranuclear inclusions linked to chromosome 17. Brain 2006; 129: 853-867.

Mackenzie IR, Feldman HH. Ubiquitin immunohistochemistry suggests classic motor neuron disease, motor neuron disease with dementia, and frontotemporal dementia of the motor neuron disease type represent a clinicopathologic spectrum. J Neuropathol Exp Neurol 2005; 64: 730- 739.

Mahapatra RK, Edwards MJ, Schott JM, Bhatia KP. Corticobasal degeneration. Lancet Neurol 2004; 3: 736-743.

Mann DM, McDonagh AM, Snowden J, Neary D, Pickering-Brown SM. Molecular classification of the dementias. Lancet 2000; 355: 626.

Markus HS, Lees AJ, Lennox G, Marsden CD, Costa DC. Patterns of regional cerebral blood flow in corticobasal degeneration studied using HMPAO SPECT; comparison with Parkinson's disease and normal controls. Mov Disord 1995; 10: 179-187.

Masellis M, Momeni P, Meschino W et al. Novel splicing mutation in the progranulin gene causing familial corticobasal syndrome. Brain 2006; 129: 3115-3123.

Masellis M, Zinman L, Black SE. More than just 'frontal': disentangling behavioural disturbances in amyotrophic lateral sclerosis. Eur J Neurol 2010; 17: 5-7.

Mathew R, Bak TH, Hodges JR. Diagnostic criteria for corticobasal syndrome: a comparative study. J Neurol Neurosurg Psychiatry 2011.

Mathuranath PS, Xuereb JH, Bak T, Hodges JR. Corticobasal ganglionic degeneration and/or frontotemporal dementia? A report of two overlap cases and review of literature. J Neurol Neurosurg Psychiatry 2000; 68: 304-312.

204

Matsuda H, Yagishita A, Tsuji S, Hisada K. A quantitative approach to technetium-99m ethyl cysteinate dimer: a comparison with technetium-99m hexamethylpropylene amine oxime. Eur J Nucl Med 1995; 22: 633-637.

Mattis S. Mental status examination for organic mental syndrome in the elderly patient. In: Bellak L, Karasu TB, editors. Geriatric Psychiatry. New York: Grune & Stratton; 1976.

McKhann GM, Albert MS, Grossman M, Miller B, Dickson D, Trojanowski JQ. Clinical and pathological diagnosis of frontotemporal dementia: report of the Work Group on Frontotemporal Dementia and Pick's Disease. Arch Neurol 2001; 58: 1803-1809.

McMonagle P, Blair M, Kertesz A. Corticobasal degeneration and progressive aphasia. Neurology 2006; 67: 1444-1451.

Mei L, Xue G, Chen C, Xue F, Zhang M, Dong Q. The "visual word form area" is involved in successful memory encoding of both words and faces. Neuroimage 2010; 52: 371-378.

Mendez MF. Corticobasal ganglionic degeneration with Balint's syndrome. J Neuropsychiatry Clin Neurosci 2000; 12: 273-275.

Mendez MF, Shapira JS, McMurtray A, Licht E, Miller BL. Accuracy of the clinical evaluation for frontotemporal dementia. Arch Neurol 2007; 64: 830-835.

Mielke R, Heiss WD. Positron emission tomography for diagnosis of Alzheimer's disease and vascular dementia. J Neural Transm Suppl 1998; 53: 237-250.

Moll J, de Oliveira-Souza R, Passman LJ, Cunha FC, Souza-Lima F, Andreiuolo PA. Functional MRI correlates of real and imagined tool-use pantomimes. Neurology 2000; 54: 1331-1336.

Momeni P, Rogaeva E, Deerlin VV. Genetic variability in CHMP2B and Frontotemporal dementia. Neurobiol Dis 2006.

Moreno F, Indakoetxea B, Barandiaran M et al. "Frontotemporoparietal" dementia: clinical phenotype associated with the c.709-1G>A PGRN mutation. Neurology 2009; 73: 1367-1374.

Morinaga A, Ono K, Ikeda T et al. A comparison of the diagnostic sensitivity of MRI, CBF- SPECT, FDG-PET and cerebrospinal fluid biomarkers for detecting Alzheimer's disease in a memory clinic. Dement Geriatr Cogn Disord 2010; 30: 285-292.

Murray R, Neumann M, Forman MS et al. Cognitive and motor assessment in autopsy-proven corticobasal degeneration. Neurology 2007; 68: 1274-1283.

Nagahama Y, Fukuyama H, Turjanski N et al. Cerebral glucose metabolism in corticobasal degeneration: comparison with progressive supranuclear palsy and normal controls. Mov Disord 1997; 12: 691-696.

Nagasawa H, Tanji H, Nomura H et al. PET study of cerebral glucose metabolism and fluorodopa uptake in patients with corticobasal degeneration. J Neurol Sci 1996; 139: 210-217.

205

Naj AC, Jun G, Beecham GW et al. Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer's disease. Nat Genet 2011; 43: 436-441.

Nalls MA, Plagnol V, Hernandez DG et al. Imputation of sequence variants for identification of genetic risks for Parkinson's disease: a meta-analysis of genome-wide association studies. Lancet 2011; 377: 641-649.

Neary D, Snowden JS, Gustafson L et al. Frontotemporal lobar degeneration: a consensus on clinical diagnostic criteria. Neurology 1998; 51: 1546-1554.

Neumann M, Sampathu DM, Kwong LK et al. Ubiquitinated TDP-43 in frontotemporal lobar degeneration and amyotrophic lateral sclerosis. Science 2006; 314: 130-133.

Neumann M, Tolnay M, Mackenzie IR. The molecular basis of frontotemporal dementia. Expert Rev Mol Med 2009; 11: e23.

Noorbakhsh F, Overall CM, Power C. Deciphering complex mechanisms in neurodegenerative diseases: the advent of systems biology. Trends Neurosci 2009; 32: 88-100.

Nothen MM, Propping P, Fimmers R. Association versus linkage studies in psychosis genetics. J Med Genet 1993; 30: 634-637.

O'Sullivan SS, Burn DJ, Holton JL, Lees AJ. Normal dopamine transporter single photon- emission CT scan in corticobasal degeneration. Mov Disord 2008; 23: 2424-2426.

Ochipa C, Gonzalez Rothi LJ. Limb apraxia. Semin Neurol 2000; 20: 471-478.

Ochipa C, Rothi LJ, Heilman KM. Conceptual apraxia in Alzheimer's disease. Brain 1992; 115 ( Pt 4): 1061-1071.

Okuda B, Kodama N, Tachibana H, Sugita M, Tanaka H. Visuomotor ataxia in corticobasal degeneration. Mov Disord 2000a; 15: 337-340.

Okuda B, Tachibana H, Kawabata K, Takeda M, Sugita M. Cerebral blood flow correlates of higher brain dysfunctions in corticobasal degeneration. J Geriatr Psychiatry Neurol 1999; 12: 189-193.

Okuda B, Tachibana H, Kawabata K, Takeda M, Sugita M. Cerebral blood flow in corticobasal degeneration and progressive supranuclear palsy. Alzheimer Dis Assoc Disord 2000b; 14: 46-52.

Osterrieth PA. The test of copying a complex figure: A contribution to the study of perception and memory. 1944. p. 286-356.

Ott J. Cutting a Gordian knot in the linkage analysis of complex human traits. Am J Hum Genet 1990; 46: 219-221.

Peigneux P, Salmon E, Garraux G et al. Neural and cognitive bases of upper limb apraxia in corticobasal degeneration. Neurology 2001; 57: 1259-1268.

206

Pick A. U¨ ber die Beziehungen der senilen Hirnatrophie zur Aphasie. Prag MedWochenschr 1892; 17: 165-167.

Pickering-Brown SM, Rollinson S, Du PD et al. Frequency and clinical characteristics of progranulin mutation carriers in the Manchester frontotemporal lobar degeneration cohort: comparison with patients with MAPT and no known mutations. Brain 2008; 131: 721-731.

Pillon B, Blin J, Vidailhet M et al. The neuropsychological pattern of corticobasal degeneration: comparison with progressive supranuclear palsy and Alzheimer's disease. Neurology 1995; 45: 1477-1483.

Pillon B, Dubois B. Memory and executive processes in corticobasal degeneration. Corticobasal degeneration and related disorders. Philadelphia: Lippincott Williams & Wilkins; 2000. p. 91- 101.

Pirker W, Asenbaum S, Bencsits G et al. [123I]beta-CIT SPECT in multiple system atrophy, progressive supranuclear palsy, and corticobasal degeneration. Mov Disord 2000; 15: 1158- 1167.

Pirker W, Djamshidian S, Asenbaum S et al. Progression of dopaminergic degeneration in Parkinson's disease and atypical parkinsonism: a longitudinal beta-CIT SPECT study. Mov Disord 2002; 17: 45-53.

Plotkin M, Amthauer H, Klaffke S et al. Combined 123I-FP-CIT and 123I-IBZM SPECT for the diagnosis of parkinsonian syndromes: study on 72 patients. J Neural Transm 2005; 112: 677-692.

Poorkaj P, Bird TD, Wijsman E et al. Tau is a candidate gene for chromosome 17 frontotemporal dementia. Ann Neurol 1998; 43: 815-825.

Rademakers R, Baker M, Gass J et al. Phenotypic variability associated with progranulin haploinsufficiency in patients with the common 1477C-->T (Arg493X) mutation: an international initiative. Lancet Neurol 2007; 6: 857-868.

Rademakers R, Eriksen JL, Baker M et al. Common variation in the miR-659 binding-site of GRN is a major risk factor for TDP43-positive frontotemporal dementia. Hum Mol Genet 2008; 17: 3631-3642.

Rahmim A, Zaidi H. PET versus SPECT: strengths, limitations and challenges. Nucl Med Commun 2008; 29: 193-207.

Ramirez J, Gibson E, Quddus A et al. Lesion Explorer: a comprehensive segmentation and parcellation package to obtain regional volumetrics for subcortical hyperintensities and intracranial tissue. Neuroimage 2011; 54: 963-973.

Rankin KP, Gorno-Tempini ML, Allison SC et al. Structural anatomy of empathy in neurodegenerative disease. Brain 2006; 129: 2945-2956.

207

Raven JC. Colored Prorgessive Matrices Sets A, Ab, B. Oxford: Oxford Psychologists Press Ltd.; 1947.

Rebeiz JJ, Kolodny EH, Richardson EP, Jr. Corticodentatonigral degeneration with neuronal achromasia: a progressive disorder of late adult life. Trans Am Neurol Assoc 1967; 92: 23-26.

Rebeiz JJ, Kolodny EH, Richardson EP, Jr. Corticodentatonigral degeneration with neuronal achromasia. Arch Neurol 1968; 18: 20-33.

Rey A. L'examen psychologique dans les cas d'encephalopathie traumatique. 1941. p. 215-85.

Riley DE, Lang AE. Clinical Diagnostic Criteria. Corticobasal degeneration and related disorders. Philadelphia: Lippincott Williams & Wilkins; 2000. p. 29-34.

Riley DE, Lang AE, Lewis A et al. Cortical-basal ganglionic degeneration. Neurology 1990; 40: 1203-1212.

Rinne JO, Lee MS, Thompson PD, Marsden CD. Corticobasal degeneration. A clinical study of 36 cases. Brain 1994; 117 ( Pt 5): 1183-1196.

Rippon GA, Staugaitis SM, Chin SS, Goldman JE, Marder K. Corticobasal syndrome with novel argyrophilic glial inclusions. Mov Disord 2005; 20: 598-602.

Risch N, Merikangas K. The future of genetic studies of complex human diseases. Science 1996; 273: 1516-1517.

Rizzo M, Eslinger PJ. Principles and Practice of Behavioral Neurology and Neuropsychology. Philadelphia: WB Saunders; 2004.

Rohrer JD, Beck J, Warren JD et al. Corticobasal syndrome associated with a novel 1048_1049insG progranulin mutation. J Neurol Neurosurg Psychiatry 2009; 80: 1297-1298.

Rohrer JD, Geser F, Zhou J et al. TDP-43 subtypes are associated with distinct atrophy patterns in frontotemporal dementia. Neurology 2010a; 75: 2204-2211.

Rohrer JD, Rossor MN, Warren JD. Apraxia in progressive nonfluent aphasia. J Neurol 2010b; 257: 569-574.

Ropper AH, Brown RH. Adams and Victor's Principles of Neurology. New York: McGraw-Hill; 2005.

Rothi LJG, Ochipa C, Heilman KM. A cognitive neuropsychological model of limb praxis. Cognitive Neuropsychology 1991; 8: 443-458.

Rouleau I, Salmon DP, Butters N, Kennedy C, McGuire K. Quantitative and qualitative analyses of clock drawings in Alzheimer's and Huntington's disease. Brain Cogn 1992; 18: 70-87.

Roy EA. Hand preference, manual asymmetries and limb apraxia. In: Elliot D, Roy EA, editors. Manual asymmetries in motor control. Boca Raton: CRC Press; 1996. p. 215-36. 208

Roy EA, Black SE, Blair N, Dimeck PT. Analyses of deficits in gestural pantomime. J Clin Exp Neuropsychol 1998; 20: 628-643.

Savoiardo M, Grisoli M, Girotti F. Magnetic resonance imaging in CBD, related atypical parkinsonian disorders, and dementias. Adv Neurol 2000; 82: 197-208.

Sawle GV, Brooks DJ, Marsden CD, Frackowiak RS. Corticobasal degeneration. A unique pattern of regional cortical oxygen hypometabolism and striatal fluorodopa uptake demonstrated by positron emission tomography. Brain 1991; 114 ( Pt 1B): 541-556.

Schneider JA, Watts RL, Gearing M, Brewer RP, Mirra SS. Corticobasal degeneration: neuropathologic and clinical heterogeneity. Neurology 1997; 48: 959-969.

Schulze TG, McMahon FJ. Defining the phenotype in human genetic studies: forward genetics and reverse phenotyping. Hum Hered 2004; 58: 131-138.

Shahani N, Brandt R. Functions and malfunctions of the tau proteins. Cell Mol Life Sci 2002; 59: 1668-1680.

Shelley BP, Hodges JR, Kipps CM, Xuereb JH, Bak TH. Is the pathology of corticobasal syndrome predictable in life? Mov Disord 2009; 24: 1593-1599.

Silverman DH. Brain 18F-FDG PET in the diagnosis of neurodegenerative dementias: comparison with perfusion SPECT and with clinical evaluations lacking nuclear imaging. J Nucl Med 2004; 45: 594-607.

Skibinski G, Parkinson NJ, Brown JM et al. Mutations in the endosomal ESCRTIII-complex subunit CHMP2B in frontotemporal dementia. Nat Genet 2005; 37: 806-808.

Skoglund L, Brundin R, Olofsson T et al. Frontotemporal dementia in a large Swedish family is caused by a progranulin null mutation. Neurogenetics 2009; 10: 27-34.

Snowden J, Neary D, Mann D. Frontotemporal lobar degeneration: clinical and pathological relationships. Acta Neuropathol 2007; 114: 31-38.

Snowden JS, Pickering-Brown SM, Mackenzie IR et al. Progranulin gene mutations associated with frontotemporal dementia and progressive non-fluent aphasia. Brain 2006; 129: 3091-3102.

Soliveri P, Monza D, Paridi D et al. Cognitive and magnetic resonance imaging aspects of corticobasal degeneration and progressive supranuclear palsy. Neurology 1999; 53: 502-507.

Soliveri P, Piacentini S, Girotti F. Limb apraxia in corticobasal degeneration and progressive supranuclear palsy. Neurology 2005; 64: 448-453.

Spatt J, Bak T, Bozeat S, Patterson K, Hodges JR. Apraxia, mechanical problem solving and semantic knowledge: contributions to object usage in corticobasal degeneration. J Neurol 2002; 249: 601-608.

209

Spillantini MG, Murrell JR, Goedert M, Farlow MR, Klug A, Ghetti B. Mutation in the tau gene in familial multiple system tauopathy with presenile dementia. Proc Natl Acad Sci U S A 1998; 95: 7737-7741.

Spina S, Murrell JR, Huey ED et al. Corticobasal syndrome associated with the A9D Progranulin mutation. J Neuropathol Exp Neurol 2007; 66: 892-900.

Stamatakis EA, Wilson JT, Wyper DJ. Spatial normalization of lesioned HMPAO-SPECT images. Neuroimage 2001; 14: 844-852.

Stamenova V, Roy EA, Black SE. A model-based approach to understanding apraxia in Corticobasal Syndrome. Neuropsychol Rev 2009; 19: 47-63.

Stamenova V, Roy EA, Black SE. Limb apraxia in corticobasal syndrome. Cortex 2011; 47: 460- 472.

Taki M, Ishii K, Fukuda T, Kojima Y, Mori E. Evaluation of cortical atrophy between progressive supranuclear palsy and corticobasal degeneration by hemispheric surface display of MR images. AJNR Am J Neuroradiol 2004; 25: 1709-1714.

Tan CF, Piao YS, Kakita A et al. Frontotemporal dementia with co-occurrence of astrocytic plaques and tufted astrocytes, and severe degeneration of the cerebral white matter: a variant of corticobasal degeneration? Acta Neuropathol 2005; 109: 329-338.

Tang-Wai DF, Josephs KA, Boeve BF, Dickson DW, Parisi JE, Petersen RC. Pathologically confirmed corticobasal degeneration presenting with visuospatial dysfunction. Neurology 2003; 61: 1134-1135.

Taniguchi S, McDonagh AM, Pickering-Brown SM et al. The neuropathology of frontotemporal lobar degeneration with respect to the cytological and biochemical characteristics of tau protein. Neuropathol Appl Neurobiol 2004; 30: 1-18.

Taniwaki T, Yamada T, Yoshida T et al. Heterogeneity of glucose metabolism in corticobasal degeneration. J Neurol Sci 1998; 161: 70-76.

The Lund and Manchester Groups. Clinical and neuropathological criteria for frontotemporal dementia. J Neurol Neurosurg Psychiatry 1994; 57: 416-418.

Thompson PD, Day BL, Rothwell JC, Brown P, Britton TC, Marsden CD. The myoclonus in corticobasal degeneration. Evidence for two forms of cortical reflex myoclonus. Brain 1994; 117 ( Pt 5): 1197-1207.

Tippett WJ, Black SE. Regional cerebral blood flow correlates of visuospatial tasks in Alzheimer's disease. J Int Neuropsychol Soc 2008; 14: 1034-1045.

Todd JJ, Marois R. Capacity limit of visual short-term memory in human posterior parietal cortex. Nature 2004; 428: 751-754.

210

Togasaki DM, Tanner CM. Epidemiologic aspects. Adv Neurol 2000; 82: 53-59.

Tokumaru AM, O'uchi T, Kuru Y, Maki T, Murayama S, Horichi Y. Corticobasal degeneration: MR with histopathologic comparison. AJNR Am J Neuroradiol 1996; 17: 1849-1852.

Tsuchiya K, Ikeda K, Uchihara T, Oda T, Shimada H. Distribution of cerebral cortical lesions in corticobasal degeneration: a clinicopathological study of five autopsy cases in Japan. Acta Neuropathol 1997; 94: 416-424.

Tuite PJ, Clark HB, Bergeron C et al. Clinical and pathologic evidence of corticobasal degeneration and progressive supranuclear palsy in familial tauopathy. Arch Neurol 2005; 62: 1453-1457.

Uchihara T, Nakayama H. Familial tauopathy mimicking corticobasal degeneration an autopsy study on three siblings. J Neurol Sci 2006; 246: 45-51.

Ungerleider LG, Courtney SM, Haxby JV. A neural system for human visual working memory. Proc Natl Acad Sci U S A 1998; 95: 883-890.

Van Deerlin V, Sleiman PM, Martinez-Lage M et al. Common variants at 7p21 are associated with frontotemporal lobar degeneration with TDP-43 inclusions. Nat Genet 2010; 42: 234-239.

Vanek Z, Jankovic J. Dystonia in corticobasal degeneration. Mov Disord 2001; 16: 252-257.

Wadia PM, Lang AE. The many faces of corticobasal degeneration. Parkinsonism Relat Disord 2007; 13 Suppl 3: S336-S340.

Wechsler D. Wechsler Memory Scale-Revised manual. San Antonio: The Psychological Corporation; 1987.

Wenning GK, Litvan I, Jankovic J et al. Natural history and survival of 14 patients with corticobasal degeneration confirmed at postmortem examination. J Neurol Neurosurg Psychiatry 1998; 64: 184-189.

Whitwell JL, Jack CR, Jr., Boeve BF et al. Imaging correlates of pathology in corticobasal syndrome. Neurology 2010; 75: 1879-1887.

Williams BW, Mack W, Henderson VW. Boston Naming Test in Alzheimer's disease. Neuropsychologia 1989; 27: 1073-1079.

Winkelmann J, Auer DP, Lechner C, Elbel G, Trenkwalder C. Magnetic resonance imaging findings in corticobasal degeneration. Mov Disord 1999; 14: 669-673.

Wong SH, Lecky BR, Steiger MJ. Parkinsonism and impulse control disorder: presentation of a new progranulin gene mutation. Mov Disord 2009; 24: 618-619.

211

Worsley KJ, Marrett S, Neelin P, Vandal AC, Friston KJ, Evans AC. A unified statistical approach for determining significant signals in images of cerebral activation. Hum Brain Mapp 1996; 4: 58-73.

Yamauchi H, Fukuyama H, Nagahama Y et al. Atrophy of the corpus callosum, cortical hypometabolism, and cognitive impairment in corticobasal degeneration. Arch Neurol 1998a; 55: 609-614.

Yamauchi H, Fukuyama H, Nagahama Y et al. Atrophy of the corpus callosum, cortical hypometabolism, and cognitive impairment in corticobasal degeneration. Arch Neurol 1998b; 55: 609-614.

Yekhlef F, Ballan G, Macia F, Delmer O, Sourgen C, Tison F. Routine MRI for the differential diagnosis of Parkinson's disease, MSA, PSP, and CBD. J Neural Transm 2003; 110: 151-169.

Yu CE, Bird TD, Bekris LM et al. The spectrum of mutations in progranulin: a collaborative study screening 545 cases of neurodegeneration. Arch Neurol 2010; 67: 161-170.

Zadikoff C, Lang AE. Apraxia in movement disorders. Brain 2005; 128: 1480-1497.

Zamboni G, Huey ED, Krueger F, Nichelli PF, Grafman J. Apathy and disinhibition in frontotemporal dementia: Insights into their neural correlates. Neurology 2008; 71: 736-742.

Zekanowski C, Peplonska B, Styczynska M, Gustaw K, Kuznicki J, Barcikowska M. Mutation screening of the MAPT and STH genes in Polish patients with clinically diagnosed frontotemporal dementia. Dement Geriatr Cogn Disord 2003a; 16: 126-131.

Zekanowski C, Styczynska M, Peplonska B et al. Mutations in presenilin 1, presenilin 2 and amyloid precursor protein genes in patients with early-onset Alzheimer's disease in Poland. Exp Neurol 2003b; 184: 991-996.

Zhang L, Murata Y, Ishida R, Saitoh Y, Mizusawa H, Shibuya H. Differentiating between progressive supranuclear palsy and corticobasal degeneration by brain perfusion SPET. Nucl Med Commun 2001; 22: 767-772.

212