neuroimagerie et pharmacothérapie de la démence atypique€¦ · neuroimagerie et...
Post on 15-Aug-2020
14 Views
Preview:
TRANSCRIPT
Neuroimagerie et pharmacothérapie de la démence atypique
Étude morphologique de la variante sémantique de l’aphasie primaire progressive et revue systématique de la pharmacothérapie en
dégénérescence lobaire fronto-temporale
Mémoire
Louis-Olivier Bouchard
Maîtrise en épidémiologie - épidémiologie clinique Maître ès sciences (M.Sc.)
Québec, Canada
© Louis-Olivier Bouchard, 2016
Neuroimagerie et pharmacothérapie de la démence atypique
Étude morphologique de la variante sémantique de l’aphasie primaire progressive et revue systématique de la pharmacothérapie en
dégénérescence lobaire fronto-temporale
Mémoire
Louis-Olivier Bouchard
Sous la direction de :
Simon Duchesne, directeur de recherche Maximiliano A. Wilson, codirecteur de recherche
iii
Résumé
Les démences sont un enjeu majeur de santé. La dégénérescence lobaire fronto-
temporale (DLFT), deuxième forme la plus prévalente de démence chez les personnes
âgées de moins de 65 ans, inclut entre autres la variante sémantique de l’aphasie primaire
progressive (svPPA), une maladie qui affecte particulièrement et initialement le langage.
Anatomiquement, on sait déjà qu’on retrouve en svPPA une atrophie principalement
marquée au niveau temporal, davantage à gauche et en antérieur. La connaissance des
atteintes de la matière blanche est toutefois moins étoffée pour l’instant. Au niveau
thérapeutique, il existe une controverse quant à l’approche à privilégier en DLFT : plusieurs
molécules ont été étudiées, plusieurs sont prescrites et pourtant il n’y a ni consensus, ni
recommandation à cet effet.
Nos objectifs dans ce mémoire sont donc d’abord de mieux caractériser les atteintes
cérébrales de la matière blanche et de la matière grise chez les patients atteints de svPPA,
par une étude tractographique et volumétrique, et ensuite d’évaluer l’efficacité de la
pharmacothérapie chez les patients avec DLFT en termes d’effet sur la cognition et sur des
symptômes neuropsychiatriques, grâce à une revue systématique avec méta-analyse.
En imagerie, notre étude a montré une diminution de la diffusion au niveau du
fascicule longitudinal supérieur gauche, de la capsule externe gauche, du cingulum droit et
du fascicule unciné bilatéralement et une atrophie plus marquée en temporal gauche, ainsi
qu’au niveau de l’amygdale et des cortex fusiforme et entorhinal. En pharmacothérapie,
aucune médication n’a démontré d’effet sur la cognition globale, mais certaines molécules
ont montré un bénéfice potentiel sur le langage, l’impulsivité et la reconnaissance des
émotions.
Ce mémoire a ainsi permis des avancées au niveau de la caractérisation des atteintes
cérébrales en svPPA et de faire le point sur l’état de la littérature en pharmacothérapie de la
DLFT.
iv
Abstract
Dementia is a major health issue. Frontotemporal lobar degeneration (FTLD), the
second most common dementia in individuals under 65 years of age, includes the semantic
variant of primary progressive aphasia (svPPA), a disease affecting mainly and initially
language. Anatomically, we know that svPPA patients show cortical atrophy, markedly in
the temporal lobes, more in the left hemisphere and anteriorly. However, our knowledge of
white matter damage is less developed. As for FTLD pharmacotherapies, there remains
much controversy. Many molecules have been studied, some are currently prescribed, but
there still is no consensus, nor any recommendation to this effect.
Our objectives in this memoir were first to better characterize cerebral damage for
white and grey matter in svPPA patients by means of a tractographic and volumetric study,
and secondly to assess the effect on global cognition and specific neuropsychiatric
symptoms of pharmacotherapy in FTLD patients, with a systematic review and meta-
analysis.
Imaging results show a diminution of fractional anisotropic diffusion in the left
superior longitudinal fasciculus, external capsule, right cingulum and bilateral uncinate
fasciculi. They also show atrophy, markedly in the left temporal lobe, amygdala, fusiform
and entorhinal cortices. As for pharmacotherapy results, no medication was shown to have
any beneficial effects on global cognition, but some drugs may improve language,
impulsivity and emotion recognition.
This memoir has indeed improved the characterization of cerebral damage in the
svPPA and reviewed thoroughly the literature on pharmacotherapy in FLTD.
v
Table des matières
RÉSUMÉ III
ABSTRACT IV
TABLE DES MATIÈRES V
LISTE DES TABLEAUX VII
AVANT-PROPOS VIII
CHAPITRE 1 : INTRODUCTION 1
Mise en contexte 1
Enjeux 2
État des connaissances 4
Objectifs du mémoire 6
CHAPITRE 2 7
BETWEEN-GROUP STUDY OF TRACTOGRAPHY AND VOLUMETRY IN THE SEMANTIC VARIANT OF PRIMARY PROGRESSIVE APHASIA 7
ABSTRACT 8
INTRODUCTION 10
METHOD 11
RESULTS 15
DISCUSSION 16
Tables 21
Figures* 24
CHAPITRE 3 27
vi
EFFECT OF PHARMACOTHERAPY IN FRONTOTEMPORAL LOBAR DEGENERATION: A SYSTEMATIC REVIEW AND META-ANALYSIS OF RANDOMISED CONTROLLED TRIALS 27
ABSTRACT 28
INTRODUCTION 30
METHODS 31
RESULTS 33
DISCUSSION 36
FIGURES 41
TABLES 43
CHAPITRE 4 : CONCLUSION 48
Rappel des objectifs 48
Résumé des résultats 49
Perspectives 51
RÉFÉRENCES 54
vii
Liste des tableaux
Chapitre 2 Table 1 Demographic and clinical data of svPPA cases and controls .............................................................. 21 Table 2 White matter study : fractional anisotropy data of svPPA cases and controls .................................... 21 Table 3 Grey matter study : volumetric data of svPPA cases and controls ....................................................... 22 Chapitre 3 Table 4 Study characteristics ............................................................................................................................ 43 Table 5 Effect of Pharmacotherapy on Cognition ............................................................................................. 44 Table 6 Effect of Pharmacotherapy on Neuropsychiatric Symptoms/Specific cognitive measures and
Safety/Tolerability ..................................................................................................................................... 45 Table 7 Meta-analyses of the effect of pharmacotherapy on cognition and on dementia symptoms ................ 46 Table 8 Risk of bias within studies with Cochrane's Collaboration tool ............................................................. 47 Liste des figures Chapitre 2 Figure 1 White matter maps : statistically significant difference of fractional anisotropy between controls and
svPPA patients ......................................................................................................................................... 24 Figure 2 Grey matter images : examples of a control compared to a svPPA patient ........................................ 26 Chapitre 3 Figure 3 Example of a Search Strategy ............................................................................................................ 41 Figure 4 Flow diagram of Study selection ......................................................................................................... 42
viii
Avant-propos
Chapitre 2
Titre : Between-Group Study of Tractography and Volumetry in the Semantic Variant of
Primary Progressive Aphasia
Auteur principal : Louis-Olivier Bouchard, M.D., louis-olivier.bouchard.1@ulaval.ca
Co-auteurs : Maximiliano Wilson, Ph.D., maximiliano.wilson@fmed.ulaval.ca
Simon Duchesne, ing., Ph.D., simon.duchesne@fmed.ulaval.ca
État de publication : À être soumis à la revue Neurology
Contributions :
- LB: Conception et design de l’étude, analyse et interprétation des données, écriture
du manuscrit et approbation pour publication.
- MW: Conception et design de l’étude, analyse des données, révision critique du
manuscrit et approbation pour publication.
- SD: Conception et design de l’étude, analyse des données, révision critique du
manuscrit et approbation pour publication.
- Tous les auteurs se portent garants de l’intégrité de l’ensemble de l’étude.
ix
Chapitre 3
Titre : Effect of Pharmacotherapy in Frontotemporal Lobar Degeneration : a Systematic
Review and Meta-analysis of Randomised Controlled Trials
Auteur principal : Louis-Olivier Bouchard, M.D., louis-olivier.bouchard.1@ulaval.ca
Co-auteurs : Yuliya Bodryzlova, MD, M.Sc., julia.bodryzlova@gmail.com
Olivier Potvin, Ph.D., olivier.potvin.1@ulaval.ca
Simon Duchesne, ing., Ph.D., simon.duchesne@fmed.ulaval.ca
État de publication : À être soumis à la revue Alzheimer & Dementia
Contributions :
- LB: Conception et design de l’étude, analyse et interprétation des données, écriture
du manuscrit et approbation pour publication.
- YB: Contribution à la conception et design de l’étude, analyse et interprétation des
données, approbation pour publication.
- OP: Contribution à la conception et design de l’étude, révision critique du
manuscrit, approbation pour publication.
- SD: Contribution à la conception et design de l’étude, révision critique du
manuscrit, approbation pour publication.
- Tous les auteurs se portent garants de l’intégrité de l’ensemble de l’étude.
1
Chapitre 1 : Introduction
Mise en contexte
La démence est définie par le Diagnostic and Statistical Manual of Mental
Disorders V (DSM-V)1 comme un trouble entraînant un déclin cognitif, par exemple des
atteintes de la mémoire, du langage ou des fonctions exécutives, et qui nuit au
fonctionnement du patient au quotidien.
C’est une condition qui, bien qu’elle puisse être fluctuante dans sa symptomatologie
à court terme, est irréversible et ne peut être mieux expliquée par un délirium ou une autre
pathologie aigüe de type infectieux, néoplasique, endocrinien, métabolique, toxique,
traumatique ou d’autre nature.
Parmi les démences, on retrouve un ensemble de maladies neurodégénératives
cliniquement et pathologiquement différentes, la maladie d’Alzheimer ayant la prévalence
la plus élevée - environ 60% des cas2. Les autres causes de démence, parfois regroupées
sous l’appellation «démences atypiques», incluent la dégénérescence lobaire fronto-
temporale (DLFT; sujet d’intérêt de ce mémoire), la démence vasculaire, la démence à
corps de Lewy et plusieurs autres maladies moins fréquentes.
À ce jour, aucune cause précise n’a été identifiée pour expliquer la pathogénèse des
démences neurodégénératives, toutes étiologies confondues. Certains facteurs de risque ont
certes été identifiés. Jouent un rôle : l’âge, la diathèse familiale, la génétique, la présence
d’un diagnostic de trouble cognitif léger, les troubles cardio-métaboliques (hypertension
artérielle, dyslipidémie, diabète sucré, obésité, tabagisme, maladie coronarienne), le stress,
la scolarité, l’emploi, les activités et d’autres facteurs (antécédents de trauma, exposition à
des toxines, etc.)3.
La DLFT est en soi une famille de maladies neurodégénératives représentant
environ 10% des cas totaux de démence4, soit la quatrième en prévalence. Cependant, sa
présentation, souvent précoce, signifie qu’elle est la seconde en importance chez les
patients de moins de 65 ans, quasiment ex-aequo avec la maladie d’Alzheimer à cet âge4.
Différents syndromes et pathologies ont été regroupés sous l’entité appelée DLFT
2
pour des raisons d’abord anatomiques, vues leurs atteintes des cortex frontaux et
temporaux, par opposition aux hippocampes, au mésencéphale, aux lobes pariétaux ou
occipitaux5. En terme de troubles cognitifs, on distingue d’abord deux grands sous-types,
soit la variante comportementale de la démence fronto-temporale (behavioral variant of
frontotemporal dementia, bvFTD) et les aphasies primaires progressives.
La bvFTD, autrefois connue sous le nom de maladie de Pick, du nom du neurologue
et pathologiste qui l’a décrite pour la première fois en 18925, affecte en premier lieu le
comportement, les réponses émotionnelles et l’inhibition.
Les aphasies primaires progressives se présentent quant à elles initialement par des
atteintes du langage, isolées ou prédominantes6. Les aphasies primaires progressives se
divisent elles-mêmes en trois variantes7. Il y a d’abord, et ce sera notre principal sujet
d’étude dans ce mémoire, la variante sémantique de l’aphasie primaire progressive
(semantic variant of primary progressive aphasia, svPPA, parfois appelée démence
sémantique), qui est caractérisée par des difficultés à nommer des objets et à comprendre
des mots, avec possible dyslexie et/ou dysgraphie de surface. Ensuite, on retrouve la
variante non-fluente ou agrammatique de l’aphasie primaire progressive, qui touche
principalement la production du langage. Finalement, le troisième sous-type du groupe, la
variante logopénique de l’aphasie primaire progressive, touche aussi le langage, avec des
atteintes marquée de la répétition et de la capacité à nommer les objets7.
Enjeux
Les démences sont un enjeu de santé majeur, affectant présentement quelque 36
millions de personnes mondialement, et près de 750 000 au Canada seulement2. Avec le
vieillissement de la population, il est attendu que ces nombres croissent de façon importante
au cours des prochaines années, atteignant selon certains estimés2 66 millions de patients
mondialement (1,4 millions au Canada) au cours des 15 prochaines années. En effet, les 65
ans et plus représentent déjà 15% de la population canadienne et cette proportion
augmentera constamment dans les prochaines décennies. Possiblement en raison de leur
plus grande longévité, 66-72% des patients atteints d’une forme ou l’autre de démence
seront des femmes8.
3
En plus de la souffrance et de la détresse vécues par les patients et leurs proches, les
démences auront aussi un impact important sur l’économie : les coûts mondiaux estimés
engendrés par ces maladies sont présentement de 600 milliards de dollars annuellement (33
milliards au Canada) et risquent de grimper en flèche8. Aux États-Unis seulement, les coûts
sont présentement de l’ordre de 226 milliards de dollars et on anticipe qu’au rythme actuel
ils seraient de 1 100 milliards de dollars d’ici 20508.
Au sommet du G8 de décembre 2013, les chefs d’état de ces pays ont même
organisé parallèlement un Sommet de la Démence pour faire face au problème9. C’est que,
encore à ce jour, la compréhension de ces maladies étant bien imparfaite, il n’existe
toujours pas de traitement curatif pour aucun type de démence, ni même de traitement
palliatif reconnu pour retarder l’évolution de la maladie ou masquer les symptômes dans
bien des cas.
Les formes de DLFT, dont la svPPA, ont une prévalence de 15 par 100 000 chez les
45-65 ans6. On estime que la survie après le diagnostic est de 3 à 10 ans selon les formes,
ce qui est d’autant plus tragique que ces maladies frappent généralement plus tôt que les
autres démences, dans la population souvent encore sur le marché du travail.
Pour l’instant, le diagnostic précis de svPPA ou d’une autre forme de DLFT peut
prendre des années à être porté, puisqu’il peut prendre souvent jusqu’à cinq ans avant que
le patient chemine dans le système pour avoir éventuellement une rencontre avec un
spécialiste de 3e ligne et accès à des évaluations neuropsychologiques. Ces délais éliminent
malheureusement toute possibilité d’un traitement ou d’une prise en charge précoces.
De ces constats émerge l’importance de mieux comprendre la présentation
anatomique et pathologique de la svPPA si on veut en accélérer le diagnostic en première
ligne et espérer améliorer la prise en charge. Caractériser et quantifier les atteintes de la
matière grise et de la matière blanche pour comprendre la maladie et en établir la signature
précise et spécifique en imagerie permettraient d’atteindre ces objectifs. Faire une revue de
la littérature de la pharmacothérapie offerte à ces patients permettra de faire le point sur la
situation et de déterminer des pistes de solution pour le bénéfice des patients atteints de
svPPA et leurs proches.
4
État des connaissances
On connaît dorénavant, au-delà des informations cliniques mentionnées ci-dessus,
plusieurs données pathologiques, génétiques et anatomiques à des degrés variables sur la
svPPA.
Il est rapporté que la cascade pathologique en cause dans la svPPA implique
l’ubiquitine et le TDP-43, sans que la protéine tau ne soit impliquée (comme dans le cas de
la variante comportementale de la démence fronto-temporale ou de la variante non-fluente
de l’aphasie primaire progressive)7.
Au niveau génétique, les mutations les plus souvent associées à la svPPA seraient
au niveau du gène GRN (codant la granuline et responsable d’un quart des cas familiaux) et
possiblement au niveau du gène MAPT (microtubule-associated protein tau et responsable
de la moitié des cas familiaux), bien que la svPPA ne soit pas considérée une tauopathie7.
Ces anomalies génétiques ne sont pas considérées présentement dans le processus
diagnostique, mais pourraient avoir un intérêt prédictif ou pronostique lorsque des
médicaments changeant le cours de la maladie seront disponibles10.
Au point de vue anatomique, plusieurs études11-15 effectuées avec l’imagerie par
résonance magnétique (IRM) chez des patients svPPA ont démontré des zones avec une
atrophie corticale, principalement au niveau des régions antérieures des lobes temporaux,
bilatéralement mais préférentiellement du côté gauche. Ces trouvailles sont acceptées dans
la littérature au point où il s’agit désormais de critères reconnus pour appuyer le diagnostic
clinique de la svPPA. Des études utilisant la tomographie par émission de positons ont
illustré des zones d’hypométabolisme du glucose sur les mêmes territoires, en antérieur des
lobes temporaux10.
En revanche, la connaissance des dommages à la matière blanche est moins étendue,
alors qu’elle est essentielle pour la compréhension de la maladie, autant pour objectiver ses
atteintes que pour tenter d’en comprendre les mécanismes. Pour l’instant, seules quelques
études16-19 basées sur les données de peu de patients ont été publiées, rapportant pour la
plupart des diminutions d’anisotropie fractionnelle (FA), un marqueur de la diffusion dans
les axones témoignant de dommages micro-structurels, parfois dans les fascicules
longitudinaux postéro-inférieurs, parfois au niveau unciné et parfois au niveau cingulaire
5
postérieur. De plus, ces études ne sont pour la plupart que conçues avec des techniques
basées sur les régions d’intérêt, une approche peu sensible, plutôt qu’avec des techniques
de tractographie plus élaborées (entre autres du fait qu’elles utilisent une approche plus
sensible basée sur des statistiques concernant l’ensemble du cerveau plutôt que des régions
prédéterminées et analysées par voxel, en plus de corriger les anomalies d’alignement et de
lissage).
D’un point de vue thérapeutique, les données sont rares en ce qui a trait à la svPPA.
Il existe déjà des solutions intéressantes non-pharmacologiques pour donner un coup de
main au quotidien aux patients et à leurs proches (entre autres, des outils développés par
des orthophonistes et des neuropsychologues et misant sur les nouvelles technologies pour
pallier à certains troubles du langage20). Un survol de la littérature concernant les approches
pharmacologiques nous confirme qu’il n’y a toutefois actuellement aucune médication
reconnue pour la prise en charge de cette clientèle, ni pour traiter la maladie, ni pour en
ralentir l’évolution10. Les recommandations actuelles portent surtout sur le traitement
symptomatique pour pallier à certains troubles du comportement ou du langage. Rien ne
semble être mis de l’avant pour attaquer l’étiologie principale et ralentir ou renverser le
déclin cognitif. Nous avons ainsi décidé d’élargir nos recherches à l’ensemble des DLFT,
afin de trouver des pistes de solutions provenant de problèmes connexes. En effet, après des
avancées intéressantes dans la recherche sur la maladie d’Alzheimer ou de Parkinson,
plusieurs études ont exploré les effets potentiels de médications impliquant les réseaux
sérotoninergiques, dopaminergiques ou cholinergiques ou ont essayé des stimulants ou des
neuromodulateurs21 dans le spectre des DLFT. Il n’y a pas de consensus; la prise en charge
est donc controversée comme en témoigne cette statistique : une étude américaine a
découvert que 42% des médecins suivant ces patients ont prescrit un inhibiteur de
l’acétylcholinestérase et 11% ont eu recours à des bloqueurs de la NMDA, bien que les
autorités gouvernementales n’aient pas reconnu d’indications à cet égard22.
6
Objectifs du mémoire
Ce mémoire se penche donc sur deux problématiques reliées aux DLFT pour faire
le point sur les connaissances sur le sujet et tenter d’y contribuer.
En premier lieu, à des fins descriptives et dans un esprit diagnostique, nous avons
procédé à une étude transversale comparant des patients atteints de svPPA à des contrôles
sains appariés sur l’âge et le sexe, acquérant pour tous un examen d’IRM comprenant des
séquences de diffusion (diffusion tensor imaging), dans le but d’évaluer quantitativement,
par des techniques tractographiques et volumétriques, l’atteinte de la matière blanche ainsi
que de la matière grise. Notre hypothèse principale était que les atteintes de la matière grise
étaient bel et bien localisées davantage au lobe temporal gauche, mais que les atteintes de la
matière blanche étaient plus étendues, que ce soit par les répercussions extrinsèques de la
mort neuronale (par dégénérescence wallérienne) ou intrinsèquement par la propagation
d’un processus neuro-inflammatoire ou d’une cascade de protéines pathologiques par les
faisceaux de matière blanche.
En deuxième lieu, pour éclaircir la controverse sur la prise en charge
pharmacothérapeutique, nous avons procédé à une revue systématique avec méta-analyse
de la littérature pour recenser et analyser toutes les publications concernant la
pharmacothérapie dans la FTLD. Notre but était de déterminer s’il y avait un bénéfice
quelconque de la pharmacothérapie pour les fonctions cognitives ou pour pallier aux
symptômes neuropsychiatriques, en plus d’évaluer la sécurité chez ces patients des
différents médicaments testés. Notre hypothèse principale était que, bien que les médecins
suivant ces patients prescrivent fréquemment différents agents pharmacologiques, aucune
molécule n’a démontré d’effet bénéfique sur la cognition, mais que certaines peuvent
pallier certains symptômes de ces maladies.
7
Chapitre 2
Between-Group Study of Tractography and Volumetry in the Semantic Variant of Primary Progressive Aphasia
L.O. Bouchard1,2; M.A. Wilson1,2,3; S. Duchesne1,2 1 Radiology Department, Université Laval, Québec, QC, Canada
2 Quebec City Mental Health Institute, Québec, QC, Canada
3 Rehabilitation Department, Université Laval, Québec, QC, Canada Corresponding author: Louis-Olivier Bouchard Institut universitaire en santé mentale de Québec F-3548, 2601 de la Canardière Québec, QC Canada G1J 2G3 louis-olivier.bouchard.1@ulaval.ca (418) 663-5000 ext. 6709
8
ABSTRACT
Background: The semantic variant of primary progressive aphasia (svPPA) is a neurodegenerative
disease, a form of dementia mainly featuring language impairment.
Objectives: To characterize cerebral white matter and grey matter damage in svPPA.
Design: A between-group study comparing svPPA patients to healthy controls.
Setting and patients: Ten patients with svPPA assessed between 2011 and 2014 from a tertiary
reference center, and nine healthy matched controls.
Measurements: Using magnetic resonance imaging (MRI), we obtained grey matter volumetric
data with an automated segmentation tool. With diffusion tensor imaging (DTI) data, we extracted
fractional anisotropy (FA) values using a tract-based spatial statistics approach. We compared both
groups using the non-parametric Wilcoxon rank-sum test.
Results: Demographic data show that patients and controls were indeed comparable. Clinical data
show lower results in svPPA than controls in the MoCA and MMSE cognitive screening tests.
Tractographic data show impaired diffusion in svPPA patients, with FA mostly decreased in the
longitudinal, uncinate, cingulum and external capsule fasciculi. Volumetric data show significant
atrophy in svPPA patients, mostly in the left entorhinal, amygdale, inferior temporal, middle
temporal, superior temporal and temporal pole cortices, and bilateral fusiform gyri (p<0.001).
Limitations: DTI and TBSS, the techniques used in this study, have difficulties dealing with white
matter crossing fibers (lowering its sensibility to detect anomalies). Also, results are limited to the
currently available atlases.
Conclusions: SvPPA patients have significant white and grey matter degeneration, as shown on
MRI and DTI. These modalities have an important role to play in the differential diagnosis of
atypical dementia in the initial assessment of these patients.
9
Keywords: Dementia, semantic variant of primary progressive aphasia, white matter, grey matter,
magnetic resonance imaging, diffusion tensor imaging, fractional anisotropy, volumetry,
tractography
10
INTRODUCTION
Background/Rationale
Primary progressive aphasia23 is a neurodegenerative disorder whose prominent feature is
language impairment24, as opposed to early memory or motor deficits that are predominant in other
dementias. The semantic variant of primary progressive aphasia (svPPA), also called semantic
dementia, is one of its three variants, alongside progressive non-fluent aphasia and logopenic
progressive aphasia. These aphasias are all forms of frontotemporal lobar degeneration, a family of
diseases that represent 15 to 20% of all dementias25.
Clinically, features of svPPA include impairment in confrontation naming and single-word
comprehension, and may include surface dyslexia or dysgraphia, while repetition and speech
production is preserved23. Pathologically, svPPA appears to be associated with ubiquitin and TDP-
4326-30, and could be genetically linked to mutations in GRN and MAPT genes31-33. Anatomically,
changes in cortical grey matter (GM) have been well studied in svPPA, to the extent that imaging
can be used as a supportive criterion for the diagnostic. These findings include the presence on
magnetic resonance images (MRI) of cortical atrophy located predominantly in the anterior
temporal lobes, bilaterally but more severely in the left hemisphere, especially in its ventral and
lateral regions34-38.
Comparatively, the extent of white matter (WM) damage in svPPA has been much less
studied. Only a few reports 39-42 have addressed this topic, always in small patient groups. They all
made use of diffusion tensor imaging (DTI), a technique that allows the study of WM fiber tracts in
MRI, and from which one can extract maps of fractional anisotropy (FA) a scalar value indicating
the directionality of diffusion in axons, itself a proxy of microstructural damage in the brain43. The
studies reported lowered FA mostly in the left postero-inferior longitudinal, uncinate and posterior
cingulate fasciculi39-42.
Furthermore, all studies but one42 used a region-of-interest rather than a tract-based
statistical (TBSS)44 approach to the analysis of FA maps. The latter is a whole-brain skeleton-based
11
technique that is more sensitive than the ROI voxel-based technique, and hence that should provide
a better resolution of WM damage in the brains of svPPA patients.
We designed this study to increase our current knowledge of concurrent GM and WM
damage in svPPA in a larger study group, with the ultimate goal of using this information to
improve diagnosis and management of patients, as we know earlier diagnosis leads to earlier care
and decreased social difficulty45.
Objectives
The objectives of this study were to:
- Characterize (localize and quantify) cerebral WM lesions in patients with svPPA through
the tractographic studies of fractional anisotropy, compared to healthy controls;
- Characterize (localize and quantify) cerebral GM lesions in patients with svPPA, using
visual scales and automated cerebral segmentation volumetric tools, compared to healthy
controls.
METHOD
Ethics
The ethics committee of the Institut universitaire en santé mentale de Québec approved the
study (project #300-2012). Informed and free consent was obtained from each participant.
Study design
We used a between-group experimental design, in which we compared patients with a
diagnosis of svPPA to cognitively and neurologically healthy matched control subjects. All
participants were native French speakers from Quebec, Canada.
12
Setting and participants
Patients
All patients that fulfilled current svPPA clinical criteria23 and were being followed at
Clinique Interdisciplinaire de Mémoire (CHU de Québec, Quebec City, Canada) between August
2013 and August 2014 were invited to participate in the study. These patients all had a neurologist’s
diagnosis of svPPA. Additionally, a comprehensive neuropsychological battery was administered to
all participants. The most salient results of this battery include:
- mild to moderate word production and anomia difficulties measured with a picture naming task
(Boston Naming Test)46 and phonological, semantic, and free fluency tasks (MEC Protocole)47;
- impaired verbal and non-verbal single-word comprehension, assessed with visual and oral word-
picture matching tasks (BECLA battery)48 and semantic association of pictures (Pyramids and palm
trees test)49;
- spared repetition (TEFREP)50 and absence of apraxia of speech (PENO)51.
Some of the patients also had impaired object recognition (object decision and object
matching subtests, BORB battery)52 with or without surface dyslexia (regular and irregular word
and pseudoword reading test)53, while others presented with surface dyslexia with unimpaired
object recognition. They all were free of noticeable deficit on tests tapping other cognitive domains
such as non-verbal episodic memory, visuospatial abilities (Rey-Osterrieth Complex Figure Test)54,
and working memory (forward and backwards digit SPANs Wechsler Memory Scale IV)55.
Controls
Cognitively healthy controls were recruited among patients’ proxies and members of the
community, and were matched by age, gender, education and manual dominance to the svPPA
13
patients. They were screened via telephone for inclusion criteria by means of a structured
questionnaire. They then went through the same neuropsychological battery as patients.
All patients and controls had no history of either moderate or severe traumatic brain injury,
cerebro-vascular disease or intracranial surgery, encephalitis or meningitis. They had no history of a
significant psychiatric syndrome, alcoholism or drug addiction or unstable medical or metabolic
condition (e.g., uncontrolled diabetes, vitamin B12 or folic acid deficiency, hypothyroidism), nor
any history of learning or reading difficulties. They were also screened for MRI compatibility by
means of another questionnaire.
Data sources and measures
Above and beyond the tests previously mentioned, the demographic and clinical data
included the presence or absence of semantic dementia, age, gender, education (in years), manual
dominance, and global cognition assessed via the Montreal Cognitive Assessment Test (MoCA)56
and Mini-Mental State Examination (MMSE)57. Since both groups of participants were matched by
age, gender, education and manual dominance, we did not make further sensitivity analyses. It
would have been unlikely that the 10:9 ratio of patients and controls affected the volumetric or FA
data.
MRIs were acquired within four weeks of the clinical assessment using the Canadian
Dementia Imaging Protocol (www.cdip-pcid.ca) on a 3-Tesla magnetic resonance scanner (Philipps
Medical Systems, Best, The Netherlands) at the IRM Québec radiology clinic in Quebec City.
Specifically, the protocol included:
- an isotropic, 1mm3 3D T1-weighted sequence (TR = 7.3ms; TE = 3.3ms; Flip angle = 9 degrees);
and
- a 60-direction, 2mm3 isotropic diffusion imaging sequence (b-value = 1000).
For cerebral WM data, fractional anisotropy maps and tract-based statistics were obtained
following these steps. First, we converted our native data from DICOM files to the NIFTI format
14
using MRIcron (http://www.mricro.com/mricron). Then, we used FMRIB’s Diffusion Toolbox36 to
create a FA image from the DTI data of each subject. This included eddy current correction, brain
masking, and diffusion tensor model fitting. Afterwards, we used the FMRIB Software Library
5.0.659 to perform voxelwise analyses, following the TBSS approach44. There are five steps to this
process: 1) preprocessing; 2) nonlinear alignment of all FA images to a standard-space image
(FMRIB58_FA); 3) creation of the mean FA image and generation of its skeleton; 4) projection of
every pre-aligned FA image on the study-specific, template skeleton; and 5) generation of
voxelwise statistics with the Randomise tool, using the Threshold-Free Cluster Enhancement. The
final result was a brain map in which we could identify voxels that showed a statistically significant
difference in FA between patients and controls. We used the Atlasquery tool to localise these voxels
and provide numerical FA values for each subject in each of the 48 regions of interest (ROI)
identified in the ICBM-DTI-81 white matter labels atlas of the Laboratory of Brain Anatomical
MRI at Johns Hopkins University60.
For GM analysis, we used an automated segmentation tool (FreeSurfer)61 to segment the
3D T1-weighted image for all subjects, producing surfaces (mm), thicknesses (mm2) and volumes
(mm3) for the whole brain and 44 different cortical and subcortical structures according to their
atlases (DKT40 and ASEG63).
Statistical methods
For all demographic, clinical and imaging data, we opted for non-parametric Wilcoxon
rank-sum test to compare cases to controls, given the size of our sample and the robustness of the
test. We used a significance level of p-value = 0.05.
Reporting
This study follows the recommendation of the STROBE Statement64.
15
RESULTS
Demographic and clinical data
Nineteen participants (ten svPPA and nine controls) were recruited in this study. Every
subject that was screened happened to be eligible for the study. Analysis of demographic data
shows no significant difference between groups in age, gender, education, or manual dominance.
Clinical data show significantly lower results for svPPA patients at both the MoCA and the MMSE
tests, which was expected (see Table 1).
White matter analysis
Volumetric data showed that both groups had similar total WM volumes. Qualitative
assessment of diffusion data shows however multiple WM regions where svPPA patients have
significantly lower FA values, slightly more severely in the left hemisphere and temporal lobe.
Statistical analysis of numerical FA values shows 20 out of 48 tract regions with significantly lower
FA values in svPPA patients than in healthy controls. The most significantly impaired regions are
the right hippocampus and uncinate fasciculus tract regions, and the left external capsule and
superior longitudinal fasciculus (see Table 3 and Figure 2).
Grey matter analysis
Volumetric data show that both groups had indistinguishable total intracranial capacity and
total brain volumes. However, our group of svPPA patients had smaller subcortical and left
hemispheric cortical GM volumes. When looking at specific regions of interest, volumes are
consistently lower in svPPA patients than in healthy controls, reaching significance in 22 of the 33
studied regions. The most significant differences were noted in the left amygdala, entorhinal,
fusiform, inferior temporal, middle temporal, and superior temporal gyri, as well as in the temporal
poles and right fusiform grey matter (see Table 2 and Figure 1).
16
DISCUSSION
The semantic variant of primary progressive aphasia is a neurodegenerative disorder with
language impairment as its hallmark. Previous literature reported in some studies cortical GM
atrophy predominantly in the anterior temporal lobes, bilaterally but markedly in the left
hemisphere. As for WM damage, literature is scarcer and always based on a very small number of
patients. Only a few reports have described abnormal diffusion, in one or more of the left postero-
inferior longitudinal, uncinate, posterior cingulate fasciculi. Of these studies, only one used a more
sensible tract-based approach instead of the region-of-interest approach. In addition to validate the
characterisation of GM damage, it is principally to address the knowledge gap we observed
concerning the WM damage that we designed this study.
The first aim of our study was to characterize cerebral WM lesions in our cohort of svPPA
patients, when compared to cognitively healthy, age-matched control subjects. Using a tract-based
whole-brain approach, we found FA differences between our two groups, including statistically
lower FA in the right hippocampus and uncinate fasciculus tract regions, as well as the left external
capsule and superior longitudinal fasciculus tracts. Our second aim was to characterize cerebral GM
lesions in these same patients. Using volumetric tools, we found significant atrophy differences
between patients and controls. Notably, larger GM atrophy in svPPA patients was found in the left
amygdala, in the left entorhinal, fusiform, inferior temporal, middle temporal and superior temporal
gyri, left temporal pole and right fusiform gyrus.
Findings
Our GM results show a lateralisation of atrophy related to svPPA, as it is more severe in the
left hemisphere. We also found significant atrophy in the temporal lobes. These results are in line
with those previously described in literature53. This is also in accordance with fundamental
neuroimaging studies on the cerebral regions associated with language53, which have shown that the
17
left anterior temporal lobe is part of the semantic brain network, whereas other functions like syntax
would be processed more in the posterior temporal or frontal lobes.
Our WM results show a significant diminution of FA in different tracts, which is known to
demonstrate microstructural damages in these fasciculi. In accordance with some of the few reports
available, we have found impairment of the longitudinal fascicule and some of the uncinate
fasciculi39,41,42; however, we also found that the damage is more extensive and less clearly
lateralized than previously thought. Interestingly, WM damage is also more extensive than that of
the GM.
Hypotheses
Two different hypotheses have been proposed to explain the pattern of WM damage 39,65,66.
The first one posits an extrinsic cause. Following GM atrophy by neuronal loss, a process of
Wallerian degeneration or axonal degeneration begins, hence causing WM loss. Considering the
severity of GM damage that we also observe in our cohort, this hypothesis would explain why not
only tracts close to the left temporal lobe are impaired, but also other connecting tracts that are
located more distantly. The second hypothesis is an intrinsic direct axonal pathology67. This
explanation puts forward that intrinsic brain networks propagate WM disease, either by means of a
process of neuroinflammation or by a cascade of misfolded protein (prion-like)68.
Since our study is purely descriptive and cross-sectional, we cannot support firmly either
hypothesis. Longitudinal studies are needed to understand the evolution of GM and WM damage in
the brain, and hence confirm or infirm either hypothesis.
Strengths and limitations
Compared to similar studies in the literature, and given the prevalence of svPPA in the
population, the number of patients that we were able to recruit is a great strength of our study, even
though it still implied the use of non-parametric tests, with their inevitably lower statistical power.
Furthermore, our inclusion and exclusion criteria prevented confounding factors to alter the internal
validity of our results, while maximising participant recruitment. First, our patients had all received
18
a diagnosis of svPPA by neurologists in a tertiary reference center, according to the latest criteria23.
Secondly, measurements were not subject to recall, interviewer or detection biases as all subjects
were imaged with the same device and an identical protocol, and analysed similarly using an
algorithm that was completely blind to the status of the participant. Thirdly, the whole-brain TBSS
approach has many advantages such as correcting for misalignment, registration and smoothing, and
removing the need to determine a priori regions of interest for the analysis, which would imply the
use of a yet incomplete knowledge of connectomics.
On the other hand, there were some limitations arising from these same measurement
techniques. The first comes from the use of DTI, as it is known that the diffusion tensor model has
limited sensitivity to crossing fibres. Future studies could use High-angular-resolution diffusion
imaging (HARDI)69 to alleviate this problem. The second limitation related to the technique comes
from the choice of FA as our sole metric to evaluate WM microstructural changes. The use of
apparent diffusion coefficient (ADC), or axial, radial and mean diffusivity could have improved the
sensitivity of our study. It is indeed known that FA is a scalar value, and that different combinations
of diffusion tensor eigenvalues that have changed concurrently could generate identical FA values65.
However, FA is a known, oft-reported metric that captures the essential changes related to WM
integrity; using other techniques could only uncover further areas or a deeper level of WM
damages. Equally, TBSS will also be subject to reduced sensibility whenever fibers cross, for
similar reasons as mentioned previously. Finally, the atlas used to extract numerical values remains
composed of rather large structures that may end up masking small, sub-regional effects. The
Human Connectome Project70 aims to improve our knowledge of WM tracts and is currently
addressing this issue.
Conclusion
In summary, our study confirms known GM atrophy in svPPA patients, as well as validates
WM damage using a more sensitive and unbiased approach than previously reported. MRI and DTI
appear to have the potential to occupy an important place in the initial investigation of dementia,
19
and could eventually be very useful in the differential diagnosis of atypical dementia, to
characterize as precisely as possible the cerebral damage in one patient and to link it to a specific
disease through its imaging signature. Our results in a French-speaking population are comparable
with previous studies conducted in native speakers of other languages, mostly English. This means
that, in terms of external validity, language does not seem to be a concern. This could lead to
multicentric studies, without language barriers, to increase the number of subjects in future studies,
especially those on a longitudinal basis and those correlating clinical, anatomical and pathological
data. As well, to better understand notions of cerebral lateralization and brain plasticity, it would
also be of interest to include, as much as available, left-handed patients to assess if findings differ
for this group.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
LB: Study conception and design, analysis and interpretation of data, manuscript drafting, and
approval for publication.
MW: Contribution to conception and design of study, interpretation of data, critical revision of
manuscript for important intellectual content, and approval for publication.
SD: Contribution to conception and design of study, interpretation of data, critical revision of
manuscript for important intellectual content, and approval for publication.
All authors agree to be accountable for all aspects of the work.
20
Acknowledgements
MW gratefully acknowledges financial support from the Réseau Québécois de Recherche sur le
Vieillissement et Fonds de recherche du Québec ‐Société et culture # 168556). SD is a Research
Scholar from the Fonds de recherche du Québec – Santé.
21
Tables
Table 1. Demographic and clinical data of svPPA patients and controls
Healthy controls (n=9) SD cases (n=10) p-value Age 67 (67.1 ± 9.7) 68 (67.3 ± 9.1) 0.90 Gender (male) 77.8% 80.0% 0.94 School years 15 (15.9 ± 3.8) 16 (16.3 ± 5.0) 0.90 Manual dominance (right) 100% 100% - MoCA 25 (24.7 ± 1.8) 19 (18.9 ± 3.4) 0.001* MMSE 28 (27.8 ± 1.5) 24 (23.7 ± 3.8) 0.01* *Statistically significant Results are presented : median (mean ± standard deviation) Abbreviations : svPPA = semantic variant of primary progressive aphasia; MoCA = Montreal Cognitive Assessment; MMSE = Mini Mental State Examination. Table 2. White matter study: fractional anisotropy data of svPPA patients and controls
White matter tracts Healthy controls SD cases p-value Cerebellar peduncle (left inferior) 0.17 (0.17 ± 0.01) 0.16 (0.15 ± 0.02) 0.03* Cerebellar peduncle (left superior) 0.63 (0.63 ± 0.02) 0.58 (0.57 ± 0.04) 0.009** Cerebellar peduncle (middle) 0.06 (0.06 ± 0.00) 0.06 (0.06 ± 0.00) 0.003** Cerebellar peduncle (right inferior) 0.16 (0.16 ± 0.01) 0.15 (0.14 ± 0.02) 0.03* Cerebellar peduncle (right superior) 0.58 (0.57 ± 0.02) 0.54 (0.53 ± 0.04) 0.03* Cerebral peduncle (left) 0.67 (0.67 ± 0.02) 0.65 (0.64 ± 0.03) 0.07 Cerebral peduncle (right) 0.65 (0.65 ± 0.01) 0.63 (0.62 ± 0.03) 0.01* Cingulum (left hippocampus) 0.45 (0.46 ± 0.03) 0.41 (0.41 ± 0.02) 0.009** Cingulum (left cingulate gyrus) 0.58 (0.59 ± 0.03) 0.52 (0.52 ± 0.03) 0.001** Cingulum (right cingulate gyrus) 0.54 (0.54 ± 0.04) 0.49 (0.49 ± 0.03) 0.002** Cingulum (right hippocampus) 0.48 (0.49 ± 0.03) 0.43 (0.42 ± 0.02) 0.0004***Corona radiata (left anterior) 0.43 (0.42 ± 0.03) 0.41 (0.40 ± 0.03) 0.16 Corona radiata (left posterior) 0.47 (0.47 ± 0.02) 0.48 (0.48 ± 0.03) 0.36 Corona radiata (left superior) 0.49 (0.48 ± 0.02) 0.50 (0.51 ± 0.04) 0.25 Corona radiata (right anterior) 0.42 (0.42 ± 0.03) 0.40 (0.41 ± 0.02) 0.36 Corona radiata (right posterior) 0.47 (0.47 ± 0.02) 0.47 (0.47 ± 0.03) 0.81 Corona radiata (right superior) 0.47 (0.46 ± 0.02) 0.47 (0.48 ± 0.04) 0.36 Corpus callosum (body) 0.68 (0.67 ± 0.03) 0.61 (0.62 ± 0.03) 0.10 Corpus callosum (genu) 0.65 (0.64 ± 0.02) 0.60 (0.60 ± 0.03) 0.003** Corpus callosum (splenium) 0.74 (0.74 ± 0.02) 0.73 (0.73 ± 0.02) 0.08 Corticospinal tract (left) 0.36 (0.36 ± 0.02) 0.35 (0.35 ± 0.02) 0.25 Corticospinal tract (right) 0.35 (0.35 ± 0.01) 0.34 (0.34 ± 0.03) 0.87 External capsule (left) 0.43 (0.43 ± 0.02) 0.37 (0.37 ± 0.02) 0.0004***
22
External capsule (right) 0.40 (0.41 ± 0.02) 0.38 (0.38 ± 0.02) 0.009** Fornix 0.38 (0.37 ± 0.09) 0.23 (0.24 ± 0.08) 0.002** Fornix (left stria terminalis) 0.51 (0.51 ± 0.03) 0.46 (0.46 ± 0.02) 0.003** Fornix (right stria terminalis) 0.50 (0.50 ± 0.04) 0.47 (0.46 ± 0.04) 0.08 Internal capsule (left anterior limb) 0.55 (0.54 ± 0.03) 0.52 (0.53 ± 0.02) 0.12 Internal capsule (left posterior limb) 0.67 (0.67 ± 0.01) 0.68 (0.68 ± 0.04) 0.56 Internal capsule (left retrolenticular) 0.57 (0.57 ± 0.02) 0.55 (0.55 ± 0.02) 0.25 Internal capsule (right anterior limb) 0.55 (0.54 ± 0.02) 0.51 (0.52 ± 0.03) 0.28 Internal capsule (right posterior limb) 0.64 (0.64 ± 0.02) 0.64 (0.65 ± 0.03) 0.81 Internal capsule (right retrolenticular) 0.55 (0.54 ± 0.02) 0.54 (0.54 ± 0.02) 0.93 Longitudinal fasciculus (left superior) 0.50 (0.51 ± 0.02) 0.47 (0.47 ± 0.02) 0.0004***Longitudinal fasciculus (right superior) 0.49 (0.48 ± 0.02) 0.48 (0.47 ± 0.02) 0.07 Medial lemniscus (left) 0.29 (0.29 ± 0.01) 0.29 (0.29 ± 0.02) 0.74 Medial lemniscus (right) 0.29 (0.29 ± 0.01) 0.29 (0.30 ± 0.02) 0.74 Pontine crossing tract 0.27 (0.27 ± 0.01) 0.26 (0.27 ± 0.02) 0.16 Sagittal stratum (left) 0.53 (0.54 ± 0.03) 0.50 (0.50 ± 0.03) 0.05 Sagittal stratum (right) 0.52 (0.53 ± 0.03) 0.49 (0.49 ± 0.02) 0.009** Superior fronto-occipital fasciculus (left) 0.48 (0.48 ± 0.04) 0.51 (0.51 ± 0.04) 0.19 Superior fronto-occipital fasciculus (right) 0.45 (0.45 ± 0.03) 0.45 (0.46 ± 0.05) 0.93 Tapetum (left) 0.61 (0.58 ± 0.08) 0.53 (0.54 ± 0.06) 0.14 Tapetum (right) 0.56 (0.53 ± 0.08) 0.49 (0.49 ± 0.06) 0.19 Thalamic radiation (left posterior) 0.58 (0.57 ± 0.03) 0.53 (0.53 ± 0.03) 0.02* Thalamic radiation (right posterior) 0.56 (0.55 ± 0.03) 0.52 (0.53 ± 0.04) 0.32 Uncinate fasciculus (left) 0.48 (0.48 ± 0.04) 0.39 (0.39 ± 0.04) 0.002** Uncinate fasciculus (right) 0.47 (0.48 ± 0.04) 0.41 (0.40 ± 0.04) 0.0004***
* p < 0.05 ** p < 0.01 *** p < 0.001 Results are presented as median (mean ± standard deviation). Abbreviations : svPPA = semantic variant of primary progressive aphasia. Table 3. Grey matter study: volumetric data of svPPA cases and controls
Anatomical regions Healthy controls (n=9) svPPA cases (n=10) p-value Total intracranial 1464158 (1445800 ± 157761) 1499366 (1514466 ± 167622) 0.51 Total brain 1050500 (1046374 ± 94260) 945580 (1004021± 120333) 0.32 Brain (without ventricles) 1031325 (101514 ± 86458) 886917 (940821 ± 108507) 0.19 Left cortical grey matter 205579 (203100 ± 16743) 172081 (172812 ± 15797) 0.002** Right cortical grey matter 207062 (204623 ± 17927) 182927 (189819 ± 16269) 0.05 Total cortical grey matter 412642 (407724 ± 34596) 351598 (362632 ± 61407) 0.007** Left cortical white matter 215561 (215029 ± 21759) 190731 (201804 ± 30910) 0.21 Right cortical white matter 222124 (217554 ± 23147) 204918 (213994 ± 34068) 0.68 Total cortical white matter 437685 (432584 ± 44871) 395650 (415799 ± 64891) 0.51 Subcortical grey matter 58410 (59075 ± 5179) 48256 (49102 ± 6890) 0.003** Total grey matter 566007 (555670 ± 44013) 487394 (501518 ± 44157) 0.03*
23
Amygdala (left) 1318 (1362 ± 229) 878 (840 ± 317) 0.0008*** Amygdala (right) 1554 (1602 ± 287) 1066 (1087 ± 248) 0.001** Bankssts (left) 2505 (2362 ± 512) 1728 (1712 ± 366) 0.007** Bankssts (right) 2308 (2327 ± 426) 1908 (2052 ± 419) 0.19 Corpus callosum (Anterior) 822 (799 ± 223) 542 (496 ± 381) 0.08 Corpus callosum (Central) 340 (353 ± 80) 295 (271 ± 105) 0.16 Corpus callosum (Mid-Ant) 395 (442 ± 135) 289 (301 ± 116) 0.03* Corpus callosum (Mid-Post) 316 (329 ± 69) 213 (229 ± 121) 0.04* Corpus callosum (Posterior) 854 (805 ± 241) 702 (557 ± 424) 0.14 Entorhinal (left) 1077 (1127 ± 201) 633 (681 ± 157) 0.0004*** Entorhinal (right) 968 (938 ± 165) 621 (644 ± 140) 0.002** Fusiform (left) 9472 (9724 ± 1200) 5415 (5661 ± 1323) 0.0006*** Fusiform (right) 9350 (9354 ± 1430) 6514 (6691 ± 1031) 0.0008*** Hippocampal (left) 3636 (3678 ± 425) 2429 (2435 ± 659) 0.001** Hippocampal (right) 3761 (3666 ± 415) 2986 (3174 ± 694) 0.06 Inferior parietal (left) 10365 (10771 ± 1378) 9947 (9664 ± 1640) 0.19 Inferior parietal (right) 13239 (12956 ± 1717) 12312 (12388 ± 1495) 0.41 Inferior temporal (left) 9258 (9424 ± 791) 4688 (5155 ± 1238) 0.0003*** Inferior temporal (right) 8993 (8997 ± 985) 6753 (6653 ± 724) 0.001** Insula (left) 6465 (6220 ± 712) 4929 (5075 ± 654) 0.004** Insula (right) 6357 (6515 ± 859) 5488 (5746 ± 859) 0.07* Lingual (left) 6162 (6212 ± 649) 5244 (5436 ± 1114) 0.1 Lingual (right) 6155 (6082 ± 622) 5938 (6026 ± 724) 0.74 Middle temporal (left) 9948 (9943 ± 1314) 5091 (5151 ± 1137) 0.0003****Middle temporal (right) 9778 (10775 ± 1511) 8403 (8019 ± 1518) 0.004** Parahippocampal (left) 2066 (2136 ± 342) 1297 (1426 ± 390) 0.002** Parahippocampal (right) 1943 (2001 ± 260) 1437 (1417 ± 303) 0.001** Superior temporal (left) 10299 (10510 ± 1078) 6920 (6949 ± 1336) 0.0004*** Superior temporal (right) 10388 (10402 ± 1280) 7723 (8117 ± 1273) 0.005** Temporal pole (left) 2304 (2257 ± 212) 904 (943 ± 195) 0.0003*** Temporal pole (right) 2091 (2110 ± 162) 1132 (1220 ± 452) 0.002** Transverse temporal (left) 1051 (1042 ± 229) 885 (889 ± 119) 0.04* Transverse temporal (right) 730 (802 ± 223) 792 (778 ± 81) 0.36
* p < 0.05 ** p < 0.01 *** p < 0.001 Results are presented as median (mean ± standard deviation). Units are mm3. Abbreviations : svPPA = semantic variant of primary progressive aphasia; Bankssts = banks of the superior temporal sulcus.
24
Figures*
Figure 1. White matter maps: statistically significant difference of fractional anisotropy between svPPA patients and controls
Controls vs svPPA (p < 0.05) Z = -27 Z = -10 Z = 24 Z = 44
X = -47 X = -14 X = 14 X = 47
*Radiological convention states that the patient’s right is the left side of the image
25
Controls vs svPPA (p < 0.01)
Z = -27 Z = -10 Z = 24 Z = 44
X = -47 X = 14 X = 14 X = 47
Abbreviation : svPPA = semantic variant of primary progressive aphasia.
26
Figure 2. Grey matter images: examples of a svPPA patient compared to a control
Control
svPPA
Abbreviation : svPPA = semantic variant of primary progressive aphasia.
27
Chapitre 3
Effect of Pharmacotherapy in Frontotemporal Lobar Degeneration: A Systematic Review and Meta-analysis of Randomised Controlled Trials
L.O. Bouchard1,2; Y. Bodryzlova2, O. Potvin2, S. Duchesne1,2 1 Radiology Department, Université Laval, Québec, QC, Canada
2 Quebec City Mental Health Institute, Québec, QC, Canada
Corresponding author: Louis-Olivier Bouchard Institut universitaire en santé mentale de Québec F-3548, 2601 de la Canardière Québec, QC Canada G1J 2G3 louis-olivier.bouchard.1@ulaval.ca (418) 663-5000 ext. 6709
28
ABSTRACT
Background: Frontotemporal lobar degeneration (FTLD) is an important cause of dementia (2nd
under 65 years old, 4th overall), yet there is currently no medication approved to treat it or alleviate
its symptoms. No review has emerged recently to survey the field of clinical trials, which may bring
therapies into clinical use.
Objectives: To systematically review randomised clinical trials (RCT) assessing the effect of
pharmacotherapy on cognition and other dementia symptoms in patients with FTTLD.
Data sources: We searched for randomised controlled trials in the Cochrane Library, MedLine,
Embase and PsycArticles databases (up to date as of January 30th 2015).
Methodology: All RCTs comparing any pharmacotherapy with a placebo in patients with FTLD
(or any subgroup of FTLD) were included. Two independent reviewers conducted the data
extraction, as well as the quality assessment, using the Risk of Bias tool and the GRADE approach.
Meta-analysis was performed for studies comparing the same drug (or the same drug family if
applicable) and the same outcomes.
Results: We included nine RCTs, studying the effect of seven different drugs (paroxetine,
trazodone, bromocriptine, methylphenidate, galantamine, memantine x3, oxytocin). Two studies
(Vercelletto 2011 and Boxer 2013) met the criteria for meta-analysis (n=130), and showed no effect
of memantine on cognition (Mini-Mental State Examination (MMSE), standardized mean
difference (SMD) = -0.83, 95% CI [-0.34, 1.69], p=0.52) and on neuropsychiatric symptoms
(Neuropsychiatric Inventory (NPI), SMD=0.28, 95% CI [-0.51, 1.06], p=0.49). Safety/tolerability
was good in all eight studies reporting it. Although no study showed significant modification of
cognition, bromocriptine (Mean Length of Utterance (MLU), p=0.043), methylphenidate
(Cambridge Gamble Task (CGT), p=0.017), oxytocin (Emotion recognition and NPI, p<0.05) and
trazodone (NPI, p=0.028) seem to relieve one or two specific neuropsychiatric symptoms. Risk of
bias was considered low in six studies, high in one study, and unclear in one study.
29
Conclusions: No pharmacotherapy appear to significantly affect cognition in FTLD patients,
though they could have minor benefits on neuropsychiatric symptoms (e.g. behaviour, impulsivity,
emotion recognition), while being safe.
Keywords: Frontotemporal lobar degeneration, FTLD, Cognition, Pharmacotherapy, Systematic
Review, Meta-analysis
30
INTRODUCTION
The second most common cause of dementia under 65 years old71 is frontotemporal lobar
degeneration (FTLD), which includes three disease subtypes: frontotemporal dementia (FTD, also
called «behavioural variant») and two forms of primary progressive aphasia (PPA), the semantic
variant of primary progressive aphasia (svPPA, also called «semantic dementia») and nonfluent
progressive aphasia (nfPPA)72. Although these diseases are heterogeneous in terms of underlying
pathologies, they are usually clustered in the same family, given their shared frontotemporal
abnormalities and the associated common behavioural and language symptoms71,73.
There is currently no medication approved or recognized to treat any FTLD subtype.
Backed by interesting results in other dementias (e.g. Alzheimer’s, Parkinson), several trials have
explored the effects of dopaminergic, serotoninergic, cholinergic or stimulant medications74. More
recently, some data have emerged concerning the use of specific disease-modifying
pharmacotherapy, directly targeting the putative causal pathological cascade75.
The literature reporting these findings is both controversial and contradictory regarding the
effects on cognition and other symptoms (behaviour, mood). The last systematic reviews on the
topic76,77,78,79 were incomplete, or published prior to recent reports studying different molecules. It is
therefore necessary to conduct a systematic review and meta-analysis to help decide whether or not
there is currently a therapeutic option to offer to this population.
Objectives
By reviewing every randomised controlled trial (RCT) in FTLD, we will:
- Evaluate the effect of pharmacotherapy (any medication) on cognition;
- Evaluate the effect of pharmacotherapy on specific cognitive measures and
neuropsychiatric symptoms; and
- Evaluate tolerability and safety of pharmacotherapy.
31
METHODS
Study eligibility
We chose to include all studies and trial reports that enrolled patients within the entire
FTLD spectrum (FTD, svPPA, nfPPA). Interventions had to be RCT of any pharmacotherapeutic
compound compared to placebo, given the lack of approved standard of care. Global and specific
cognitive measures and neuropsychiatric symptoms (e.g. behaviour and mood) as well as safety and
tolerability had to be assessed as part of outcomes.
In order to be as exhaustive as possible, we did not set any exclusion criteria, such as setting
of the study, comorbidities, use of other medication or other forms of therapy (e.g. psychotherapy),
type of publication, year of release and language of publication, and duration of follow-up.
Design, sources and search strategy
We performed a systematic review of the literature using the Cochrane Library80,
MedLine81 (via PubMed), Embase82 and PsycArticles83 (via PsycNet), up to January 30th 2015. To
exemplify the process, the complete strategy for MedLine is described in Figure 3. We used
EndNote X4.0.2 (Thomson Reuters, Toronto, ON) to manage references. We followed the
methodology proposed by the Cochrane collaboration for systematic reviews84.
Study selection
After identifying studies within the aforementioned databases, two independent reviewers
(L.O.B., Y.B.) went through the study selection process. In the event of any disagreement, a third
reviewer (O.P.) made the final decision. First, reading the title and abstract, we removed duplicates
and screened for eligible articles, according to the criteria set above. For the eligible articles, we
then used the full text to ensure that they were indeed pharmacotherapy RCT addressing the right
population.
Data extraction
Each independent reviewer extracted salient data from the studies, using a standardised
Microsoft Excel form. In addition to the characteristic of each study (title, authors, publication
32
name, year, type of publication), we compiled demographic information of the study’s population
(sample size, age, sex, cognitive function at baseline) and the particular type of FTLD. We noted
the intervention (medication used and posology), duration of treatment and follow-up.
As a primary outcome, we were interested in the change in global cognitive function,
quantified with psychometric tests85 such as the Mini-Mental State Exam (MMSE)86, Clinician’s
Interview Based Impression of Change, with caregiver input (CIBIC-Plus)87, Clinician’s General
Impression of Change (CGIC)88, Cambridge Neuropsychological Test Automated Battery
(CANTAB)89 or Aphasia Quotient90. We selected as secondary outcomes the effect of
pharmacotherapy on specific cognitive measures and neuropsychiatric symptoms and were
compiled as available, including Neuropsychiatric Inventory91 (NPI), Frontal Behavior Inventory92
(FBI), Boston Naming Test93 (BNT), Cambridge Gamble Task94 (CGT), Cambridge Behavioural
Inventory95 (CBI), Emotion recognition96 or Mean length of utterance97. Also as a secondary
outcome, we then extracted data concerning adverse effects to gauge safety/tolerability.
Risk of bias
Once the data was independently extracted by both reviewers, we graded various potential
sources of bias (selection, performance, detection, attrition and reporting), as “low” or “high”, or
"unclear" if there was insufficient information to judge properly, in accordance with the Cochrane
Collaboration’s tool for assessing risk of bias98. This validity assessment was performed at the study
level.
Qualitative assessment
Following the recommendations of the GRADE Working Group99, we intended to rate the
quality of evidence (“very low”, “low”, “moderate” or “high”), using the GRADE approach, which
considers methodological quality, directness of evidence, heterogeneity, precision of effect
estimates and risk of publication bias.
33
Quantitative assessment and data synthesis
Once data were extracted, we intended to perform a meta-analysis using RevMan 5.2
software (The Nordic Cochrane Centre, Copenhagen, Denmark). We planned to analyse together all
studies with outcomes concerning cognition and dementia symptoms, namely those that
investigated the effect of the same drug (or at least the same drug family) and used comparable
outcomes. Although dementia scores are based on measurement scales, we were interested in the
global score and consequently intended to use the standardized mean difference (with 95% CI) as a
summary measure to compare studies, and conduct meta-analyses with the random-effects model.
For the same two outcomes, we also intended to identify and measure heterogeneity with a forest
plot as well as using the I2 test, classifying heterogeneity as considerable (75-100%), substantial
(50-90%), moderate (30-60%) or not important (0-40%)91. To address heterogeneity rated as
moderate or higher, we intended to conduct subgroup analyses to study the effect in specific FTLD
subgroups and drugs of the same pharmacologic family. Finally, we planned to assess the presence
of a publication bias by producing a funnel plot.
RESULTS
Study selection
We identified two hundred and eighty publications from the databases listed earlier. We
removed 44 duplicates, and after screening titles and abstracts, a further 189 non-pharmacological
studies as well as ten studies not including participants with FTLD. There remained 28 studies to be
assessed for eligibility. As 12 were not randomised controlled trials and seven were only title
registrations or protocols, only nine studies were finally included in our systematic review100-108
(Figure 4: Flow diagram of Study selection). When we compare to previous systematic reviews, we
see that they are either too old (lacking three to five of our included studies) or fail to include two
relevant studies while including two that do not fulfill the criteria.
34
Study characteristics
Of the nine included studies (all conducted between 2004 and 2013), seven were crossover-
designed RCTs100-106 and two were classic RCTs107-108. Seven were full publications and two were
presented as letters to the editors102,105. These trials studied participants with FTD100,101,106,107 (5/9),
PPA102,105 (2/9) or a combination (1 FTD + PPA104; 1 FTD + SD108). Patients had a mean age
between 61.7 and 66.8 years old. The number of participants ranged from 6 to 81 between studies.
Each study compared a single pharmacologic agent against a placebo: selective serotonin reuptake
inhibitor (paroxetine)100, tetracyclic antidepressant (trazodone)101, dopamine agonist
(bromocriptine)102, psychostimulant (methylphenidate)103, acetylcholinesterase inhibitor
(galantamine)104, neuromodulating agent (oxytocin)106 and NMDA antagonist (memantine)105,107,108.
The duration of the studies ranged between six to 52 weeks (with washouts of one to six weeks in
crossover RCTs), apart from two trials that studied the effect of a single dose103,106.
All but three studies 102,105,106 assessed both global cognition and specific cognitive measures
or neuropsychiatric symptoms were assessed. Only one study did not mention adverse events100
(Table 4).
Risk of bias
Of the nine included studies, six were rated as having a low overall risk of bias101,102,105-108,
one was rated as having a high risk104, and two were unclear100,103. Only three trials mentioned how
they did their sequence generation and only one wrote about allocation concealment. The risk of
bias associated with performance was low for all, irrelevant for detection risk of bias, and low or
unclear for reporting bias. On the attrition criteria, one study was considered at high risk because it
had lost half its patients104. (Table 8: Risk of bias)
Global cognition
Seven trials studied cognition. Two of them (galantamine and memantine) measured the
Aphasia quotient, and although both seemed to favour medication over placebo, none were actually
statistically significant. Three studies (memantine x 2 and galantamine) measured cognition using
35
the MMSE, with p-values ranging between 0.27 and 0.88. Results on CIBIC-Plus (memantine,
p=0.11) and CGIC (memantine, p=0.90) were not significant either. On CANTAB, one study
(paroxetine) showed statistically significant impairment with the medication on reversal learning
(p=0.05) and pattern recognition (p=0.02), and one study (methylphenidate) showed impairment on
spatial span with the medication, although not statistically significant (p=0.096).
The meta-analysis of the two studies107,108 on memantine that had comparable outcomes
gave standardized mean differences between placebo and medication on MMSE in the pooled
results (n=130) of -0.83 with a 95% confidence intervals ranging from -3.34 to 1.69 (p=0.52). A
forest plot did not seem relevant for only two studies. For the same reason, we did not proceed to
any subgroup analysis, even though we calculated an I2=68%, which would be considered
substantial heterogeneity (Table 5 and Table 7).
Specific cognitive measures and neuropsychiatric symptoms
Regarding specific dementia symptoms, eight trials studied at least one particular symptom.
Results on specific language tests (naming, fluency, narrative language for bromocriptine) were not
statistically significant, apart for mean length of utterance in favour of the medication (p=0.01).
However, BNT (with memantine) showed a significant deterioration with the medication (p=0.004).
CGT (with methylphenidate) showed improvement with the medication (p=0.017). NPI, CBI or FBI
(with paroxetine, galantamine and memantine) did not reach significance. In one study (oxytocin),
both anger recognition test (MD=10%, p<0.05) and NPI (MD=-2.7, p=<0.05) favoured significantly
the medication over the placebo.
Again, a meta-analysis was possible only for the same two memantine trials. The standardized mean
difference for the pooled results (n=130) was 0.28, with 95% confidence intervals from -0.51 to
1.06 (p=0.49, not statistically significant). With only two studies, we did not conduct a forest plot
(I2=79%) or subgroup analyses (Table 6 and Table 7).
36
Safety/Tolerability
The safety/tolerability outcome was available in eight studies. When present, adverse events (and
serious adverse events) were mostly equal between medication and placebo groups. One trial
specifically studied the effect of its drug (methylphenidate) on pulse and blood pressure (p=0.12
and 0.53, respectively). Another trial specifically studied the effect of its drug (memantine) on
cognitive side effects (6:1 ratio against medication with p=0.056) and psychiatric side effects (1:6
ratio against placebo with p=0.03). (Table 6)
Quality assessment
A meta-analysis was possible only with two studies 100,101, as they were the only trials that
studied the same medication (memantine) and had evaluated comparable outcomes. Therefore, the
GRADE methodology could not be fully applied to our systematic review. However, based on the
quality of the design (only double-blind RCTs), and the directness of evidence (outcomes
specifically measured), the evidence would seem of moderate to high quality. As we could not
statute on the heterogeneity (seemingly substantial but on only two trials), imprecision (confidence
intervals are rather large, which comes as no surprise considering the small samples in the studies)
and publication bias (a funnel plot was not indicated with only two studies), we were unable to
evaluate precisely the quality of evidence with the GRADE approach as planned.
DISCUSSION
FTLD is the second most common cause of dementia under 65 years old. Although there
have been several trials exploring different molecules, there is currently no medication approved to
treat patients suffering from any subtype of FTLD. We also found that this topic is controversial as
some physicians already use pharmacotherapy with these patients, despite the lack of consensus110.
To clearly assess the effect of pharmacotherapy on cognition and on specific cognitive measures or
37
neuropsychiatric symptoms, we decided to make a systematic review of the literature, ending up
with nine RCT and meta-analysing two of them.
The principal outcome we were interested in was global cognition. This systematic review
showed that there is currently no evidence of any effect whatsoever on cognition by
acetylcholinesterase inhibitors (for FTD and PPA) and NMDA blockers (for FTD, PPA and SD).
Actually, as for the NMDA blockers (in FTD and SD), the evidence gathered in the meta-analysis
tends to show that they have no effect on global cognition (SMD = -0.83, p=0.52). We had no
information on cognition for bromocriptine (a dopamine agonist) and oxytocin (a neuromodulating
agent).
The second outcome we studied is specific cognitive measures or neuropsychiatric
symptoms. We found some evidence that selective serotonin reuptake inhibitors (paroxetine,
p=[0.02-0.05]) and psychostimulants (methylphenidate, p=0.096) could be detrimental to
performance on some precise cognitive functions (such as reversal learning, pattern recognition and
spatial span) in FTD patients. The meta-analysis, although limited, shows that NMDA blockers (for
FTD and SD) have no effect on neuropsychiatric symptoms (NPI: mean difference = 0.28, p=0.49).
As well, the results of this systematic review show that there is no effect of paroxetine (on FTD) or
galantamine (on FTD and PPA) on symptomatic issues. Bromocriptine (on PPA) seems to improve
(p=0.043) the mean length of utterances (subtask of global language testing) but otherwise no
significant effect was observed on other language aspects. Methylphenidate also had no effect on
most behavioural symptoms associated with FTD, except for the gambling behaviour that appears to
be reduced (CGT: p=0.017). Evidence of the study on oxytocin (for FTD, single dose) shows a
benefit for anger recognition (p<0.05) and neuropsychiatric symptoms (NPI: mean difference = -
2.7, p<0.05). Memantine (for FTD and SD) was detrimental for naming objects (BNT: p=0.004).
As for the safety and tolerability outcome, all drugs studied in this systematic review were
well tolerated overall and did not globally differ from the placebo concerning adverse events and
safety.
38
Strengths and limitations
The strengths of this systematic review come from its methodology, including its double-
reviewer process. We thoroughly searched the Cochrane library, MedLine, Embase and
PsycArticles and included every RCT published to date that compared any pharmacotherapy to a
placebo, without regard to publication date, language, format or other exclusion criteria. It is
therefore the only exhaustive and up-to-date review on the topic. Additionally, the methodology of
this systematic review followed as carefully as possible the recommendation of the Cochrane
collaboration84 and the PRISMA group109, enhancing its quality.
The major limitation of this study comes from the small number of existing studies on the
topic, driving the fact that a meta-analysis was possible with only two studies. We were
consequently unable to proceed to the planned heterogeneity assessments and to evaluate the
presence of a publication bias. The overall strength of evidence (following the GRADE approach)
was hard to determine and is undoubtedly lowered by unknown variables that are usually
considered in this method. Furthermore, this report is limited by the quality of the included studies
themselves: the risk of bias was overall considered low, but one study (galantamine) had a high risk
of bias and two had an unclear risk (paroxetine and methylphenidate).
Conclusions
There is currently no approved medication to improve FTLD or even simply to address its
symptoms, neither is there a treatment recognized for a specific subgroup of FTLD. This systematic
review and meta-analysis demonstrated that pharmacotherapy trials have not yet shown any
statistically significant cognitive benefit for these patients, even though some earlier reports
expressed the view that serotoninergic, dopaminergic or NMDA pathways could be involved.
Pharmacotherapies targeting these pathways seem to have failed, except to modify performance on
specific cognitive measures (e.g. naming) and neuropsychiatric symptoms, in some cases.
39
Clinically, this systematic review should encourage some physicians to think twice before
prescribing off-label medication as is sometimes done (acetylcholinesterase inhibitors and NMDA
blockers are prescribed for FTLD in as much as 42% and 11% of patients, respectively)110. There is
indeed currently no evidence to support such practice to expect any effect on global cognition.
However, we have found that perhaps in precise situations pharmacotherapy could have its place to
address specific symptoms.
As for research application, the first conclusion is that we do need more randomised, placebo-
controlled trials for new molecules, or based on previous studies. For example, Oxytocin gave
interesting results with a single dose, and thus could benefit from being tested within a more
prolonged trial. Keeping in mind that FTLD refers to a broad spectrum of underlying pathologies,
the field would likely benefit from targeting specific subgroups in further trials111, in an effort to
directly address pathological cascades: tauopathy in FTD (a good example being the upcoming
leucomethylthionium, or LTMX, trial75) or TDP43/Ubiquitin dysfunction in svPPA. Recent trials
with such "disease-modifying" drugs failed to improve cognition in Alzheimer’s disease112,113, but
aetiological differences prevent us from drawing similar conclusions across diseases.
40
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
LB: Study conception and design, analysis and interpretation of data, manuscript
drafting, and approval for publication.
YB: Contribution to conception and design of study, analysis and interpretation of
data, approval for publication.
OP: Contribution to conception and design of study, critical revision of manuscript
for important intellectual content, and approval for publication.
SD: Contribution to conception and design of study, critical revision of manuscript
for important intellectual content, and approval for publication.
All authors agree to be accountable for all aspects of the work.
Acknowledgements
SD is a Research Scholar from the Fonds de recherche du Québec – Santé.
41
FIGURES
Figure 3 Example of a Search Strategy
- MedLine (via PubMed) : Advanced Search Builder, Title/Abstract, Filter : «Randomized Controlled Trial» = 35 results Search (((((((((Frontotemporal lobar degeneration[Title/Abstract]) OR
Frontotemporal degeneration[Title/Abstract]) OR Frontotemporal
dementia[Title/Abstract]) OR Semantic dementia[Title/Abstract]) OR
Semantic aphasia[Title/Abstract]) OR Nonfluent aphasia[Title/Abstract])
OR Nonfluent progressive aphasia[Title/Abstract]) OR Agrammatic
aphasia[Title/Abstract]) OR Agrammatic dementia [Title/Abstract]) OR
Primary progressive aphasia[Title/Abstract]
42
Figure 4 Flow diagram of Study selection
Records identified through database searching (n=280)
Records after duplicates removed (n=236)
Records screened (n=236)
Records excluded (n=208, 198 non-pharmacological studies, 10 studying non-
FTLD population)
Full-text articles assesed for eligibility (n=28)
Full-text articles excluded (n=19, 11 non-RCTs, 8 not published (title registration
or protocol))
Studies included in qualitative synthesis (n=9)
Studies included in quantitative synthesis (meta-analysis) (n=2)
43
TABLES
Table 4 Study characteristics
Source Design Duration Dx n Lost participa
nts
Rx* Family Cognition Age Male (%)
Deakin et al. (2004)
RCT-Crossover
9 weeks x2 (Washout : 5 w)
FTD 10 0 Paroxetine 40mg die
SSRI MMSE = 23.2 66.3 70.0
Lebert et al. (2004)
RCT-Crossover
6 weeks x2 FTD 31 5 (3 AE, 2 exclusion)
Trazodone 150-300mg die
Tetracyclic antidepressant
MMSE = 20.8 (8.3)
61.7 48.4
Reed et al. (2004)
RCT-Crossover
7 weeks x2 (Washout : 4w)
PPA 6 1 (AE) Bromocriptine 7.5mg TID
Dopamine agonist
Aphasia quotient = [49.1-83.8]
66.8 N/A
Rahman et al. (2006)
RCT-Crossover
Single dose (Washout : 1-2w)
FTD 8 0 Methylphenidate 40 mg
Pycho-stimulant
MMSE = 27 (1.7) 62 87.5
Kertesz et al. (2008)
RCT-Crossover
8 weeks (All patients tried Rx for 18 months before random.)
FTD (39%) + PPA
39 5 (5? AE) Galantamine 8-12mg BID
Acetyl-cholinesterase inhibitor
MMSE = 19 (7.1) 63.3 70.6
Johnson et al. (2010)
RCT-Crossover
26 weeks x2 (Washout : 6w)
PPA 18 9 (2 quit, 7 left area)
Memantine 10mg BID
NMDA blocker Aphasia quotient = 72.4
N/A N/A
Jesso et al. (2011)
RCT-Crossover
Single dose (Washout : 2w)
FTD 20 0 Oxytocin 24 IU intranasal
Neuro-modulator
MMSE = 23.4 (4.32)
64.4 N/A
Vercelletto et al. (2011)
RCT 52 weeks FTD 49 (Rx=23)
7 (6AE, 1 quit)
Memantine 10mg BID
NMDA blocker MMSE = 24.8 (3.2)
65.5 73.8
Boxer et al. (2013)
RCT 26 weeks FTD (79%) + SD
81 (Rx=39)
5 (3AE, 2 exclusion)
Memantine 10mg BID
NMDA blocker MMSE = 24.7 [21.3-29.1]
66 67.1
Abbreviations : RCT (Randomised controlled trial), Dx (Diagnosis), Pts (Patients), Rx (Medication), FTD (Frontotemporal dementia), PPA (Primary progressive aphasia), SD (Semantic dementia), AE (Adverse event), SSRI (Selectve serotonin re-uptake inhibitors), NMDA (N-methyl-D-aspartate), MMSE (Mini-mental state examination) * All trials compared a drug to a placebo.
44
Table 5 Effect of Pharmacotherapy on Global Cognition
Source Cognition assessment Cognitive test Results
Deakin (Paroxetine) CANTAB items Reversal learning (Impaired with Rx, p=0.05)
Pattern recognition (Impaired with Rx, p=0.02) Others : NSS
Lebert (Trazodone) MMSE MD = -3.6 (p=0.08) Reed (Bromocriptine) N/A
Rahman (Methylphenidate) CANTAB items Spatial span (Impaired with Rx, p=0.096)
Others : NSS
Kertesz (Galantamine) Aphasia quotient MMSE
Favors Rx but NSS NSS (p=0.88)
Johnson (Memantine) Aphasia quotient MD = 4.29 (NSS) Jesso (Oxytocin) N/A
Vercelletto (Memantine) MMSE CIBIC-Plus
MD = -2.6 (p=0.27) MD = 0.8 (p=0.11)
Boxer (Memantine) MMSE CGIC
MD = 0.1 (p=0.69) MD = 0 (p=0.90)
Abbreviations : CANTAB (Cambridge Neuropsychological Test Automated Battery), MMSE (Mini-mental state examination), CIBIC-Plus (Clinician’s Inverview-Based Impression of Change, with caregiver input), CGIC (Clinician’s Global Impression of Change), Rx (Medication), NSS (not statistically significant), MD (mean difference)
45
Table 6 Effect of Pharmacotherapy on Neuropsychiatric Symptoms/Specific cognitive measures and Safety/Tolerability
Source Neuropsychiatric symptoms/Specific cognitive measures Safety/Tolerability Neuropsy. tests Results AEs Results
Deakin (Paroxetine) NPI CBI
MD = 3.6 (p=0.74) MD = 2.6 (p=0.18)
N/A
Lebert (Trazodone) NPI MD = 16.9 (p=0.028) AEs 11 Rx, 3 P (all mild)
Reed (Bromocriptine)
Naming, Fluency, Narrative language
MLU : deteriorates with Rx (p=0.043) but less than with P (p=0.01). Other tests NSS.
AEs 1 with Rx (frustration)
Rahman (Methylphenidate)
CGT Improved with Rx (p=0.017) Pulse and BP
NSS : p=[0.12-0.53]
Kertesz (Galantamine) FBI Favours P but NSS AEs
SAE 4 Rx, 5 P 2 Rx, 1 P
Johnson (Memantine) N/A AEs None
Jesso (Oxytocin)
Anger recognition NPI FBI
MD=10% (p<0.05)
MD = -2.7 (p<0.05)
MD = -1.2 (p= 0.08)
AEs SAE
3 Rx, 7 P None
Vercelletto (Memantine)
NPI FBI
MD = -2.1 (p=0.21) MD = 3.4 (p=0.12)
Serious AEs AEs
Rx = 26% P = 31% Rx = 35% P = 39%
Boxer (Memantine) NPI BNT
MD = 2.2 (p=0.47) MD = 2.2 (p=0.004)
Cognitive Psychiatric
Rx vs P = 6:1 (p=0.056) Rx vs P = 1:6 (p=0.03)
Abbreviations : NPI (Neuropsychiatric Inventory), CBI (Cambridge Behavioural Inventory), CGT (Cambridge Gamble Task), BNT (Boston naming test), FBI (Frontal Behavioural Inventory), MD (Mean difference), AEs (Adverse events), SAE (Severe adverse events), MLU (Mean length of utterance), NSS (not statistically significant), Rx (Medication), P (Placebo)
46
Table 7 Meta-analyses of the effect of pharmacotherapy on cognition and on dementia symptoms
Table 7.1: Meta-analysis (Cognition)
Source
Weight
SMD (95%
CI)
p-value
HeterogeneityCognitive test Results
Vercelletto (Memantine)
MMSE MD = -2.6 (p=0.27)
37.7% -0.83 (-3.34 to 1.69)
0.52
I2 = 68%
Boxer (Memantine)
MMSE MD = 0.1 (p=0.69)
62.8%
Table 7.2: Meta-analysis (Dementia symptoms: NPI)
Source
Weight
SMD (95% CI)
p-value
Heterogeneity Neuropsychiatric test Results
Vercelletto (Memantine)
NPI MD = -2.1 (p=0.21)
37.7% 0.28 (-0.51 to 1.06)
0.49
I2 = 79%
Boxer (Memantine) NPI MD = 2.2
(p=0.47) 62.8%
Abbreviations: MMSE (Mini-mental state examination), MD (Mean difference), SMD (Standardized mean difference), CI (Confidence interval), I2 (Heterogeneity test), NPI (Neuropsychiatric Inventory).
47
Table 8 Risk of bias within studies with Cochrane's Collaboration tool
Source Selection Performance Detection Attrition Reporting
Summary Author Sequence generation
Allocation concealment
Blinding Blinding Incomplete Selective reporting
Deakin Unclear Unclear Low Low Low Unclear Unclear Lebert Unclear Unclear Low Low Low Low Low Reed Unclear Unclear Low Low Low Low Low Rahman Unclear Unclear Low Low Low Unclear Unclear Kertesz Low Unclear Low Low High Low High Johnson Unclear Unclear Low Low Low Low Low Jesso Unclear Unclear Low Low Low Low Low Vercelletto Low Low Low Low Low Unclear Low Boxer Low Unclear Low Low Low Low Low
48
Chapitre 4 : Conclusion
Rappel des objectifs
Ce mémoire a permis de se pencher sur deux sujets complémentaires, l’aspect
diagnostique et l’aspect thérapeutique, dans un contexte de dégénérescence lobaire fronto-
temporale (FTLD) et plus spécifiquement de la variante sémantique de l’aphasie primaire
progressive (svPPA).
En premier lieu, si les connaissances acquises à ce jour au niveau de la localisation
anatomique des atteintes de la matière grise dans la svPPA sont nombreuses114-118, les
études qui portent sur les atteintes de la matière blanche sont rares. Donc, l’étude des
atteintes de la matière blanche méritait une attention plus particulière, avec davantage de
patients et avec une méthodologie plus robuste, entre autres grâce aux techniques
tractographiques qui tirent davantage profit de l’imagerie de diffusion (diffusion tensor
imaging). Que les futurs travaux longitudinaux en imagerie ou les études pathologiques
démontrent que les atteintes de la matière blanche soient secondaires aux dommages
neuronaux ou en lien avec une propagation d’un processus pathologique intrinsèque à la
matière blanche, il apparaît en tous les cas évident que l’étude de la matière blanche au
moment de porter le diagnostic de svPPA permet d’ajouter à la fois de la sensibilité et de la
spécificité à l’étude de la matière grise, pour bien identifier la signature anatomique de la
maladie. Ceci a conduit à l’étude (cf. Chapitre 2) comparant des patients atteints de svPPA
à des sujets sains sur une base tractographique et volumétrique pour mieux caractériser les
lésions cérébrales secondaires à la maladie.
Deuxièmement, une étude approfondie pour connaître l’état actuel de l’art était
nécessaire considérant le manque de cohésion dans les approches thérapeutiques
pharmacologiques en svPPA et en FTLD en général, ainsi que l’inexistence d’un consensus
clair recommandant une molécule, alors que sur le terrain plusieurs cliniciens prescrivent
tout de même différents médicaments à cette clientèle. De cette façon, la controverse
entourant l’usage de pharmacothérapie en FTLD a été le sujet de la revue systématique
49
avec méta-analyse de l’ensemble des essais cliniques randomisés contrôlés avec placebo où
les patients étaient atteints d’une forme ou l’autre de FTLD.
Résumé des résultats
Notre étude intitulée «Between-Group Study of Tractography and Volumetry in the
Semantic Variant of Primary Progressive Aphasia» a permis de mesurer l’anisotropie
fractionnelle de la matière blanche chez 10 patients atteints de svPPA et neuf contrôles
sains appariés pour l’âge et le sexe. Cette technique permet de déceler des atteintes micro-
structurelles. Nous démontrons ainsi des altérations principalement au niveau du fascicule
longitudinal supérieur gauche, de la capsule externe gauche, du cingulum droit et des
fascicules uncinés. Dans l’ensemble, nous avons remarqué une prédominance des atteintes
à l’hémisphère gauche et au niveau du lobe temporal. Cette étude nous a aussi permis de
valider les résultats de volumétrie corticale décrits dans la littérature. L’atrophie était
encore une fois plus marquée à l’hémisphère gauche, plus particulièrement au niveau de
l’amygdale et des gyri entorhinal, fusiforme, temporal inférieur, temporal moyen et
temporal supérieur, de même qu’au niveau des pôles temporaux bilatéralement et au niveau
du gyrus fusiforme droit. Ces trouvailles ont ainsi permis de mieux caractériser les atteintes
cérébrales présentes chez les patients atteints de la svPPA.
Ceci confirme notre hypothèse initiale, qui était que, si les atteintes de la matière
grise sont particulièrement marquées à gauche et en temporal, les atteintes de la matière
blanche sont plus étendues. Considérant les deux grandes hypothèses tentant d’expliquer les
dommages de la matière blanche, soit une atteinte secondaire par dégénérescence
wallérienne suite à une atteinte de la matière grise, soit une atteinte primaire par un
processus dégénératif intrinsèque de la matière blanche, le fait que les dommages de la
matière blanche soient plus extensifs que ceux de la matière grise peuvent faire pencher la
balance du côté de l’atteinte intrinsèque. Ces résultats alimentent donc la réflexion sur ce
sujet et encouragent la recherche longitudinale pour pouvoir éventuellement statuer sur la
question. Il est permis de croire qu’une meilleure compréhension du processus
50
pathologique pourrait orienter à la fois les méthodes diagnostiques, pour bien cerner la
maladie et l’identifier plus précocement, et les recherches thérapeutiques.
Notre seconde étude, «Effect of Pharmacotherapy in Frontotemporal Lobar
Degeneration: A Systematic Review and Meta-analysis of Randomised Controlled Trials»,
a répertorié les neuf essais cliniques randomisés contrôlés publiés à ce jour. Ces études
portent sur sept médicaments différents chez des patients atteints d’une forme ou l’autre de
FTLD, que ce soit la variante comportementale (ou démence fronto-temporale) ou les
aphasies primaires progressives (dont la svPPA). Deux études seulement ont pu être
incluses dans la méta-analyse. Il en ressort qu’aucune médication n’a été démontrée
efficace pour améliorer les fonctions cognitives de façon globale. Il a été démontré
toutefois que certains médicaments ont des effets positifs intéressants pour s’attaquer
spécifiquement à un symptôme neuropsychiatrique ou améliorer une mesure cognitive
précise, dont la bromocriptine pour la production du langage, le methylphénidate pour les
comportements impulsifs, l’oxytocine pour la reconnaissance de certaines émotions et le
trazodone pour certains symptômes neuropsychiatriques comme les difficultés dans la
reconnaissance des émotions ou l’impulsivité.
Ces résultats confirment notre hypothèse initiale, qui était qu’aucune médication
actuellement disponible n’améliore la cognition, bien que plusieurs soient tentées par les
cliniciens, mais que certaines situations, certaines molécules peuvent pallier à certains
symptômes.
Il semble donc que, s’il n’y a pas d’indication d’utiliser quelque substance que ce
soit dans une visée thérapeutique globale chez cette clientèle, il pourrait être avantageux, et
ce de façon personnalisée, d’utiliser certaines molécules ciblant certains symptômes pour
améliorer la qualité de vie du patient et de ses proches.
Au terme de cette revue systématique avec méta-analyse, notre recommandation aux
cliniciens serait donc d’éviter de prescrire quelque médication que ce soit pour des visées
cognitives chez les patients souffrant d’une FTLD, si ce n’est que pour s’attaquer à certains
symptômes précis, comportementaux entre autres.
Il ressort aussi de cette revue qu’il sera crucial, à l’avenir, de bien catégoriser les
patients par pathologie précise avant de procéder à des études pharmacologiques. De même,
51
nous constatons que pour l’instant l’avenue la plus prometteuse pour aider à court terme les
patients est davantage au niveau d’outils neuropsychologiques et orthophoniques.
Perspectives
Ce mémoire a permis d’amener un éclairage nouveau sur nos questions initiales, et
permet également d’en soulever de nouvelles, et ainsi amène quelques pistes de réflexion
intéressantes.
Considérons d’abord le point de vue diagnostique. L’enjeu est ici, bien entendu, de
mieux caractériser les atteintes présentes chez les patients souffrant de la svPPA, à la fois
pour améliorer l’efficacité et la spécificité du processus diagnostique et pour améliorer la
compréhension de la pathophysiologie de la maladie.
Diagnostiquer précisément le sous-type de démence et le plus tôt possible dans la
maladie permettrait, d’une part, une prise en charge précoce ciblée du patient et de ses
proches. Il faut rappeler que le temps d’établissement d’un diagnostic est souvent de
plusieurs années (environ 4-6 ans dans nos milieux) après la première présentation clinique,
ce qui limite de façon drastique toute intervention et adaptation possible pour le patient
comme pour ses proches aidants. D’autre part, un diagnostic précis et précoce pourrait
permettre de recruter davantage de patients dans des catégories bien délimitées pour les
études futures.
Au moment du diagnostic différentiel, l’idéal serait donc d’avoir, en plus des
critères cliniques, une signature para-clinique spécifique, principalement via l’imagerie,
pour distinguer la svPPA des autres DLFT et des autres démences. Le patron que nous
avons décrit (atteintes du fascicule longitudinal supérieur gauche, de la capsule externe
gauche, du cingulum droit et des fascicules uncinés, ainsi qu’atrophie de l’amygdale et des
gyri entorhinal, fusiforme, temporal inférieur, temporal moyen et temporal supérieur, de
même qu’au niveau des pôles temporaux bilatéralement et au niveau du gyrus fusiforme
droit) semble sensible à la svPPA. Une plus grande généralisation à un échantillon plus
large est nécessaire, permettant également de déterminer la spécificité de ce patron. De
même, l’inclusion, autant que faire se peut, de patients gauchers permettrait d’évaluer si
52
leurs résultats diffèrent au niveau de la latéralisation des trouvailles, ce qui pourrait aussi
faire avancer la compréhension de la plasticité cérébrale et de ses anomalies dans les
démences. Des efforts sont en cours, par exemple, dans le projet du connectome humain119
pour obtenir des atlas plus fins de la matière blanche, ce qui augmenterait du coup la
spécificité des études ultérieures. En plus de sa disponibilité, l’IRM nous apparaît toujours
comme la modalité de choix en raison des études de diffusion qu’elle permet. Toutefois,
pour en tirer pleinement profit, il sera intéressant de suivre les avancées de méthodes de
plus en plus sophistiquées, dont le high angular resolution diffusion imaging (HARDI)120,
qui prendrait mieux en considération les nombreuses zones de croisement de fibres dans la
matière blanche, et de valider la pertinence de mesurer d’autres paramètres de diffusion
(par exemple, en plus de l’anisotropie fractionnelle, on pourrait utiliser les diffusivités
moyenne, axiale et radiale).
Améliorer la compréhension de la physiopathologie de la maladie pourrait ensuite,
on l’espère, se traduire par des progrès dans le développement de nouvelles thérapies. Des
études longitudinales pour comprendre l’évolution de la maladie pourrait apporter des
éléments de réponse, entre autre, quant aux mécanismes de propagation des dommages
cérébraux. Dans le cas de la matière blanche, par exemple, ces études permettraient
possiblement de savoir si les dommages axonaux sont secondaires à la dégénérescence
wallérienne (suite à la mort du neurone)121 ou secondaires à un processus intrinsèque (par
neuro-inflammation ou par une cascade de protéines pathologiques)122. Par ailleurs, bien
que les coûts et l’accessibilité en limitent parfois l’usage, la tomographie par émission de
positron (TEP) pourrait aussi contribuer à la compréhension de la maladie, en évaluant le
métabolisme de certaines zones cérébrales123. De plus, si les études sur l’imagerie
amyloïdienne dans la maladie d’Alzheimer sont concluantes, peut-être que des composés
radio-pharmaceutiques permettront éventuellement de faire de l’imagerie-tau ou de
l’imagerie-TDP. Les études pathologiques post-mortem et les avancées de la génétique
auront aussi assurément une contribution intéressante à apporter dans la compréhension de
la maladie et pour valider la corrélation entre les modèles cliniques et radiologiques par
rapport aux pathologies sous-jacentes.
53
D’un point de vue thérapeutique, on distinguera d’une part les visées curatives et
d’autre part les visées palliatives. D’un point de vue curatif, vraisemblablement l’objectif
ultime, du moins pour cesser la progression de la pathologie, il y a encore du chemin à
parcourir. Les études testant des molécules affectant la neurotransmission (voies
sérotoninergiques, dopaminergiques, NMDA) se sont soldées par des échecs et il apparaît
donc peu probable que la solution vienne de ce côté. C’est dans ce contexte qu’on s’est
tourné vers les disease-modifying agents (molécules qui visent à changer le cours de la
maladie plutôt que pallier ou masquer les symptômes). En maladie d’Alzheimer, les deux
premières études sur ce type de molécules, anti-amyloïde dans ce cas, n’ont pas montré de
bénéfice (bapineuzumab et solanezumab en 2014)124,125, mais les pathologies sous-jacentes
étant différentes en DLFT, il sera intéressant de voir la contribution de médicaments par
exemple anti TDP-43 ou anti-tau (comme le leucomethylthionium ou LMTX)126.
Pour ce qui est de pallier aux symptômes, il existe déjà des solutions intéressantes
non-pharmacologiques pour donner un coup de main au quotidien aux patients et à leurs
proches (entre autres, des outils développés par des orthophonistes et des
neuropsychologues et misant sur les nouvelles technologies pour pallier à certains troubles
du langage127). Au niveau pharmacologique, notre revue systématique a fait ressortir la
pertinence de certaines molécules pour s’attaquer à un symptôme ou un paramètre cognitif
précis, ce qui peut tout de même amener une amélioration considérable de la qualité de vie.
D’autres études, que ce soit en prolongeant l’essai (dans le cas de l’ocytocine par exemple),
en ciblant davantage la population étudiée (en allant au-delà du phénotype clinique dans les
critères d’inclusion) ou en utilisant de nouvelles molécules, pourraient répondre à ce
besoin.
En résumé, ce mémoire aura permis de mieux caractériser les atteintes cérébrales
chez les patients atteints de svPPA, décrivant l’atrophie de la matière grise et les dommages
davantage étendus de la matière blanche, dans le but d’en faciliter le diagnostic et de faire
progresser la compréhension de la maladie. Ce mémoire aura aussi fait le point sur la
pharmacothérapie en DLFT, confirmant qu’il n’y a pas de médication reconnue pour traiter
les troubles cognitifs de ces patients, mais faisant ressortir la pertinence de certaines
molécules pour pallier à certains symptômes.
54
Références
1. Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-V). American
Psychiatric Association. 2013; 2. Dementia numbers in Canada. Alzheimer Society of Canada. http://alzheimer.ca. 2015; 3. Arcand M, Hébert R. Précis pratique de gériatrie. EDISEM, 3e édition. 2007; 4. NINDS Frontotemporal Dementia Information Page. National Institute of Neurological
Disorders and Stroke. http://ninds.nih.gov. 2015; 5. Barkhof F, Fox NC, Bastos-Leite AJ, Scheltens P. Neuroimaging in Dementia. Springer.
2011; 6. Mesulam, MM. Primary progressive aphasia. Ann Neurol. 2001, 49: 425–432; 7. Gorno-Tempini ML, Hillis AE, Weintraub S. Classification of primary progressive aphasia
and its variants. Neurology. 2011; 76:1006-14; 8. Frontotemporal Disorders: Information for patients, families and caregivers. National
Institutes of Health. 2015; 9. Fox NC, Petersen RC. The G8 Dementia Research Summit – a starter for eight? The
Lancet. 2013, 382: 1968-1969; 10. Riedl L, Mackenzie IR, Forstl H. Frontotemporal lobar degeneration: current perspectives.
Neuropsychiatr Dis Treat. 2014, 10:297-310; 11. Gorno-Tempini ML, Dronkers NF, Rankin KP. Cognition and anatomy in three variants of
primary progressive aphasia. Ann Neurol. 2004, 55: 335–346; 12. Mummery CJ, Patterson K, Price CJ. A voxel-based morphometry study of semantic
dementia: relationship between temporal lobe atrophy and semantic memory. Ann Neurol. 2000, 47: 36–45;
13. Mesulam M, Weineke C, Rogalski E. Quantitative template for subtyping primary progressive aphasia. Arch Neurol. 2009, 66: 1545–1551;
14. Galton CJ, Patterson K, Graham K. Differing patterns of temporal atrophy in Alzheimer's disease and semantic dementia. Neurology. 2001, 57: 216–225;
15. Rosen HJ, Kramer JH, Gorno-Tempini ML. Patterns of cerebral atrophy in primary progressive aphasia. Am J Geriatr Psychiatry. 2002, 10: 89–97;
16. Galantucci S, Tartaglia MC, Wilson SM. White matter damage in primary progressive aphasias: a diffusion tensor tractography study. Brain. 2011, 134: 3011-3029;
17. Whitwell JL, Avula R, Senjem ML. Gray and white matter diffusion in the syndromic variants of frontotemporal dementia. Neurology. 2010: 74(16): 1279-1287;
18. Agosta F, Scola E, Canu E. White matter damage in frontotemporal lobar degeneration spectrum. Cereb Cortex. 2012, 22(12): 2705-2714;
19. Mahoney CJ, Malone IB, Ridgway GR. White matter tract signatures of the progressive aphasia. Neurobiol Aging. 2013, 34(6): 1687-1699;
20. Gravel-Laflamme K, Routhier S, Macoir J. Non-pharmacological therapies of language deficits in semantic dementia. Gériatrie et psychologie neuropsychiatrie du vieillissement. 2012, 10(4): 427-4236;
21. Rabinovici G, Miller B. Frontotemporal lobar degeneration: epidemiology, pathophysiology, diagnosis and management. CNS Drugs. 2010: 375-398;
22. Lopez-Pousa S, Calvo-Perxas L, Lejarreta S. Use of Antidementia Drugs in Frontotemporal Lobar Degeneration. American Journal of Alzheimer’s Disease and Other Dementias. 2012; 27-4:260-266.
23. Gorno-Tempini ML, Hillis AE, Weintraub S. Classification of primary progressive aphasia and its variants. Neurology. 2011; 76:1006-14;
55
24. Mesulam, MM. Primary progressive aphasia. Ann Neurol. 2001, 49: 425–432; 25. Riedl L, Mackenzie IR, Forstl H. Frontotemporal lobar degeneration: current perspectives.
Neuropsychiatr Dis Treat. 2014, 10:297-310; 26. Hodges JR, Davies RR, Xuereb JH. Clinicopathological correlates in frontotemporal
dementia. Ann Neurol. 2004, 56: 399–406; 27. Mesulam M, Wicklund A, Johnson N. Alzheimer and frontotemporal pathology in subsets
of primary progressive aphasia. Ann Neurol. 2008, 63: 709–719; 28. Davies RR, Hodges JR, Kril JJ. The pathological basis of semantic dementia. Brain. 2005,
128: 1984–1995; 29. Grossman M, Wood EM, Moore P. TDP-43 pathologic lesions and clinical phenotype in
frontotemporal lobar degeneration with ubiquitin-positive inclusions. Arch Neurol. 2007, 64: 1449–1454;
30. Nestor PJ, Balan K, Cheow HK. Nuclear imaging can predict pathologic diagnosis in progressive nonfluent aphasia. Neurology. 2007, 68: 238–239;
31. Snowden JS, Pickering-Brown SM, Mackenzie IR. Progranulin gene mutations associated with frontotemporal dementia and progressive non-fluent aphasia. Brain. 2006, 129: 3091–3102;
32. Mesulam M, Johnson N, Krefft T. Progranulin mutations in primary progressive aphasia: the PPA1 and PPA3 families. Arch Neurol. 2007, 64: 43–47;
33. Van Swieten J, Spillantini MG. Hereditary frontotemporal dementia caused by tau gene mutations. Brain Pathol. 2007, 17: 63–73;
34. Gorno-Tempini ML, Dronkers NF, Rankin KP. Cognition and anatomy in three variants of primary progressive aphasia. Ann Neurol. 2004, 55: 335–346;
35. Mummery CJ, Patterson K, Price CJ. A voxel-based morphometry study of semantic dementia: relationship between temporal lobe atrophy and semantic memory. Ann Neurol. 2000, 47: 36–45;
36. Mesulam M, Weineke C, Rogalski E. Quantitative template for subtyping primary progressive aphasia. Arch Neurol. 2009, 66: 1545–1551;
37. Galton CJ, Patterson K, Graham K. Differing patterns of temporal atrophy in Alzheimer's disease and semantic dementia. Neurology. 2001, 57: 216–225;
38. Rosen HJ, Kramer JH, Gorno-Tempini ML. Patterns of cerebral atrophy in primary progressive aphasia. Am J Geriatr Psychiatry. 2002, 10: 89–97;
39. Galantucci S, Tartaglia MC, Wilson SM. White matter damage in primary progressive aphasias: a diffusion tensor tractography study. Brain. 2011, 134: 3011-3029;
40. Whitwell JL, Avula R, Senjem ML. Gray and white matter diffusion in the syndromic variants of frontotemporal dementia. Neurology. 2010: 74(16): 1279-1287;
41. Agosta F, Scola E, Canu E. White matter damage in frontotemporal lobar degeneration spectrum. Cereb Cortex. 2012, 22(12): 2705-2714;
42. Mahoney CJ, Malone IB, Ridgway GR. White matter tract signatures of the progressive aphasia. Neurobiol Aging. 2013, 34(6): 1687-1699;
43. Basser PJ, Pierpaoli C. Microstructural and physiological features of tissues elucidated by quantitative-diffusion-tensor MRI. J Magnet Reson B. 1996, 111(3): 209-219;
44. Smith SM, Jenkinson M, Johansen-Berg D. Tract-based spatial statistics: Voxelwise analysis of multi-subject diffusion data. Neuroimage. 2006, 31: 1487-1505;
45. Routhier S, Gravel-Laflamme K, Macoir J. Non-pharmacological therapies for language deficits in the agrammatic and logopenic variants of primary progressive aphasia : a literature review. Gériatrie et Psychologie Neuropsychiatrie du Vieillissement. 2013, 11: 87-97;
46. Kaplan E, Goodglass H, Weintraub S. Boston Naming Test. Lee & Febiger. Philadelphia. 1983;
56
47. Joanette Y, Ska B, Côté H. Protocole Montréal d'Evaluation de la Communication MEC. Ortho Editions. Isbergues, France. 2004;
48. Gauthier C, Jean C, Macoir J. Batterie BECLA. Université Laval, Québec. 2005; 49. Howard D, Patterson K. Pyramids and palm trees: A test of semantic access from pictures
and words. Thames Valley Test Company. Bury St. Edmunds, England. 1992; 50. Bourgeois-Marcotte J, Wilson MA, Forest M. TEFREP: Épreuve de répétition de phrases
en franco-québécois. Développement, validation et normalisation [TEFREP: Sentence repetition test in French-Canadian. Development, validation and norms]. Canadian Journal on Aging. 2015, 34(3): 1-6;
51. Joanette Y, Ska B, Belleville S. Évaluation neuropsychologique dans la démence de type Alzheimer: Un compromis optimal. L’Année Gérontologique. 1995, 3: 69–83;
52. Riddoch MJ, Humphreys GW. BORB: Birmingham Object Recognition Battery. Erlbaum. Hove. 1993;
53. Wilson MA, Joubert S, Ferré P. The role of the left anterior temporal lobe in exception word reading: Reconciling patient and neuroimaging findings. NeuroImage. 2012, 60: 2000-2007;
54. Meyers J, Lange D. The Complex Figure: A recognition subtest. The Clinical Neuropsychologist. 1994, 8: 153-166;
55. Wechsler D. Wechsler Memory Scale. Fourth Edition, San Antonio, Texas. 2009; 56. Nasreddine Z, Phillips N, Bédirian V. The Montreal Cognitive Assessment (MoCA): a brief
screening tool for mild cognitive impairment. Journal of the American Geriatrics Society. 2005, 53(4); 695-699;
57. Folstein M, Folstein S, McHugh P. Mini-mental state: a practical method of grading the cognitive state of patients for the clinician. Journal of Psychiatric. 1975, 12; 189-198;
58. Behrens TEJ, Johansen-Berg H, Woolrich MW. Probabilistic diffusion tractography with multiple fibre orientations. What can we gain? Proc Natl Acad Sci. 2004, 101(36): 1335-13340;
59. Woolrich MW, Jbabdi S, Patenaude B. Bayesian analysis of neuroimaging data in FSL. Neuroimage. 2009, 45: 173-186;
60. Oishi K, Faria A, Jiang H. Atlas-based whole brain white matter analysis using large deformation diffeomorphic metric mapping: Application to normal elderly and Alzheimer’s disease participants. Neuroimage. 2009: 46(2): 486-499;
61. Dale AM, Fischl B, Sereno MI. Cortical surface-based analysis, Segmentation and surface renconstruction. Neuroimage. 1999, 9: 179-194;
62. Desikan R, Ségonne F, Fischl B. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage. 2006, 31: 968-980;
63. Fischl B, Salat D, Busa E. Whole Brain Segmentation: Automated Labeling of Neuroanatomical Structures in the Human Brain. Neuron. 2002, 33: 341-355;
64. Von Elm E, Altman DG, Egger M. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Ann Intern Med. 2007, 147(8): 573-577;
65. Aghakhanyan G, Martinuzzi A, Frijia F. Brain white matter involvement in hereditary spastic paraplegias: analysis with multiple diffusion tensor indices. American Journal of Neuroradiology. 2014, 35(8): 1533-1538;
66. Worker A, Blain C, Jarosz J. Diffusion tensor imaging of Parkinson’s disease, multiple system atrophy and progressive supranuclear palsy: a tract-based spatial statistics study. PLoS One. 2014, 9(11): e114167;
67. Dagher A, Nagano-Saito A. Functional and anatomical magnetic resonance imaging in Parkinson’s disease. Mol Imaging Biol. 2007, 9(4): 234-242;
57
68. Heneka MT, Carson MJ, El Khoury J. Neuroinflammation in Alzheimer’s disease. Lancet Neurol. 2015, 14(4): 388-405;
69. Descoteaux M, Angelinao E, Fitzgibons S. Apparent diffusion coefficients from high angular resolution diffusion imaging: estimation and applications. Magn Reson Med. 2006, 56(2): 395-410;
70. Van Essen DC, Ugurbil K, Auerbach E. The Human Connectome Project: a data acquisition perspective. Neuroimage. 2012, 62(4): 2222-2231.
71. Riedl L, Mackenzie IR, Forstl H. Frontotemporal lobar degeneration : current perspectives. Neuropsychiatric Disease and Treatment. 2014; 10 : 297-310;
72. Gorno-Tempini ML, Hillis AE, Weintraub S. Classification of primary progressive aphasia and its variants. Neurology. 2011; 76:1006-14;
73. Galantucci S, Tartaglia MC, Wilson SM. White matter damage in primary progressive aphasias : a diffusion tensor tractography study. Brain. 2011; 134:3011-29;
74. Freedman M. Frontotemporal dementia : Recommendations for therapeutic studies, designs, and approaches. Canadian Journal of Neurological Sciences. 2007; 34:118-24;
75. NHS. Leuco-methylthioninium (LMTX) for behavioural variant frontotemporal dementia - first line. Horizon Scanning Centre – National Institutes for Health Research. 2013;
76. Kaye ED, Petrovic-Poljak A, Verhoeff NPLG. Frontotemporal dementia and pharmacologic interventions. Journal of Neuropsychiatry and Clinical Neurosciences. 2010; 22:19-29;
77. Huey E, Putnam K, Grafman J. A systematic review of neurotransmitter deficits and treatments in frontotemporal dementia. Neurology. 2006; 66-1:17-22;
78. De Gloria Portugal M, Marinho V, Laks J. Pharmacological treatment of frontotemporal lobar degeneration : systematic review. Revista Brasileira de Psiquiatria. 2011;33-1:81-90;
79. Nardell M, Tampi RR. Pharmacological treatments for frontotemporal dementias : a systematic review of randomized controlled trials. American Journal of Alzheimer’s Disease and Other Dementias. 2014;29-2:123-132;
80. The Cochrane Collaboration. The Cochrane Library. John Wiley & Sons. 2013; 81. PubMed. MedLine. National Center for Biotechnology Information; 82. Embase. Elsevier; 83. American Psychological Association. PsyArticles. PsycNet. 2014; 84. Higgin JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions.
The Cochrane Collaboration. Version 5.1.0, March 2011; 85. Sheehan B. Assessment scales in dementia. Therapeutic advances in neurological
disorders. 2012; 5:349-58; 86. Folstein M, Folstein S, McHugh P. ‘Mini-Mental State’: a practical method for
grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975, 12 : 189-198;
87. Schneider L, Olin J, Doody R. Validity and reliability of the Alzheimers Disease cooperative study – clinician global impression of change. Alzheimer Dis Assoc Disord. 1997, 11: S1-S12;
58
88. Schneider L, Olin J. Clinical global impressions if Alzheimer’s clinical trials. 1996, 8(2); 277-288;
89. Robbins TW, James M, Owen AM. Cambridge Neuropsychological Test Automated Battery (CANTAB): a factor analytic study of a large sample of normal elderly volunteers. Dementia. 1994, 5(5): 266-281;
90. Kertesz A, Poole E. The aphasia quotient: the taxonomic approach to measurement of aphasic disability. Can J Neurol Sci. 1974, 1(1); 7-16;
91. Cummings JL, Mega M, Gray K. The Neuropsychiatric Inventory: comprehensive assessment of psychopathology in dementia. Neurology. 1994, 44(12): 2308-2314;
92. Blair M, Kertesz A, Davis-Faroque N. Behavioural measures in frontotemporal lobar dementia and other dementias: the utility of the frontal behavioural inventory and the neuropsychiatric inventory in a national cohort study. Dement Geriatr Cogn Disord. 2007; 23(6): 406-415;
93. LaBarge E, Edwards D, Knesevich JW. Performance of normal elderly on the Boston Naming Test. Brain Lang. 1986, 27(2): 380-384;
94. Zois E, Kortlang N, Vollstadt-Klein S. Decision-making deficits in patients diagnosed with disordered gambling using the Cambridge Gambling task. Brain Behav. 2014, 4(4): 484-494;
95. Wedderburn C, Wear H, Brown J. The utility of the Cambridge Behavioural Inventory in neurodegenerative disease. J Neurol Neurosurg Psychiatry. 2008, 79(5): 500-503;
96. Tottenham N, Tanaka JW, Leon AC. The NimStim set of facial expressions: judgents from untrained research participants. Psychiatry Res 2009; 168: 242-249;
97. Flege JE, Brown WS. Effects of utterane position on English speech timing. Phonetica. 1982, 39(6): 337-357;
98. Higgins JPT, Altman DG, Gotzsche PC. The Cochrane Collaboration’s tool for assessing risk of bias in randomized trials. BMJ. 2011; 343;
99. Guyatt GH, Oxman AD, Vist GE. GRADE : an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. 2008, 336: 924-6;
100. Deakin JB, Rahman S, Nestor PJ. Paroxetine does not improve symptoms and impairs cognition in frontotemporal dementia: a double-blind randomized controlled trial. Psychopharmacology 2004; 4:400-8;
101. Lebert F, Stekke W, Hasenbroekx. Frontotemporal dementia : A randomized, controlled trial with Trazodone. Dementia and Geritatric Cognitive Disorders 2004; 17 : 355-359;
102. Reed DA, Johnson NA, Thompson C. A clinical trial of bromocriptine for treatment of primary progressive aphasia. Annals of Neurology 2004; 5:750;
103. Rahman S, Robbins TW, Hodges JR. Methylphenidate ('Ritalin') can ameliorate abnormal risk-taking behavior in the frontal variant of frontotemporal dementia. Neuropsychopharmacology 2006; 3:651-8;
104. Kertesz A, Morlog D, Light M. Galantamine in frontotemporal dementia and primary progressive aphasia. Dementia and Geriatric Cognitive Disorders. 2008; 2:178-85;
105. Johnson NA, Rademaker A, Weintraub S. Pilot trial of memantine in primary progressive aphasia. Alzheimer Disease and Associated Disorders. 2010; 3:308;
59
106. Jesso S, Morlog D, Ross S. The effects of oxytocin on social cognition and behavior in frontotemporal dementia. Brain. 2011; 9:2493-501;
107. Vercelletto M, Boutoleau-Bretonnière C, Volteau C. Memantine in behavioral variant frontotemporal dementia : negative results. Journal of Alzheimer’s disease. 2011; 4:749-759;
108. Boxer AL, Knopman DS, Kaufer DI. Memantine in patients with frontotemporal lobar degeneration: a multicentre, randomised, double-blind, placebo-controlled trial. Lancet neurology. 2013; 2:149-56;
109. Moher D, Liberati A, Tetzlaff J. Preferred Reporting Items for Systematic Reviews and Meta-Analyses : The PRISMA Statement. The Prisma Group. 2009;
110. Lopez-Pousa S, Calvo-Perxas L, Lejarreta S. Use of Antidementia Drugs in Frontotemporal Lobar Degeneration. American Journal of Alzheimer’s Disease and Other Dementias. 2012; 27-4:260-266;
111. Barkhof F, Fox NC, Bastos-Leite AJ. Neuroimaging in dementia. Springer. 2011; 278p.;
112. Salloway S, Sperling R, Fox NC. Two Phase 3 Trials of Bapineuzumab in Mild-to-Moderate Alzheimer’s Disease. The New England Journal of Medicine. 2014;370:322-33;
113. Doody RS, Thomas RG, Farlow M. Phase 3 Trials of Solanezumab for Mild-to-Moderate Alzheimer’s Disease. The New England Journal of Medicine. 2014; 370:311-321.
114. Gorno-Tempini ML, Dronkers NF, Rankin KP. Cognition and anatomy in
three variants of primary progressive aphasia. Ann Neurol. 2004, 55: 335–346; 115. Mummery CJ, Patterson K, Price CJ. A voxel-based morphometry study of
semantic dementia: relationship between temporal lobe atrophy and semantic memory. Ann Neurol. 2000, 47: 36–45;
116. Mesulam M, Weineke C, Rogalski E. Quantitative template for subtyping primary progressive aphasia. Arch Neurol. 2009, 66: 1545–1551;
117. Galton CJ, Patterson K, Graham K. Differing patterns of temporal atrophy in Alzheimer's disease and semantic dementia. Neurology. 2001, 57: 216–225;
118. Rosen HJ, Kramer JH, Gorno-Tempini ML. Patterns of cerebral atrophy in primary progressive aphasia. Am J Geriatr Psychiatry. 2002, 10: 89–97;
119. Van Essen DC, Ugurbil K, Auerbach E. The Human Connectome Project: a data acquisition perspective. Neuroimage. 2012, 62(4): 2222-2231;
120. Descoteaux M, Angelinao E, Fitzgibons S. Apparent diffusion coefficients from high angular resolution diffusion imaging: estimation and applications. Magn Reson Med. 2006, 56(2): 395-410;
121. Dagher A, Nagano-Saito A. Functional and anatomical magnetic resonance imaging in Parkinson’s disease. Mol Imaging Biol. 2007, 9(4): 234-242;
122. Heneka MT, Carson MJ, El Khoury J. Neuroinflammation in Alzheimer’s disease. Lancet Neurol. 2015, 14(4): 388-405;
123. Riedl L, Mackenzie IR, Forstl H. Frontotemporal lobar degeneration : current perspectives. Neuropsychiatric Disease and Treatment. 2014; 10 : 297-310;
124. Salloway S, Sperling R, Fox NC. Two Phase 3 Trials of Bapineuzumab in Mild-to-Moderate Alzheimer’s Disease. The New England Journal of Medicine. 2014;370:322-33;
60
125. Doody RS, Thomas RG, Farlow M. Phase 3 Trials of Solanezumab for Mild-to-Moderate Alzheimer’s Disease. The New England Journal of Medicine. 2014; 370:311-321;
126. NHS. Leuco-methylthioninium (LMTX) for behavioural variant frontotemporal dementia - first line. Horizon Scanning Centre – National Institutes for Health Research. 2013.
127. Gravel-Laflamme K, Routhier S, Macoir J. Non-pharmacological therapies of language deficits in semantic dementia. Gériatrie et psychologie neuropsychiatrie du vieillissement. 2012, 10(4): 427-4236.
top related