University of Groningen. Visual hallucinations in Parkinson's disease Meppelink, Anne Marthe

Transcription

1 University of Groningen Visual hallucinations in Parkinson's disease Meppelink, Anne Marthe IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 2011 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Meppelink, A. M. (2011). Visual hallucinations in Parkinson's disease: clinical and fmri studies Groningen: s.n. Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date:

2 Visual Hallucinations in Parkinson s Disease; clinical and fmri studies

3 The work described in this thesis was performed at the Department of Neurology and Neuro Imaging Center of the University Medical Center Groningen, The Netherlands. Printing of this thesis was financially supported by: School of Behavioural and Cognitive Neuroscience (BCN) University of Groningen (RuG) Cover: Artistic modulation of a drawing of Leonardo da Vinci, combined with a brain activation image from this thesis and stills from the noise movies by Frans Cornelissen. Printed by: Gildeprint, Enschede Copyright c A. M. Meppelink, Groningen 2011 ISBN ISBN (digitaal)

4 Visual Hallucinations in Parkinson s Disease; clinical and fmri studies Proefschrift ter verkrijging van het doctoraat in de Medische Wetenschappen aan de Rijksuniversiteit Groningen op gezag van de Rector Magnificus, dr. E. Sterken, in het openbaar te verdedigen op woensdag 13 april 2011 om uur door Anne Marthe Meppelink geboren op 25 juni 1981 te Drachten

5 Promotor: Copromotores: Beoordelingscommissie: Prof. dr. K.L. Leenders Dr. T. van Laar Dr. B.M. de Jong Prof. dr. J.A. den Boer Prof. dr. R.A.C. Roos Prof. dr. A. Aleman

6 Contents Contents i 1 General introduction Parkinson s disease Basal ganglia circuits and cortical involvement Complex visual hallucinations Visual hallucinations in Parkinson s disease Visual processing and attention Therapeutic options for treatment of VH Brain imaging methods Outline of the thesis Visual recognition in PD with VH Abstract Introduction Patients and Methods Results Discussion Conclusions Attention and perception in PD with VH Abstract Introduction Methods Results Discussion Conclusion Impaired visual processing in PD with VH Abstract i

7 ii CONTENTS 4.2 Introduction Methods Results Discussion Conclusion Acknowledgements Cortical grey matter changes in PD Abstract Introduction Methods Results Discussion Conclusions Acknowledgements Cholinergic modulation during fmri Abstract Introduction Methods Results Discussion Apomorphine in PD with VH Abstract Introduction Patients and Methods Results Discussion Hallucinations in visual deprivation Abstract Case Report Discussion Acknowledgements General Discussion Introduction Bottom-up processing Top-down visual processing and attention Parkinson s disease and visual hallucinations

8 CONTENTS iii 9.5 Modulation VH in PD: from phenomenology to functional anatomy Conclusions Summary and conclusion 119 Samenvatting 125 Bibliography 133 Dankbrein 155

9

10 Chapter 1 General Introduction 1

11 2 CHAPTER 1. GENERAL INTRODUCTION This thesis topic is visual hallucinations (VH) in Parkinson s disease (PD), studied clinically and with functional magnetic resonance imaging (fmri). In this chapter, we will provide a brief introduction with background information on PD, basal ganglia circuits, VH in general, VH in PD, the visual system and attention. 1.1 Parkinson s disease PD is a neurodegenerative disorder primarily affecting the neuromelanin containing dopaminergic neurons in the substantia nigra, that project to the striatum (Lang and Lozano, 1998). Clinically, it is characterized typically by motor symptoms like bradykinesia, akinesia, tremor, rigidity and gait disturbances. In addition, non-motor symptoms, like depression, sleep disorders, autonomic dysfunction, cognitive impairment and VH widely occur in PD. The latter is the main topic of this thesis and raises the question of cortical involvement in PD (see below). PD is the second most common neurodegenerative disease after Alzheimer s disease and has a prevalence of approximately 1.8 in 1000 in the European population, but about 10 times higher in the elderly (von Campenhausen et al., 2005). Several genes have been identified that can lead to familial PD (Tan and Skipper, 2007), but this genetic form of PD only constitutes a minor proportion of the total PD patient population. The cause of the disease in sporadic cases is unknown, but several pathophysiological mechanisms have been proposed. The generally accepted hypothesis is that PD results from an interaction between inherited predisposition and environmental or endogenous toxic agents, resulting in mitochondrial respiratory failure, oxidative stress and subsequent cell death of nigral neurons (Schapira et al., 1992). Neurodegeneration in PD might also result from a decreased activity of the ubiquitin-proteasome system, which degrades misfolded or excess protein. Accumulation of the protein α-synuclein and formation of α-synuclein inclusion bodies, called Lewy bodies, is one of the hallmarks of PD. 1.2 Basal ganglia circuits and cortical involvement The degeneration of dopaminergic neurons of the substantia nigra pars compacta (SNc) and, to a lesser extent, of the ventral tegmental area (VTA) in PD, results in reduced dopamine in the basal ganglia (BG). The BG play an important role in motor, cognitive and affective behavioral functions. The mechanism by which the BG contribute to these functions, seems to be through

12 1.3. COMPLEX VISUAL HALLUCINATIONS 3 the selection of an appropriate response in a particular context and, in parallel, the suppression of inadequate responses (Redgrave et al., 1999). The functional-anatomical organization of the BG, with several feedback loops between nuclei in the BG and between the BG, thalamus and cortex, is in accordance with its role in complex behaviors. The BG receive input from the whole cortex and project, via the thalamus, back to the cortex, mainly the frontal lobe. They form a complex network of parallel and segregated cortico-basal ganglia-thalamo-cortical loops that can be clustered into three main functional categories; sensorimotor, cognitive and limbic. The main input structure of the BG is the striatum, encompassing the caudate nucleus, putamen and nucleus accumbens. These input nuclei receive dopaminergic modulation from the SNc (nucleus caudatus and putamen) and the VTA (nucleus accumbens), both localized in the ventral mesencephalon. The striatum projects to the output nuclei of the BG, the internal globus pallidus and the substantia nigra pars reticulata (SNr), via a direct pathway and via an indirect pathway. This indirect pathway also encompasses intrinsic nuclei of the BG; the external globus pallidus and the subthalamic nucleus. Both the internal globus pallidus and the SNr have an inhibitory connection to the thalamus. Dopamine from the mesencephalon that is released in the striatum stimulates the direct pathway via D1 dopamine receptors and inhibits the indirect pathway via the D2 dopamine receptors. The net result is a disinhibition of the thalamus, leading to activation of the premotor and prefrontal cortices and subsequent activation of behavioral output (see figure 1.1). In PD, less dopamine is delivered to the striatum, resulting in reduced stimulation of the direct pathway and reduced inhibition of the indirect pathway. The consequence of these effects is an increased inhibition of the thalamus and thus reduced motor or behavioral output. 1.3 Complex visual hallucinations VH have been defined as involuntary visual perceptions in the waking state without external visual stimulation (Collerton et al., 2005). VH can be simple, characterized by the absence of form, or complex, with a clearly defined specific form. Complex VH can occur in many different pathological conditions like PD, schizophrenia, narcolepsy, hallucinogen-induced states, epilepsy and eye disease (Charles Bonnet Syndrome, CBS), amongst others (Manford and Andermann, 1998). Although the underlying pathology is different for every condition, complex hallucinatory experiences are phenomenologically quite

13 4 CHAPTER 1. GENERAL INTRODUCTION Figure 1.1: Direct and indirect striatal output pathways and the influence of dopamine on these routes, represented in a semi-sagittal scheme of the cerebral cortex and the basal ganglia. Abbreviations: Acb, nucleus accumbens; Caud, caudate nucleus; GPe, external segment of the globus pallidus; GPi, internal segment of the globus pallidus; MC, primary motor cortex; MD, mediodorsal thalamic nucleus; O, occipital cortex; P, parietal cortex; PFC, prefrontal cortex; Put, putamen; sc, central sulcus; SNC, substantia nigra, pars compacta; SNR, substantia nigra, pars reticulata; STN, subthalamic nucleus; T, temporal cortex; VA, ventral anterior thalamic nucleus; VL, ventral lateral thalamic nucleus. (Groenewegen, 2009) similar. In line with this, similar cortical activation patterns during VH have been demonstrated in schizophrenia, CBS and PD (Silbersweig et al., 1995; Ffytche et al., 1998; Kataoka et al., 2008). Disturbances at disparate sites and mutual connections between sites in a distributed network (Manford and Andermann, 1998), including visual cortices, prefrontal cortex and subcortical structures, involved in visual processing, attention and reality monitoring, might explain this. Phenomenology and pathophysiology of VH in PD are discussed below. 1.4 Visual hallucinations in Parkinson s disease VH are common in PD with a prevalence of approximately 30 percent (Barnes and David, 2001). VH in PD typically comprise complex visual, commonly

14 1.4. VISUAL HALLUCINATIONS IN PARKINSON S DISEASE 5 moving images lasting for seconds to minutes. Animals, people and objects define the categories of images that are often perceived by hallucinating PD patients (Fenelon et al., 2000; Mosimann et al., 2006). Some have defined minor forms of hallucinations, which consist of a sensation of presence, a sensation of a sideways passage and illusions, in which an external stimulus is perceived but misinterpreted (Fenelon et al., 2000; Barnes and David, 2001; Fenelon et al., 2008). All above mentioned categories of hallucinations can occur in PD and often overlap, making it likely that they are related. Though mostly these VH are non-threatening and insight is retained (Fenelon et al., 2000; Barnes and David, 2001), in 80 percent they progress to hallucinations with loss of insight or delusions and they constitute a risk factor for nursing home placement (Goetz and Stebbins, 1993; Lo et al., 2009). The exact pathophysiological mechanism of VH in PD is unknown, but a combination of impaired visual processing and reduced attention has been proposed (Flowers and Robertson, 1995; Collerton et al., 2005). VH in PD have been commonly viewed as an adverse effect of dopaminergic treatment for PD, causing a relative overstimulation of the limbocortical dopaminergic receptors (Bosboom et al., 2004). All types of dopaminergic drugs are associated with the induction or exacerbation of VH, although the evidence is stronger for dopaminergic agonists than for levodopa (Baker et al., 2009; Diederich et al., 2009). The hypothesis that VH in PD are simply caused by dopaminergic overstimulation has been challenged by several observations. First, a majority of PD patients on dopaminergic treatment do not report VH, while several studies report that the mean levodopa-equivalent dose is equal in PD patients with and without VH (Fenelon et al., 2000; Merims et al., 2004). Second, high-dose challenge with levodopa in non-demented PD patients with daily VH does not precipitate hallucinations (Goetz et al., 1998). Moreover, VH have already been reported in the pre-levodopa era (Fenelon et al., 2006) and recent studies have reported a prevalence of percent of VH in drug naïve PD patients (Biousse et al., 2004; Dotchin et al., 2009). This suggests that disease-related factors play a role in the genesis of VH in PD. Clinical studies have shown that cognitive impairment in PD patients is associated with the occurrence of VH (Fenelon et al., 2000; Williams and Lees, 2005). Several cognitive domains (executive functioning, visuospatial abilities, attention) have been shown to be impaired in non-demented PD patients with VH, when compared to PD patients without VH (Imamura et al., 2007; Ramirez- Ruiz et al., 2007a). Other clinical factors that might be associated with VH in PD are disease duration (Fenelon et al., 2000), REM sleep behavioural dis-

15 6 CHAPTER 1. GENERAL INTRODUCTION order with vivid dreaming (Goetz et al., 2005) and visual disorders (Matsui et al., 2006b). Neurochemically, several neurotransmitter systems have been postulated to be involved in the genesis of VH in PD. Apart from the before mentioned dopaminergic hypothesis, other mono-amines, like serotonin and adrenalin, and acetylcholine could play a role. According to the monoaminergic-cholinergic imbalance hypothesis, VH in PD are associated with a cholinergic deficit in combination with a relatively preserved (or overstimulated) dopamine system (Perry and Perry, 1995; Francis and Perry, 2007). This thesis main topic of investigation is the association of VH in PD with impaired visual processing and decreased attention. 1.5 Visual processing and attention Visual processing In normal visual processing light is absorbed in the retina and transduced into electrical signals by the photoreceptors (rods and cones). Signals are transferred from the rods to the magnocellular ganglion cells and from the cones to the parvocellular and koniocellular ganglion cells. The axons of the ganglionic cells form the optic nerve and fibres of each eye partly cross at the chiasm to form the optic tract, carrying the representation of the contralateral visual field. The parvo- and koniocellular pathway are mainly involved in colour discrimination, while the magnocellular pathway is involved in achromatic contrast discrimination. The lateral geniculate nucleus (LGN) of the thalamus is the main relay between the retina and the visual cortex, but about 10 percent of retinal axons project to the superior colliculus, involved in head movements and saccadic eye movements. In addition, some axons of the retinal ganglionic cells project to the suprachiasmatic nucleus of the hypothalamus, involved in synchronizing biological rhythms, and the pretectum, controlling pupillary reflexes. The LGN consists of 6 different layers, with magnocellular projections to layers 1 and 2 and parvocellular projections to layers 3-6. Koniocellular cells project to the interlaminar zones between the 6 layers (Hendry and Yoshioka, 1994). From the LGN, optic radiations project to the primary visual cortex (V1) in the occipital lobe. V1 is also called the striate cortex, because of the presence of a stripe of white matter, the stria of Gennari, in layer 4. Magno-, parvo- and koniocellular axons project to different sublayers of V1, thus maintaining the segregation of these cellular pathways at this level of processing. Each half of the visual field is represented upside-down by the

16 1.5. VISUAL PROCESSING AND ATTENTION 7 contralateral striate cortex around the calcarine sulcus, the fovea being represented in the most posterior half of V1. After processing of visual stimuli in V1, information is conveyed through extrastriate areas to the occipito-parietal and occipito-temporal stream, also called the dorsal and ventral visual stream, respectively. Input to the occipito-parietal pathway (containing the motion sensitive area V5) derives mainly from the magnocellular cells, while input to the occipito-temporal pathway (with colour- and form-sensitive V4) derives from cells in both the magnocellular and parvocellular layers of the LGN (Ungerleider and Haxby, 1994). Retino-geniculate signals also project directly to V5, mostly via koniocellular neurons (Sincich et al., 2004). The ventral and lateral occipito-temporal areas are important in perceiving and recognizing visual objects (Grill-Spector, 2003; Downing et al., 2006). Several subregions in the occipito-temporal cortex exist that respond more strongly to specific object categories, such as the fusiform face area for faces and the parahippocampal place area for scenes (Kanwisher et al., 1997; Epstein et al., 1999b). Other regions that are important in visual object recognition are the fusiform gyrus (including the fusiform face area), the lingual gyrus, the lateral occipital complex and the middle temporal gyrus (Malach et al., 1995; Downing et al., 2006) Impaired visual processing and VH Visual deprivation is a well known risk factor for the occurrence of VH. VH can occur in sighted subjects after prolonged blind-folding (Merabet et al., 2003) or in otherwise healthy subjects with visual impairment, also called the Charles Bonnet syndrome (CBS) (Merabet et al., 2003; Teunisse et al., 1996). A wide range of visual perceptual disturbances has been described in PD, attributing the underlying causes to different levels of the visual-cognitive system, from retina to frontal cortex. PD patients regularly complain of difficulty reading and blurred vision, despite normal visual acuity (Pieri et al., 2000; Biousse et al., 2004). Abnormalities in early visual processing have been shown to occur in PD patients studying Visual Evoked Potentials and contrast sensitivity, reflecting dopaminergic malfunction in the amacrine cells of the retina (Bodis-Wollner, 1990). Also, autopsy studies have shown reduced retinal dopamine (DA) levels of unmedicated PD patients, but normal DA levels in PD patients that received levodopa until death (Harnois and Di Paolo, 1990). Colour discrimination can be reduced in PD, but probably only a subset of PD patients may show chromatic deficits (Buttner et al., 1995; Pieri et al., 2000; Silva et al., 2005). Several studies have shown that visual perceptual impairments in PD are as-

17 8 CHAPTER 1. GENERAL INTRODUCTION sociated with the occurrence of VH. Matsui and colleagues showed that VH in PD were closely related to impaired visual acuity (Matsui et al., 2006b). Furthermore, it was shown that contrast sensitivity and colour discrimination were significantly more impaired in PD patients with VH, compared to PD patients without VH, regardless of visual acuity (Diederich et al., 1998; Davidsdottir et al., 2005). Other visual disturbances associated with VH in PD include visual space perception (Ramirez-Ruiz et al., 2007a) and visual object perception (Barnes et al., 2003; Ramirez-Ruiz et al., 2006) Attention Attention is a heterogeneous process and one can attend to a stimulus (object, location or moment) in different ways, leading to several sub-processes of attention. In selective (or focussed) attention priority is given to one stimulus in favour of another, while sustained attention and divided attention encompass attending to one stimulus or dividing attention between two or more different stimuli, respectively, over an increasing period of time. Attending to specific features of a visual image elicits activations in brain regions that are involved in the processing of these features, for example activation of the fusiform gyrus when attending to changes in shape (Corbetta et al., 1991). Attentive, i.e. top-down, processes are considered to play an important role in the identification of objects in suboptimal visual circumstances (Bar et al., 2006). Interestingly, VH in PD tend to occur in these suboptimal visual circumstances, mostly during the evening (Fenelon et al., 2000) Impaired attention and VH Cognitive dysfunction is common in PD patients and consists mainly of impairment of executive function and attention (Muslimovic et al., 2005; Verleden et al., 2007). In a recent study, it was shown that about a quarter of de novo diagnosed PD patients without dementia were cognitively impaired, defined as impairment on at least three neuropsychological tests. All cognitively impaired PD patients had attentional and executive deficits (Muslimovic et al., 2005). Although only few studies on attention in PD exist, it seems that executive and attentional functions are more impaired in non-demented PD patients with VH than in non-demented PD patients without VH (Grossi et al., 2005; Imamura et al., 2007; Barnes and Boubert, 2008). The influence of executive dysfunction is widespread and might therefore ex-

18 1.6. THERAPEUTIC OPTIONS FOR TREATMENT OF VH 9 plain a considerable part of previous behavioral and imaging results regarding visual perception and attention. The Mini Mental State Examination (MMSE) was developed as a screening test for Alzheimer s type dementia (Folstein et al., 1975), but seems less sensitive to cognitive impairment in PD. The Frontal Assessment Battery (FAB) was developed to assess frontal lobe function to identify a dysexecutive syndrome (Dubois et al., 2000). Even in PD patients with dementia, MMSE scores can be relatively spared, while scores on the FAB are decreased and thus seem to better reflect cognitive impairments in PD patients. For this reason, an effort was made to match participating patients of the studies described in this thesis on both MMSE and FAB. 1.6 Therapeutic options for treatment of VH Non-pharmacological interventions Effective non-pharmacological interventions to improve VH in PD include correction of reduced visual acuity (glasses) or, when indicated, removing cataract (Matsui et al., 2006b). Coping strategies for PD patients with VH include looking in another direction or at another object, turning on the light during the night or speaking to the spouse or caregiver in order to check the non-reality of the phenomenon (Diederich et al., 2003, 2009) Dopaminergic modulation Typical neuroleptics have been used in the past at low doses to reduce hallucinations and other psychotic symptoms in PD, but have the disadvantage of worsening motor function. Newer, so called atypical, neuroleptics have the potential to treat hallucinations in PD with less negative effects on motor function. The only atypical neuroleptic with confirmed benefit without worsening motor function is clozapine, which exhibits weak antagonism for the D2 dopamine receptor subtype. In addition, it blocks other neurotransmitter receptors, including the serotonergic 2A receptor (Schotte et al., 1993) Cholinergic modulation Other studies have shown that, apart from the dopaminergic neurotransmitter system, the cholinergic system is likely to be involved in the pathogenesis of

19 10 CHAPTER 1. GENERAL INTRODUCTION VH in PD (Perry and Perry, 1995). The cholinergic system plays an important role in awareness. A decrease in the cortical acetylcholine (Ach) levels impairs the selection of subcortical information streams, causing unselected and chaotic cortical activation, which may predispose to hallucinations (Perry and Perry, 1995). Clinical evidence shows that visual hallucinations can be induced by anti-cholinergics, while cholinesterase inhibitors (ChE-I) ameliorate cognitive dysfunction and VH in PD (Burn et al., 2006; Wesnes et al., 2005). The cholinergic innervation of the human cortex is derived from cholinergic cell groups in the basal forebrain. The nucleus basalis of Meynert (nbm) contains cholinergic neurons that project to the entire cerebral cortex, including primary and associative visual cortical regions (Mesulam and Geula, 1994). Hilker and colleagues have shown that non-demented PD patients have reduced acetylcholinestesterase binding, reflecting reduced cholinergic activity, in occipital and temporal cortical regions, while PD patients with dementia (PDD) have a more extensive cholinergic deficit, especially in the middle temporal gyrus, compared to healthy controls (Hilker et al., 2005). 1.7 Brain imaging methods Magnetic Resonance Imaging (MRI) Magnetic Resonance Imaging (MRI) is a tool to acquire detailed images of the human brain in vivo, making use of the magnetic properties of different tissues. Hydrogen atoms in the brain tissue align with the static magnetic field inside the scanner, called longitudinal magnetization. A dynamic magnetic field is subsequently applied in which hydrogen atoms are excited by radio frequency pulses in a direction perpendicular to the static magnetic field, called transverse magnetization. When the radio frequency is turned off, hydrogen atoms return to equilibrium, emitting energy that gives rise to the MRI signal. Recovery of the longitudinal signal (T1) or decay of the transverse signal (T2) can be measured, signal strength depending on tissue characteristics in which the atoms reside. Signal strength is also affected by changes in homogeneity of the local magnetic field due to changes in blood oxygenation (T2*), used in functional MRI (Huettel et al., 2004).

20 1.8. OUTLINE OF THE THESIS Basic principles of functional MRI Functional MRI (fmri) is a technique that uses magnetic properties of blood to determine indirectly which brain regions are active. When a certain brain region becomes activated, for example during a specific task, blood flow increases and oxygen rich blood (containing oxyhemoglobin) is supplied to this region, replacing the deoxygenated blood (containing deoxyhemoglobin). This response is called the hemodynamic response and can be measured over time: the Hemodynamic Response Function (HRF). For reasons unknown, more oxygenated blood is supplied than needed (Fox and Raichle, 1986), the so-called overshoot. So even when activated tissues use the supplied oxygen, the overshoot causes a local relative increase in oxygenated blood. Deoxyhemoglobin has paramagnetic properties, which means that it introduces inhomogeneity into the nearby magnetic field. Oxyhemoglobin, on the other hand, is weakly diamagnetic and has little effect on the local magnetic field, leading to an increased signal (Pauling and Coryell, 1936). The fmri signal that arises is an indirect measure of neuronal activity and is also called the Blood Oxygenation Level Dependent (BOLD) signal. Cerebral activation during a specific task is compared to another task or with baseline. Every 2-3 seconds whole brain volume images can be acquired, consisting of so-called voxels (volume pixels) of approximately 3 mm 3. Voxels that show differences during a task condition can be displayed as a map (image) of t- values, using Statistical Parametric Mapping (SPM) (Friston et al., 1995). 1.8 Outline of the thesis This thesis topic is visual hallucinations (VH) in Parkinson s disease (PD). The main objectives of this thesis were to investigate: 1) the association between VH in PD and impairments of visual processing and attention and 2) cerebral functional organization underlying visual processing in PD patients with VH. In addition, the influence of pharmacological and non-pharmacological interventions on visual processing in PD and on VH in CBS, respectively, is explored. This thesis is divided in two parts. In the first part (chapters 2, 3, 4 and 5), underlying mechanisms of VH in PD are investigated. In chapter 2 we investigated visual perception of gradually revealed images and sustained attention in PD patients with VH, compared to PD patients without VH and to healthy controls. In chapter 3 visual object and space perception was investigated in the same subjects. To gain further insight in underlying mechanisms of VH in

21 12 CHAPTER 1. GENERAL INTRODUCTION PD, we investigated cerebral activation patterns with fmri before and during recognition of gradually revealed images in these patients, compared to PD patients without VH and to controls (chapter 4). In chapter 5 we used Voxel Based Morphometry (VBM) to investigate whether the functional differences in PD patients with VH (from chapter 4) were associated with structural, i.e. grey matter volume, changes. The second part (chapters 6, 7 and 8) focuses on therapeutic interventions in patients with VH. Chapter 6 describes preliminary data on the effect of the cholinesterase-inhibitor rivastigmine on visual object processing in healthy controls. Chapter 7 is a pilot study in which the effect of apomorphine on visual perception and attention in PD patients with VH is described. Chapter 8 describes fmri correlates of VH in one patient with CBS and the influence of repetitive transcranial magnetic stimulation (rtms). In chapter 9 the results of our clinical and imaging studies are discussed in a broader perspective, focusing on possible mechanisms of impaired visual perception and attention leading to VH in PD, using a functional network approach.

22 Chapter 2 Visual object recognition and attention in Parkinson s disease patients with visual hallucinations A. M. Meppelink 1,2, J. Koerts 1,2, M. A. Borg 1, K. L. Leenders 1,2, T. van Laar 1,2 (1) Department of Neurology, University Medical Center Groningen, The Netherlands (2) School of Behavioral and Cognitive Neurosciences, University of Groningen, the Netherlands Accepted for publication in Movement Disorders 13

23 14 CHAPTER 2. VISUAL RECOGNITION IN PD WITH VH 2.1 Abstract Visual hallucinations (VH) are common in Parkinson s disease (PD) and are hypothesized to be due to impaired visual perception and attention deficits. We investigated whether PD patients with VH showed attention deficits, a more specific impairment of higher order visual perception, or both. Fortytwo volunteers participated in this study, including 14 PD patients with VH, 14 PD patients without VH and 14 healthy controls (HC), matched for age, gender, education level and for level of executive function. We created movies with images of animals, people and objects dynamically appearing out of random noise. Time until recognition of the image was recorded. Sustained attention was tested using the Test of Attentional Performance. PD patients with VH recognized all images but were significantly slower in image recognition than both PD patients without VH and HC. PD patients with VH showed decreased sustained attention compared to PD patients without VH who again performed worse than HC. In conclusion the recognition of objects is intact in PD patients with VH, however, these patients where significantly slower in image recognition than patients without VH and HC, which was not explained by executive dysfunction. Both image recognition speed and sustained attention decline in PD, in a more progressive way if VH start to occur.

24 2.2. INTRODUCTION Introduction Visual hallucinations (VH) are common in Parkinson s disease (PD) with a prevalence of approximately 30 percent (Barnes and David, 2001). VH are defined as involuntary visual perceptions in the waking state without external visual stimulation (Collerton et al., 2005). Minor forms of hallucinations consist of sensation of presence, a sideways passage or illusions, in which an external stimulus is perceived but misinterpreted (Fenelon et al., 2000, 2008; Barnes and David, 2001). VH can be simple, characterized by the absence of form, or complex, with a clearly defined specific form. All above mentioned categories of hallucinations can occur in PD and often overlap, making it likely that they are related. VH in PD typically comprise complex visual, commonly moving images lasting for seconds to minutes. Animals, people and objects define the categories of images that are often perceived by hallucinating PD patients (Fenelon et al., 2000; Mosimann et al., 2006). The exact etiology of VH in PD is unknown, however a combination of impaired visual processing and attention may be involved (Diederich et al., 2005). While reduced visual acuity is a risk factor for the occurrence of VH (Matsui et al., 2006b), it was shown that PD patients with VH have additional impairments on facial recognition and object perception compared to PD patients without VH (Ramirez-Ruiz et al., 2006; Barnes et al., 2003). Other studies have shown attentional deficits in non-demented PD patients experiencing VH (Barnes and Boubert, 2008). In a recently described model on VH, it was hypothesized that a combination of impaired visual processing and attention is required for VH to occur (Collerton et al., 2005). VH in PD tend to occur in dim, suboptimal visual circumstances (Sanchez- Ramos et al., 1996). Several techniques have been used to mimic suboptimal visual situations, such as backward masking, in which briefly presented images are immediately followed by a masking stimulus, and gradual revelation of objects using panels and noise (Bar et al., 2001; James et al., 2000; Grill-Spector et al., 2000). By creating movies in which an image slowly and dynamically appears out of random noise, the speed and content of conscious perception of images can be assessed (James et al., 2000; Reinders et al., 2006; Kleinschmidt et al., 2002). In this study movies were created of pictures of animals, objects and people gradually appearing out of noise. The aim was to investigate visual perception of gradually revealed images in PD patients with VH compared to PD patients without VH and healthy controls. Additionally, it was investigated whether PD patients with VH showed

25 16 CHAPTER 2. VISUAL RECOGNITION IN PD WITH VH Figure 2.1: Images revealed out of noise. The images Queen (A), Dog (B), Telephone (C), and Rabbit (D) that were dynamically revealed out of noise. decreased sustained attention and whether this was associated with the visual perception of gradually revealed objects. 2.3 Patients and Methods Subjects Forty-two volunteers participated in this study, including 14 PD patients who experienced VH at least weekly during the last month, 14 PD patients without VH and 14 healthy controls. PD was diagnosed according to the criteria of the UK Parkinson s Disease Society Brain Bank. These three groups were matched for age (ANOVA: F=0.91 p=0.41), gender (Chi-Square test p=0.73) and level of education (Kruskal-Wallis Test p=0.86). The latter was rated with a Dutch education scale ranging from 1 (elementary school not finished) to 7 (university degree). Both PD groups were also matched for their level of executive functioning (i.e. score on Frontal Assessment Battery (FAB; t-test: t=-0.44 p=0.67). All PD patients were on during the assessment. The levodopaequivalent daily dose (LEDD) was calculated for all patients, according to the formula: LEDD= levodopa dose (mg) + ( 0.3 *levodopa dose if using entacapone with each dose) + (slow release levodopa *0.7) + (bromocriptine *10) + (ropinirole *20) + (pergolide *100) + (pramipexole *100) + (apomorphine *10). In the PD with VH group three patients used stable medication against their VH (one patient used clozapine, one patient used reminyl and one used clozapine and reminyl), while none of the subjects used anticholinergics. Demographic and illness characteristics are described in table 2.1. Exclusion criteria were dementia (Mini Mental State Examination (MMSE) score < 24), neurological disorders other than PD, psychiatric disorders, visual acuity less than 50 percent (Snellen chart) and visual field defects.

26 2.3. PATIENTS AND METHODS 17 PD + VH PD - VH HC M(SD) range M(SD) range M(SD) range Age (years) 69.0 (5.0) (6.8) (5.9) Education 4.4 (1.7) (1.9) (1.7) 2-7 Disease duration 10.7 (4.9) (5.7) 1-23 LEDD 944 (509) (446) SPESS 9.4 (4.3) (5.2) 1-18 MMSE 26.2 (1.3) (1.6) (1.5) FAB 14.3 (1.7) (2.6) (1.0) Males: n (%) 9 (64 %) 10 (71 %) 8 (57 %) Females: n (%) 5 (36 %) 4 (29 %) 6 (43 %) Table 2.1: Demographic and illness characteristics of PD patients with visual hallucinations (PD + VH; n=14), PD patients without VH (PD - VH; n=14) and healthy controls (HC; n=14) This study was approved by the Medical Ethical Committee of the University Medical Center Groningen. All participants signed an informed consent prior to study inclusion Stimulus Material The motor severity of PD patients was rated with the Shortened Parkinson Evaluation Scale (SPES) of the SCales for Outcomes in Parkinson s disease (SCOPA) (Marinus et al., 2004). The severity of VH of PD patients was assessed with part B Hallucinations of the Neuropsychiatric Inventory and a questionnaire based on the characteristics of visual hallucinations in PD patients as described by Barnes and David (Barnes and David, 2001). All groups were presented with the following tests: Image recognition movies Four newly developed movies, in which images gradually and dynamically appeared out of random noise, were shown on a computer screen. The movies contained pictures of animals (dog, rabbit), a well-known person (Dutch queen) and an object (telephone, see figure 2.1). Movies were presented in the same order (Queen, Dog, Rabbit, Telephone) to all participants. Subjects had to verbally name the image when recognized, while time until recognition of the image was recorded by stopwatch. The movie was stopped when the image was recognized. The session duration varied between 91 and 248 seconds. Test battery for Attention Performances (TAP) The subtest Optical vigilance of this battery was used to assess the ability to focus attention for 10 minutes. Participants had to push a button when recognizing irregularities in a normally regular movement pattern of an object

27 18 CHAPTER 2. VISUAL RECOGNITION IN PD WITH VH on a computer screen. The test contained twelve irregularities. The number of times the participant did not recognize an irregularity was rated. Mean reaction times were also measured, by calculating the time between the irregularity presentation and the button press Statistical Analysis Not all variables were normally distributed in all three groups. Also, concerning the variable number omitted of the optical vigilance test of the TAP the assumption of equality of variance was violated. Therefore, the non-parametric Welch test, which is very robust when sample sizes are equal and the normality and equality of variance assumptions are violated (Buning, 1997) was used to verify the results of the parametric tests. The results of the non-parametric test supported our parametric findings, therefore only the results of the parametric tests are described. The three groups were compared concerning their time until correct image recognition using MANOVA. MANOVA was also used to compare the scores of all groups on the optical vigilance test of the TAP. Subsequently, a Helmert contrast, only applied to variables on which groups significantly differed according to the MANOVA, was used to determine firstly, if the group of PD patients with VH differed significantly from both PD patients without VH and healthy controls and secondly, if PD patients without VH differed from healthy controls. To investigate if the 3 PD patients using medication for their VH biased the PD patients with VH group results, we performed a second MANOVA with Helmert contrast, now excluding these subjects. Because much of our data was not normally distributed we calculated within subject variability and made scatter plots to investigate if our results were biased by outliers. Image recognition times on the 4 movies were converted to z-scores and the difference between the highest and lowest z-score was used to calculate within subject variability. This variability was compared between groups using ANOVA. Mean reaction times (see TAP) of the three groups were compared using ANOVA. A Difference contrast was applied to further differentiate differences between groups. Spearman correlations within groups were calculated to investigate associations between visual acuity, mean reaction times and disease severity on the one side and image recognition time and number omitted on the TAP on the other side. Additionally, Spearman correlations were calculated between sustained attention and the time until image recognition within each group.

28 2.4. RESULTS 19 PD + VH PD - VH HC F p M (SD) M (SD) M (SD) Queen 42.6 (16.3) 35.7 (7.9) 34.1 (6.6) Dog* 23.2 (5.3) 20.2 (3.3) 18.9 (3.9) Rabbit** 50.2 (19.4) 41.6 (6.6) 37.6 (5.6) Telephone 23.9 (5.2) 21.7 (6.1) 19.3 (5.3) Table 2.2: Comparison of time until recognition of movies in PD patients with visual hallucinations (PD+VH; n=14), PD patients without VH (PD-VH; n=14) and healthy controls (HC; n=14) using MANOVA (two-tailed). * Helmert contrast: PD+VH are slower in recognition than PD-VH and HC (p=0.01); no differences between PD-VH and HC (p=0.42). **Helmert contrast: PD+VH are slower in recognition than PD-VH and HC (p=0.01); no differences between PD-VH en HC (p=0.39) 2.4 Results In the group of PD patients with VH, 7 percent (n=1) reported having VH about once a week, 50 percent had VH several times per week and 43 percent reported having VH several times a day. Twenty-nine percent of the hallucinating PD patients reported that they became upset during their VH and 21 percent considered their VH as a moderate to severe emotional burden. None of the PD patients with VH experienced VH during the testing. Twentyone percent of hallucinating PD patients associated their VH with the use of dopaminergic medication. LEDD scores were not significantly different in PD patients with VH, compared to PD patients without VH (t=1.38, p=0.18). The motor severity of all PD patients was assessed using the SPES-SCOPA (table 2.1). All subjects were able to correctly name all images that were gradually revealed out of noise and considered the test as non-fatiguing. The MANOVA showed significant differences between groups for the time until recognition of the images Dog and Rabbit. Moreover, the Helmert contrast showed that PD patients with VH recognized the images significantly slower than both PD patients without VH and healthy controls, while no significant differences between PD patients without VH and healthy controls were found. For the images Queen and Telephone, a trend to similar results was seen, but these differences were not significant (see table 2.2). Excluding the PD patients that used medication for their VH did not alter the results (Helmert contrast com-

29 20 CHAPTER 2. VISUAL RECOGNITION IN PD WITH VH Figure 2.2: Relative increase of recognition time in Parkinson s disease patients. Bar diagram showing the mean percentage of extra time until recognition for PD patients with and without VH (PD+VH, respectively PD-VH), relative to healthy controls. group differs significantly from both other groups, shown by MANOVA and Helmert contrast (see table 2.2). paring PD patients with VH with both PD patients without VH and healthy controls: Queen: p=0.03, Dog: p=0.04, Rabbit: p=0.02, Telephone: p=0.07). Within subject variability did not differ between groups (p=0.59). A MANOVA showed that both PD patients with and without VH had longer reaction times than healthy controls (p=0.01). Furthermore, there was no difference in reaction times between PD patients with VH and PD patients without VH (p=0.64). There was no correlation between the recognition performance and mean reaction times within groups (data not shown). Image recognition times were not associated with visual acuity or disease severity (data not shown). The relative increase of image recognition time in PD patients with and without VH, both compared to healthy controls, is depicted in figure 2.2. Sustained attention was measured by rating the number of non-recognized ir-

30 2.5. DISCUSSION 21 PD + VH PD - VH HC r(p) r(p) r(p) Queen (0.66) 0.19 (0.50) 0.04 (0.90) Dog 0.22 (0.48) 0.56 (0.04) 0.47 (0.08) Rabbit 0.39 (0.18) 0.33 (0.26) 0.20 (0.50) Telephone 0.09 (0.78) 0.00 (1.00) 0.30 (0.30) Table 2.3: Spearman correlations(two-tailed) between sustained attention and time until recognition of movies in PD patients with visual hallucinations (PD+VH; n=13), PD patients without VH (PD-VH; n=14) and healthy controls (HC; n=14). regularities on the optical vigilance test of the TAP. On average PD patients with VH did not recognize 4.3 irregularities (SD=3.4), PD without VH 2.5 (SD=2.1) and healthy controls 0.3 irregularities (SD=0.5). The MANOVA showed that the groups differed significantly from each other on sustained attention (F=10.7, p=0.00). Additionally, the Helmert contrast showed that PD patients with VH showed a significantly decreased sustained attention as compared to both PD patients without VH and healthy controls (p=0.00). PD patients without VH showed a significantly decreased sustained attention as compared to healthy controls (p=0.00). The Spearman correlation did not show an association between sustained attention and visual acuity or disease severity (data not shown). Spearman correlations were also calculated between sustained attention and image recognition times within each group. PD patients without VH showed an association between sustained attention and the movie Dog, while no other associations were found between the time until recognition of images gradually revealed out of noise and sustained attention (table 2.3). Scatter plots of these associations are shown in figure Discussion The aim of this study was to investigate visual perception of gradually revealed images, mimicking suboptimal situations in which VH often occur, in PD patients with VH compared to PD patients without VH and healthy controls. Additionally, it was investigated whether PD patients with VH showed decreased sustained attention and whether this was associated with the time until correct recognition of gradually revealed objects. We suggest that our movies with pictures of animals, people and objects gradually appearing out of random noise mimic suboptimal visual perceptual circumstances. In contrast to static images, which are generally characterized by a large degree of

31 22 CHAPTER 2. VISUAL RECOGNITION IN PD WITH VH Figure 2.3: Scatter plots showing data of PD patients with VH(), PD patients without VH() and healthy controls( ). perceptual clarity and constancy, a dynamic presentation of stimuli like in our movies probably provides a more natural visual perceptual situation. Regarding the gradual appearance of the images, conscious perception is postponed, creating an additional temporal dimension in the assessment of visual perception. These movies are applied here to measure differences in visual perception associated with the occurrence of VH in PD. Cognitive impairment is often found in PD patients, if compared to healthy controls. A dysexecutive syndrome is one of the core features of cognitive dysfunction in PD and is usually one of the earliest cognitive symptoms in PD (Dubois and Pillon, 1997; Bosboom et al., 2004). The unique feature of this study is the cognitively matched set-up of the PD patient groups with special focus on executive functioning, which makes it unlikely to explain our findings by a difference in executive functioning. While others used the MMSE to match groups of PD patients with and without VH, we used the FAB, which is a useful screening test to assess executive functioning in PD (Dubois et al., 2000). PD patients with VH were slower in recognizing images gradually revealed out of random noise than both PD patients without VH and healthy controls. These results were independent of visual acuity, disease severity or reaction

32 2.5. DISCUSSION 23 times. Although recognition of these images in PD patients with VH was slower, our data show that the final recognition and naming of the image was unimpaired. It is unclear why only Dog and Rabbit were recognized significantly slower. It is likely that Queen and Telephone become significant if group sizes are increased (also see figure 2.2). Other factors might have influenced the results. Firstly the fixed order of presentation of the movies. However, the Telephone movie, which was the final movie in all sessions, was not recognized significantly slower, which makes this influence very unlikely. Secondly, as PD patients are generally more fatigued later during the day, when VH also tend to be worse, it is important to consider the time of day at which the assessment was done, because this could have influenced the performance. We tested our subjects arbitrarily over the day, making it difficult to estimate the possible effect of this factor on our results. The current study also showed that PD patients with VH had decreased sustained attention, compared to PD patients without VH, who in turn performed worse than healthy controls. Sustained attention refers to the ability to focus on an enduring, monotone task for a longer period of time. Attentional deficits may fluctuate as is seen frequently in PD Dementia and Dementia with Lewy bodies, in which VH are common (Walker et al., 2000). Moreover, reduced sustained attention in PD patients with VH, as compared to PD patients without VH and healthy controls, is consistent with recent data, showing a decreased focused attention in PD patients and an even worse attention in PD patients with VH (Barnes and Boubert, 2008). Regarding this study and the previous studies on visual processing in PD patients with VH, it is difficult to compare the behavioral data, because all studies used different techniques to investigate visual perception. We used a different time frame as compared to others, taking into account visual object perception as well as time until recognition of the object. The slower image recognition in PD patients with VH might be caused by a reduced bottom-up visual processing. Hypothetically, this may lead to a higher demand on the top-down system, resulting in activation of visual images by a kind of overcompensation, causing VH in PD. Imaging studies have shown that the primary visual cortex (V1) is equally affected in PD patients with VH as compared to PD patients without VH (Boecker et al., 2007; Oishi et al., 2005), while secondary visual cortices, i.e. occipital-temporal and occipitalparietal regions, show hypoperfusion and reduced glucose metabolism in PD patients with VH, as compared to PD patients without VH (Okada et al., 1999; Oishi et al., 2005; Matsui et al., 2006a; Boecker et al., 2007). This may reflect reduced bottom-up visual processing, in which the visual image is processed

33 24 CHAPTER 2. VISUAL RECOGNITION IN PD WITH VH ventrally from V1 to the occipital-temporal region, which is involved in object perception, and dorsally to the occipital-parietal region, which is involved in space and movement perception. In higher visual processing, attentive, i.e. top-down, processes are considered to play an important role in the identification of objects in suboptimal visual circumstances (Hopfinger and West, 2006). It has been hypothesized that VH and attentional deficits are both due to a degeneration of the cholinergic system and that attentional deficits may play a causative role in the generation of VH in PD (Francis and Perry, 2007). This study has shown that both image recognition and sustained attention are impaired in PD patients with VH, as compared to PD patients without VH, in turn being more impaired than healthy controls. However, the time until recognition did not show significant differences between PD patients without VH and healthy controls. This suggests that in PD sustained attention as well as image recognition both deteriorate, probably in consequence of the same underlying disease process, but with different time intervals. Irrespective of the overall level of executive functioning, attention seems to decline linearly from the beginning in PD patients, while visual perceptual functions remain relatively intact in the course of PD, starting to deteriorate when VH become apparent (figure 2.4). All movies, except one ( Dog ), failed to show positive correlations between the time until recognition and sustained attention, in all three groups. Only the Dog movie did show a correlation between recognition time and sustained attention in the PD patient group without VH, for reasons we do not understand. These limited correlations suggest that impaired sustained attention does not seem to cause an increased recognition-time of images revealed out of random noise. However, the somewhat ambiguous results make it difficult to draw firm conclusions. Whether sustained or other forms of attention, such as focused, selective or divided attention are involved in the generation of VH, should be determined in future research. 2.6 Conclusions In conclusion, our dynamic image recognition paradigm provides a natural means to present visual stimuli and creates an additional temporal dimension in the assessment of visual perception. The recognition of these images in PD patients with VH was intact, but significantly slower than in patients without VH and HC. Sustained attention was also shown to be impaired in PD patients with VH, compared to PD patients without VH and HC. This was not explained by a difference in executive functioning. Future research has

34 2.6. CONCLUSIONS 25 to validate the image recognition paradigm described in this article, using a greater number of movies of comparable duration, in order to average data and to investigate test-retest reliability. Currently we use similar movies with images revealed from random noise, in an fmri paradigm, to elucidate cerebral patterns involved in image recognition and possible higher order deficits in PD patients with VH during this task.

35

36 Chapter 3 Attentional and perceptual impairments in Parkinson s disease with visual hallucinations J. Koerts 1,2,3, M. A. Borg 1, A. M. Meppelink 1,3, K. L. Leenders 1,3, M. van Beilen 1,3, T. van Laar 1,3 (1) Department of Neurology, University Medical Center Groningen, The Netherlands (2) Department of Neuropsychology, Faculty of Behavioral and Social Sciences, University of Groningen, The Netherlands (3) School of Behavioral and Cognitive Neurosciences, University of Groningen, the Netherlands Accepted for publication in Parkinsonism and Related Disorders 27

37 28 CHAPTER 3. ATTENTION AND PERCEPTION IN PD WITH VH 3.1 Abstract Visual hallucinations (VH) are common in Parkinson s Disease (PD). Both deficits of perception and attention seem to play a role in the pathogenesis of VH in PD. However, the possible coexistence of impairments in attention and visual perception in PD with VH is not known. This study investigated both attention and visual perception in non-demented PD patients with VH, compared to PD patients without VH and healthy controls. Fourteen participants were included in each group. All patients were assessed with sustained visual attention and object and space perception tests. Only PD patients with VH showed impairments on object and space perception. In addition, PD patients with and without VH showed impairments on sustained visual attention, being more severely affected in PD patients with VH. Only in PD patients with VH sustained visual attention was associated with a decreased object and space perception. The results of our study thus suggest that in PD patients with VH an impairment of object and space perception, possibly in association with a decreased sustained visual attention, might play a role in the pathogenesis of VH.

38 3.2. INTRODUCTION Introduction Visual hallucinations (VH) are common in Parkinson s Disease (PD) and often occur in dim, suboptimal visual surroundings. VH in Parkinson s disease have been commonly viewed as an adverse effect of dopaminergic treatment for PD, causing a relative over- stimulation of the limbocortical dopaminergic receptors (Bosboom et al., 2004). However, VH have already been reported in the pre-levodopa era and may not be associated with the dose or duration of treatment of dopaminergic drugs (Fenelon et al., 2006; Goetz et al., 1998; Holroyd et al., 2001). In addition, cognitive and visual impairments and a dysregulation of the sleep-wake cycle are suggested to be contributing factors to the pathophysiology of VH in PD (Diederich et al., 2009). Perception normally consists of a combination of bottom-up and top-down processing. Bottom-up processing involves data-driven processing, i.e. the incoming visual stimuli in the occipital cortex and the perception of spaces and objects in the occipito-parietal and inferior temporal lobe, respectively. Top-down processing, on the other hand, involves the process that contributes to perception, but which does not originate in the external world, e.g. perceptual expectations and attentional modulation (Aleman et al., 2008). Several models have been introduced to explain (visual) hallucinations (Aleman et al., 2008; Collerton et al., 2005; Diederich et al., 2005). Although these models are different, all models suggest that hallucinations are due to impairments in both bottom-up and top-down information processing. In patients with PD and VH several studies have shown object and space perception impairments, as compared to PD patients without VH (Barnes et al., 2003; Ramirez-Ruiz et al., 2007a). Secondly, several studies reported PD patients with VH showing a decreased ability to focus attention and fluctuating levels of attention, relative to PD patients without VH (Ballard et al., 2002; Barnes and Boubert, 2008). Studies on PD patients with VH so far were solely focused on visual perception or on attention. However, both perception and attention play a role in perceiving the external world and the possible coexistence of impairments in visual perception and attention in PD with VH should be investigated. We previously investigated object perception in PD patients with VH, mimicking suboptimal visual situations by creating movies with objects dynamically appearing out of random noise, with no direct manipulations on the objects themselves. In addition, the associations between visual sustained attention and object perception were determined (Meppelink et al., 2008). It was concluded that the

39 30 CHAPTER 3. ATTENTION AND PERCEPTION IN PD WITH VH recognition of objects in that way is not impaired in PD patients with VH, however these patients were significantly slower in image recognition than PD patients without VH and healthy controls, which was associated with a decreased sustained visual attention. When perceiving objects it is assumed that we make use of all objects properties for optimal recognition (Newell et al., 2004). The present study focuses on the recognition and identification of differentially manipulated objects, not taking into account the background environment. Also, the associations between visual sustained attention and object perception will be determined. In addition, space perception will be investigated in the present study. Previous studies on visual perception in PD patients with and without VH, had matched groups on MMSE scores and level of intelligence (Barnes et al., 2003). However, PD patients with VH show decreased levels of executive functioning, as compared to PD patients without VH (Barnes and Boubert, 2008). The influence of executive dysfunctions is widespread and may have altered previously described results of PD with VH. The present study therefore investigated matched patient groups for their level of executive functioning (assessed with the Frontal Assessment Battery (FAB)) (Dubois et al., 2000), in addition to age, sex and level of education. 3.3 Methods Subjects Forty-two volunteers participated in this study, including 14 PD patients with VH, 14 PD patients without VH and 14 healthy controls. These subjects were included in a larger study focused on perception and attention in PD patients with and without VH and healthy controls, of which part of our results are already described (Meppelink et al., 2008). The groups of PD patients without VH and healthy controls were matched to the group of PD patients with VH in such a way that no group differences existed for age (ANOVA: F= 0.91; p= 0.41), gender (Chi-Square test p= 0.73) and level of education (Kruskal-Wallis Test p= 0.86; rated with a Dutch education scale ranging from 1 (elementary school not finished) to 7 (university degree)). Both PD groups were also matched for their level of executive functioning (i.e. score on FAB (Dubois et al., 2000), t-test: t= -0.44; p= 0.67). PD was diagnosed according to the criteria of the UK Parkinson s Disease Society Brain Bank by experienced neurologists (K.L. Leenders and T. Van

40 3.3. METHODS 31 PD + VH PD - VH HC M(SD) range M(SD) range M(SD) range Age (years) 69.0 (5.0) (6.8) (5.9) Education 4.4 (1.7) (1.9) (1.7) 2-7 Disease duration 10.7 (4.9) (5.7) 1-23 LEDD 944 (509) (446) SPESS 9.4 (4.3) (5.2) 1-18 MMSE 26.2 (1.3) (1.6) (1.5) FAB 14.3 (1.7) (2.6) (1.0) Visual acuity 0.6 (0,2) (0.2) (0.2) Males: n (%) 9 (64 %) 10 (71 %) 8 (57 %) Females: n (%) 5 (36 %) 4 (29 %) 6 (43 %) Table 3.1: Demographic and illness characteristics of PD patients with visual hallucinations (PD + VH; n=14), PD patients without VH (PD - VH; n=14) and healthy controls (HC; n=14) Laar). Exclusion criteria were patients with moderate to severe cognitive impairments (Mini Mental State Examination score< 24), neurologic disorders other than PD, psychiatric disorders, visual acuity less than 50 percent (Snellen chart) and visual field defects (Donders technique). All PD patients were on during the assessment. A levodopa-equivalent daily dose score (LEDD) was calculated for all patients (Esselink et al., 2004). In the PD with VH group three patients used stable medication for their VH (one used clozapine, one used galantamine and one used clozapine and galantamine), while none of the subjects used anticholinergics. The motor severity of PD patients was rated with the Shortened Parkinson Evaluation Scale (SPES) of the SCales for Outcomes in Parkinson s disease (SCOPA) (Marinus et al., 2004). The severity of VH was assessed with part B Hallucinations of the Neuropsychiatric Inventory and a questionnaire based on the characteristics of VH in PD patients (see Barnes and David (2001)). All demographic and clinical characteristics are described in Table 3.1. PD patients with VH showed higher LEDD scores than PD patients without VH, however this difference was not significant (t= 1.38; p= 0.18). PD patients with VH did however show a significantly longer disease duration (t= 2.03; p= 0.05) and a decreased visual acuity (F= 4.3; p= 0.02) compared to PD patients without VH. Patient groups did not differ concerning the severity of their motor symptoms (t= 0.75; p= 0.46). This study was approved by the Medical Ethical Committee of the University Medical Center Groningen. All participants signed an informed consent prior to study inclusion.

41 32 CHAPTER 3. ATTENTION AND PERCEPTION IN PD WITH VH Stimulus material All three groups were presented with the following neuropsychological tests: Visual Object and Space Perception battery (VOSP, (Warrington, 1991)) This battery comprises four tests for object perception and four tests for space perception. The tests for object perception include identifying incomplete letters ( Incomplete letters ), identifying silhouettes of objects and animals ( Silhouettes ), identifying a silhouette of a real object out of four silhouettes of which three were silhouettes of non-sense objects ( Object decision ) and identifying rotated silhouettes of objects ( Progressive silhouettes ). The tests for space perception include counting dots ( Dot counting ), determining whether dots are placed in the middle of a square or not ( Position discrimination ), determining with which number of at randomly placed numbers in a square the location of a dot in an other square corresponded ( Number location ) and counting the number of bricks of a structure of bricks ( Cube analysis ). The subtests of the VOSP were presented in an at random order. Test battery for Attention Performances (TAP) PD patients with VH might show fluctuating levels of attention and might consequently obtain a relatively normal score on an attention test with a short duration (Ballard et al., 2002). Therefore a test (the subtest Optical vigilance of the TAP), during which participants had to focus their visual attention for a longer period of time, was applied in this study. During 10 min participants had to push a button if they recognized irregularities in a normally regular movement pattern of an object on a computer screen. The test contained a total of twelve irregularities. The number of omitted irregularities was rated as a measure for sustained visual attention Statistical analysis Not all variables were normally distributed within all three groups. Also, some variables violated the assumption of equality of variance. Therefore, the nonparametric Welch test, which is robust when sample sizes are equal and the normality and equality of variance assumptions are violated (Buning, 1997), was used to verify the results of the parametric tests. The results of the nonparametric test supported our parametric findings, therefore only the results of the parametric tests are described. All three groups were compared on all subtests of the VOSP and on the sus-

42 3.4. RESULTS 33 tained visual attention test, using MANOVA. Since PD patients with VH showed a higher LEDD score (even though this was not significant), had a longer disease duration and showed a lower visual acuity than PD patients without VH, these variables were entered as covariates (NB. The Welch test does not allow the inclusion of covariates. Therefore, a regression analysis was performed and residual variables, which represent the performances on the VOSP subtests and the sustained visual attention test without the influence of the above described covariates, were saved. These residual variables were entered in the Welch test). Subsequently, a Helmert contrast was used to determine first of all whether PD patients with VH differed significantly from both PD patients without VH and healthy controls and second whether PD patients without VH differed from healthy controls. The achieved effect sizes and power of all comparison between groups were calculated post hoc, using G Power (Faul et al., 2007). In addition, Spearman correlations were calculated within groups between sustained visual attention and the subtests of the VOSP on which PD patients with VH differed significantly from both other groups. 3.4 Results Of all PD patients with VH, 7 percent reported having VH about once a week, 50 percent had VH several times per week and 43 percent reported to have VH several times a day. In total 29 percent of the hallucinating PD patients became upset during their VH and 21percent of the hallucinating PD patients considered their VH as a moderately to severely emotional burden. The performance on the subtests of the VOSP of all three groups is shown in Table 3.2 and Fig The groups showed a statistical significant difference on Object decision and trends towards significant differences between groups were found for Dot counting and Number location. No significant differences were found between the 3 groups on the other VOSP subtests. The differences were further analyzed using the Helmert contrast, which showed that PD patients with VH scored significantly lower on Object decision than both PD patients without VH and healthy controls (p= 0.02). PD patients without VH and healthy controls scored similarly on Object decision (p= 0.88). Concerning Number location the same pattern was found; PD patients with VH scored significantly lower than both PD patients without VH and healthy controls (p= 0.05) and PD patients without VH and healthy controls obtained similar scores (p= 0.74). On Dot counting PD patients with and without VH obtained similar scores (p= 0.44), while both PD patient groups showed

43 34 CHAPTER 3. ATTENTION AND PERCEPTION IN PD WITH VH Figure 3.1: Mean performance of PD patients with VH (PD + VH; n = 14), PD patients without VH (PD - VH; n = 14) and healthy controls (HC; n = 14) on VOSP subtests. significantly lower scores than healthy controls (p= 0.02). The MANOVA showed that the three groups differed significantly on sustained visual attention (F= 4.1; p= 0.03; effect size= 0.74; power= 0.99). On average PD patients with VH did not recognize 4.3 (SD= 3.4) irregularities, PD without VH did not recognize 2.5 (SD= 2.1) irregularities and healthy controls did not recognize 0.3 (SD= 0.5) irregularities. The Helmert contrast showed that PD patients with VH had significantly more difficulties sustaining their visual attention than both PD patients without VH and healthy controls (p= 0.02) and PD patients without VH had significant more difficulties sustaining their visual attention than healthy controls (p= 0.05). The correlations between sustained visual attention and the subtests of the VOSP on which PD patients with VH obtained significantly lower scores than both PD patients without VH and healthy controls (Object decision and Number location) are shown in Table 3.3. Sustained visual attention was significantly correlated with Object decision and Number location within PD patients with VH (see Fig. 3.2 and Table 3.3). In both other groups no correlations were found between Ob-

44 3.5. DISCUSSION 35 Figure 3.2: Correlations between Object decision, Number location and sustained visual attention in PD patients with VH (PD + VH; n = 14). PD + VH PD - VH HC F p M(SD) M(SD) M(SD) Incomplete letters 18.1 (1.7) 18.9 (1.0) 19.0 (0.9) Silhouettes 18.6 (4.7) 20.7 (3.5) 22.1 (4.1) Object decision 15.7 (1.8) 17.8 (1.0) 18.1 (1.6) Progressive silhouettes 10.4 (2.7) 11.5 (2.5) 10.6 (3.4) Dot counting 9.9 (0.3) 9.8 (0.4) 10.0 (0.0) Position discrimination 18.4 (2.8) 19.5 (0.8) 19.6 (0.6) Number location 8.9 (1.1) 9.6 (0.5) 9.3 (0.8) Cube analysis 8.5 (2.2) 9.2 (1.1 ) 9.6 (0.6) Table 3.2: Differences between PD patients with VH (PD + VH; n=14), PD patients without VH (PD VH; n=14) and healthy controls (HC; n=14) on the different subtests for object and space perception ject decision, Number location and sustained visual attention (see Table 3.3). 3.5 Discussion Sustained visual attention and object and space perception were investigated in PD patients with VH, compared to PD patients without VH and healthy controls. PD patients with VH showed decreased sustained visual attention, compared to both other groups, independent of visual acuity or disease severity. These results are consistent with studies reporting a decreased ability to focus attention and fluctuating levels of attention in respectively PD patients with VH (Barnes and Boubert, 2008) and patients with PD dementia, with very frequent VH (Ballard et al., 2002). Our results also showed that PD patients with VH scored significantly lower on the object perception test Object decision compared to both PD patients without VH and healthy controls, independently of visual acuity and disease severity. Importantly, no differences were found between PD patients without

45 36 CHAPTER 3. ATTENTION AND PERCEPTION IN PD WITH VH VH and healthy controls, which confirms previous data (Barnes et al., 2003; Meppelink et al., 2008). The Object decision test used in this study required participants to identify a silhouette of a real object amongst the silhouettes of three nonsense objects. These silhouettes showed a lack of perceptual clarity and a lack of detail. Previously this was explained by suggesting that PD patients with VH can not resolve ambiguities of these stimuli and therefore activate previously stored internal schemata in higher-level visual areas (Barnes et al., 2003). These suggestions are consistent with the models that explain (visual) hallucinations (Aleman et al., 2008; Collerton et al., 2005; Diederich et al., 2005) and suggest that perception in patients with VH is influenced by top-down factors, such as perceptual expectations and visual representations of the exterior world, which are stored in memory. In addition, PD patients with VH showed a decreased performance on a space perception test (Number location), relative to both PD patients without VH and healthy controls. It is not likely that perception during this task was influenced by visual representations stored in memory or perceptual expectations, however this does not exclude the possibility that PD patients with VH also increasingly relied on top-down processes during this task. Previously we mimicked suboptimal visual situations by creating movies of undegraded objects dynamically appearing out of random noise, with no direct manipulations on the objects themselves and showed that PD patients with VH had no difficulties recognizing objects, they were however significantly slower than PD patients without VH and healthy controls (Meppelink et al., 2008). The present study extends these findings by showing that PD patients with VH do show object perception impairments if objects are manipulated without taking into account the background environment, specifically if the objects that need to be recognized lack perceptual clarity and detail. PD patients with VH thus show impairments if the objects that need to be recognized lack clarity, not if the surroundings of the object (random noise) are unclear. Decreased levels of sustained visual attention were correlated with a decreased object and space perception in PD patients with VH. Importantly, no correlations were found between sustained visual attention and object and space perception in PD patients without VH, nor in healthy controls. Since sustained visual attention and object and space perception are both processes in which bottom-up and top-down factors interact, these results confirm the models that have been introduced to explain (visual) hallucinations (Aleman et al., 2008; Collerton et al., 2005; Diederich et al., 2005). A possible explanation for

46 3.5. DISCUSSION 37 Figure 3.3: Course of Object decision, Number location and sustained visual attention. Line diagram showing the performances on the Test battery for Attention Performances (TAP) and the tests Object decision and Number location of the VOSP in healthy controls, PD patients without VH (PD - VH) and PD patients with VH (PD + VH). this correlation could be that the decreased performance of PD patients with VH on the object and space perception tests was influenced by an impaired sustained visual attention. However, since the object and space perception tests were presented in an at random order, this is not likely. It is more likely that sustained visual attention and the perception of objects and spaces are two cognitive functions that both decline with the progression of PD (see Fig. 3.3) and both play a role in the pathogenesis of VH. It should be noted that PD patients with VH did not show a decreased performance on all subtests for object and space perception. A first explanation for these findings may be the fact that participants who obtained a score of 24 or lower on the MMSE were excluded. Consequently, the group of cognitively most severely affected PD patients, in which VH are common, were not included in this study. Secondly relatively small groups of participants were included in this study, causing a relatively small distribution of data. The effect sizes and power of our main findings (differences on sustained visual attention and Object decision) were however high, suggesting that the decreased sustained visual attention and object perception are evident in PD patients with VH and can even be detected in small groups. All other results show low to average effect sizes and power and thus need to be confirmed in a larger samples of subjects.

47 38 CHAPTER 3. ATTENTION AND PERCEPTION IN PD WITH VH 3.6 Conclusion In conclusion, the results of this study suggest that in PD patients with VH impairment of visual processing of objects, possibly in association with a decreased sustained visual attention, might play a role in the pathogenesis of VH.

48 Chapter 4 Impaired visual processing preceding image recognition in Parkinson s disease with visual hallucinations A. M. Meppelink 1,2,3, B. M. de Jong 1,2,3,R.Renken 2, K. L. Leenders 1,2,3, F. W. Cornelissen 3,4, T. van Laar 1,2,3 (1) Department of Neurology, University Medical Center Groningen, The Netherlands (2) Neuro Imaging Center (NIC) Groningen, University of Groningen, the Netherlands (3) School of Behavioral and Cognitive Neurosciences, University of Groningen, the Netherlands (4) Laboratory of Experimental Ophthalmology, University Medical Center Groningen, University of Groningen, the Netherlands Accepted for publication in Brain 39

49 40 CHAPTER 4. IMPAIRED VISUAL PROCESSING IN PD WITH VH 4.1 Abstract Impaired visual processing may play a role in the pathophysiology of visual hallucinations in Parkinson s disease. In order to study involved neuronal circuitry, we assessed cerebral activation patterns both before and during recognition of gradually revealed images in Parkinson s disease patients with visual hallucinations, Parkinson s disease patients without visual hallucinations and healthy controls. We hypothesized that, before image recognition, Parkinson s disease with visual hallucinations would show reduced bottom-up visual activation in occipital-temporal areas and increased (pre)frontal activation, reflecting increased top-down demand. Overshoot of the latter has been proposed to play a role in generating visual hallucinations. Nine non-demented Parkinson s disease patients with visual hallucinations, 14 Parkinson s disease patients without visual hallucinations and 13 healthy controls were scanned on a 3 Tesla MRI scanner. Static images of animals and objects gradually appearing out of random visual noise were used in an eventrelated design paradigm. Analyses were time-locked on the moment of image recognition, indicated by the subjects button-press. Subjects were asked to press an additional button on a colour-changing fixation dot, to keep attention and motor action constant and to assess reaction times. Data pre-processing and statistical analysis were performed with SPM5. Bilateral activation of the fusiform- and lingual gyri was seen during image recognition in all groups (p<0.001). Several seconds before image recognition, Parkinson s disease patients with visual hallucinations showed reduced activation of the lateral occipital cortex, compared to both Parkinson s disease patients without visual hallucinations and healthy controls. In addition, reduced activation of extrastriate temporal visual cortices was seen just before image recognition in Parkinson s disease patients with visual hallucinations.

50 4.2. INTRODUCTION Introduction Parkinson s disease (PD) is a multisystem neurodegenerative disorder, in which deterioration of dopaminergic neurons in the substantia nigra, that project to the striatum, is a classical hallmark (Lang and Lozano, 1998; Braak et al., 2004). Motor symptoms like bradykinesia, rigidity and tremor are dominant characteristics of PD, while non-motor symptoms such as cognitive impairment and visual hallucinations (VH) may additionally occur (Aarsland et al., 2005; Barnes and David, 2001). VH in PD typically consist of complex visual, commonly moving, images lasting for seconds to minutes, experienced in the alert state with eyes open, affecting percent of all PD patients (Barnes and David, 2001; Williams and Lees, 2005). Animals, people and objects define the three categories of images that are most frequently seen by hallucinating PD patients. Mostly these VH are non-threatening and the patient maintains insight in the fact that the experiences do not reflect real events (Fenelon et al., 2000; Barnes and David, 2001). Nevertheless, VH may progress to hallucinations with loss of insight or delusions in 80 percent and constitute an important risk factor for nursing home placement (Goetz and Stebbins, 1993). In the present study, we employed fmri in PD patients with VH (PDwithVH), PD patients without VH (PDnonVH) and healthy control subjects. By assessing functional differences in brain regions implicated in visual perception, we aimed to gain more insight in the origin of VH in PD. In this respect we were particularly interested in visual processing stages preceding actual image recognition. Hallucinations are defined as involuntary perceptual experiences in the waking state without external visual stimulation (Collerton et al., 2005). Particularly auditory and visual hallucinations are core symptoms in schizophrenia and can be treated with dopamine receptor antagonists. This suggests involvement of the striatum in the pathophysiology of hallucinations. Such association is indeed supported by functional imaging with H 2 O-PET in schizophrenia, showing increased striatum (and cingulate) activation during hallucinations, together with activations in distinct auditory- and visual cortical regions during respectively auditory- and visual hallucinations (Silbersweig et al., 1995). VH may, however, also occur without psychiatric or neurological disease. For example, in Charles Bonnet Syndrome (CBS) complex VH occur secondary to profound visual loss in cognitively normal people (Teunisse et al., 1996). These observations have provided support for the concept that impaired processing of externally presented stimuli may lead to an increased reliance on top-down mechanisms, in which an internal generator may contribute to activations in

51 42 CHAPTER 4. IMPAIRED VISUAL PROCESSING IN PD WITH VH appropriate perceptive cortical regions (Silbersweig et al., 1995; Stebbins et al., 2004) Visual Hallucinations in Parkinson s disease With regard to VH in PD, a combination of impaired visual processing and attention has been reported (Flowers and Robertson, 1995; Collerton et al., 2005; Meppelink et al., 2008; Diederich et al., 2005). VH in PD have been commonly viewed as an adverse effect of dopaminergic treatment for PD, causing a relative overstimulation of the limbocortical dopaminergic receptors (Bosboom et al., 2004). However, VH in PD may not be associated with the dose or duration of treatment of dopaminergic drugs (Goetz et al., 1998; Holroyd et al., 2001). Moreover, VH have already been reported in the pre-levodopa era (Fenelon et al., 2006). Neuropathologically, VH in PD are associated with increased Lewy body deposition in the temporal lobe (Harding et al., 2002), suggesting that VH in PD are at least partially caused by the disease itself. With fmri, increased caudate activation has been demonstrated during visual stimulation in PDwithVH, compared to PDnonVH, thus showing some resemblance with the above described functional imaging findings in schizophrenia (Stebbins et al., 2004). This raises the question how the basal ganglia play a role in the generation of hallucinations. The basal ganglia are involved in switching behavior to internal or external sensory stimuli (Redgrave et al., 1999). Their well-structured interconnections with the cerebral cortex by cortico-basal ganglia-thalamo-cortical circuits are organized in parallel loops, which enable the regulation of normal adaptive behavior by selection of motor and non-motor behavioral responses (Redgrave et al., 1999; de Jong and Paans, 2007). Some imaging studies have further shown hyperperfusion of the frontal lobe during VH in PD (Kataoka et al., 2008) and schizophrenia (Silbersweig et al., 1995). Joint activations in the basal ganglia and frontal lobes might thus reflect an aspect of internal image generation in these patient groups. In patients with PD, schizophrenia and CBS increased perfusion or activation of visual association cortices was seen during the occurrence of VH (Kataoka et al., 2008; Silbersweig et al., 1995; Ffytche et al., 1998), while in other studies reduction of either activation, perfusion or metabolism in visual association cortices was seen during rest or simple visual stimulation (Stebbins et al., 2004; Okada et al., 1999; Matsui et al., 2006a; Boecker et al., 2007). In CBS, the latter likely reflects reduced visual cortical processing due to visual deprivation (Ffytche et al., 1998). In PDwithVH, cortical visual processing itself seems to be impaired, as explained below.

52 4.2. INTRODUCTION Impaired visual processing A wide range of visual perceptual disturbances has been associated with VH in PD, including reduced visual acuity (Matsui et al., 2006b), contrast sensitivity, colour discrimination (Diederich et al., 1998), visual space perception (Ramirez-Ruiz et al., 2007a) and visual object perception (Barnes et al., 2003; Ramirez-Ruiz et al., 2006). Relative hypometabolism of the ventral visual stream in PD patients compared to patients with progressive supranuclear palsy has been proposed to reflect the vulnerability for VH particularly in PD (Klein et al., 2005). Relatively impaired visual processing in PDwithVH could hypothetically lead to compensatory visual processing and internal image generation. In this respect, some resemblance can be seen with the occurrence of VH in CBS, although the cause of underlying visual dysfunction is different. VH in PD tend to occur in dim suboptimal visual circumstances, mostly during the evening (Fenelon et al., 2000). In suboptimal visual circumstances, top-down processes are considered to play an important role in the recognition of objects (Bar et al., 2006). It is thought that a partially analyzed version of the input image is rapidly projected from early visual areas to the prefrontal cortex, where it activates an initial guess, which is projected back to the temporal cortex (Bar et al., 2006). There, it is integrated in bottomup visual processing, in which the visual image is processed from V1 to the occipital-temporal cortex. Several techniques have been used to mimic these suboptimal visual situations, such as backward masking, in which briefly presented images are immediately followed by a masking stimulus, and gradual revelation of objects using panels and visual noise (Bar et al., 2001; James et al., 2000; Grill-Spector et al., 2000). By presenting images that slowly and dynamically appear out of random noise, the speed and content of conscious perception of images can be assessed (James et al., 2000; Reinders et al., 2006; Kleinschmidt et al., 2002; Meppelink et al., 2008). This dynamic presentation of stimuli probably mimics a situation of more natural visual perception. Previous data from our own group has shown that non-demented PDwithVH were slower in recognizing images dynamically popping-out of noise, compared to both PDnonVH and healthy controls (Meppelink et al., 2008) Hypothesis The aim of the current study was to investigate the distribution of cerebral activations during visual processing in non-demented PDwithVH compared to non-demented PDnonVH and healthy controls. Our hypothesis was that in PDwithVH, impaired bottom-up visual processing induces top-down com-

53 44 CHAPTER 4. IMPAIRED VISUAL PROCESSING IN PD WITH VH pensation. During actual hallucinations, this compensation may change into an overshoot of top-down exerted activity, resulting in an increase of activations within a visuo-frontal neuronal network. In the present fmri study, we did not aim to detect the effect of actual hallucinations, because the included PDwithVH did not perceive them during scanning. The employed paradigm was designed to provoke and identify successive activation patterns during the observation of a display with gradually revealed images. By focusing particularly on activation in the stage preceding image recognition, we aimed to investigate whether the previously found delay in image recognition in PDwithVH (Meppelink et al., 2008) is reflected in changed activations in respectively ventral extrastriate visual cortex, basal ganglia and prefrontal cortex. 4.3 Methods Subjects Thirty-six subjects participated in this study, divided in 3 groups; 9 PD patients who experienced complex VH at least weekly during the last month, 14 PDnonVH and 13 healthy controls. Originally, 12 PD patients with VH and 14 healthy controls were included in this study. Two subjects (one healthy control and one PDwithVH) were excluded because of motion artifacts and two PDwithVH were excluded because they were unable to perform the task as instructed once they were in the scanner. E.g., they pressed the recognitionbutton several times per movie while only one response was requested. PD was diagnosed according to the criteria of the UK PD Society Brain Bank. These three groups were matched for age and level of education. The latter was rated with a Dutch education scale ranging from 1 (elementary school not finished) to 7 (university degree). Both PD groups were also matched for cognition [assessed with the Mini Mental State Examination (MMSE) (Folstein et al., 1975)] and for their level of executive functioning [assessed with the Frontal Assessment Battery (FAB) (Dubois et al., 2000)]. All PD patients were on during the assessment. The levodopa-equivalent daily dose (LEDD) was calculated for all patients, according to the formula: LEDD=levodopa dose (mg) + (0.3 *levodopa dose if using entacapone with each dose) + (slow release levodopa *0.7) + (bromocriptine *10) + (ropinirole *20) + (pergolide *100) + (pramipexole *100) + (apomorphine *10) (Esselink et al., 2004). Visual acuity was assessed with the Snellen chart. Demographic and clinical characteristics are described in table 4.1. Exclusion criteria were dementia (MMSE score < 24), neurological disorders other than PD, psychiatric disor-

54 4.3. METHODS 45 PD + VH PD - VH HC M(SD) range M(SD) range M(SD) range Age (years) 61.2 (8.2) (7.8) (7.5) Education 5.7 (0.7) (0.7) (0.8) 5-7 Disease duration 8.1 (5.0) (4.7) 4-24 LEDD 855 (543) (362) UPDRS-III 21.4 (7.0) (7.3) 6-30 MMSE 26.8 (1.0) (1.3) (0.9) FAB 15.7 (1.8) (1.5) (1.2) Visual acuity* (Snellen) 0.94 (0.1) (0.1) (0.1) Males: n (%) 5 (55 %) 11 (79 %) 9 (69 %) Females: n (%) 4 (45 %) 3 (21 %) 4 (31 %) Table 4.1: Demographic and illness characteristics of Parkinson s disease patients with visual hallucinations (PD+VH; n=9), PDnonVH (PD-VH; n=14) and healthy controls (HC; n=13). * Corrected with glasses, when necessary. ders, visual acuity less than 50 percent (Snellen chart) and visual field defects. This study was approved by the Medical Ethical Committee of the University Medical Center Groningen. All participants signed an informed consent prior to study inclusion Clinical tests and statistics Contrast sensitivity was assessed in all subjects using the Mars contrast sensitivity test (Arditi, 2005). All PD patients were asked to fill in a self-report depression scale [Beck Depression Inventory, BDI (BECK et al., 1961)]. The severity of motor symptoms in PD patients was rated with the Unified PD Rating Scale (UPDRS), part III. Severity of VH in PD patients was assessed with part B Hallucinations of the Neuropsychiatric Inventory and a questionnaire based on the characteristics of VH in PD patients as described by Barnes and David (Barnes and David, 2001). Not all variables were normally distributed in all three groups. Therefore, the non-parametric Mann-Whitney and Kruskal-Wallis tests were used to investigate these non-normally distributed variables. An ANOVA was used to compare normally distributed variables. The three groups were compared concerning their contrast sensitivity and visual acuity using the Kruskal-Wallis test. In addition, the Mann-Whitney test was used to investigate firstly, if PDwithVH differed from PDnonVH and secondly, if PDnonVH differed from healthy controls. Differences in age were investigated using an ANOVA. Scores of PD patients on the BDI, LEDD, MMSE and UPDRS-III were compared using the Mann-Whitney test.

55 46 CHAPTER 4. IMPAIRED VISUAL PROCESSING IN PD WITH VH Figure 4.1: fmri paradigm and design. (A) Image of a dog, gradually appearing out of noise. Subjects pressed a button when the image was recognized. An additional button press was required on an infrequently colour changing fixation square, to keep attention constant and to assess reaction times. (B) fmri design; 1. Image recognition movies start with 100 random white noise at t=0 sec. Images gradually appear, with the pop-out (recognition) being between 0 and 30 seconds; 2. Hemodynamic Response following the pop-out; 3. Visual percept of the image; 4. Block of 30 seconds, representing motor response on the colour change and visual input.

56 4.3. METHODS fmri paradigm and experimental procedure A total of 50 pictures of animals (22), well-known objects (22) and meaningless objects (6, control) were used to create a paradigm in which pictures gradually pop out of random uniform visual white noise (figure 4.1A). Movie stimuli were generated in Matlab 5 on an Apple Macintosh computer running Mac OS using some of the routines of the Psychtoolbox (Brainard, 1997; Pelli, 1997). Movies were created from gray scale pictures that were first normalized to have their mean luminance equal to the background level. Noise contrast remained constant throughout the duration of the 30-second movie. Image contrast (and thus signal-to-noise) increased linearly over time causing the image to gradually appear out of the noise. Perceptual recognition ( pop-out ) occurred from 10 to 28 seconds after initial movie onset. All movies were created from grey-scale pictures with a resolution of of 300 by 300 pixels. Movies were shown at twice this size (600x600 pixels). The movies were presented using the Presentation program (Neuro Behavioral Systems, Inc. CA, USA). ÄThey were projected by a beamer (resolution 1024 x 768 pixels, Barco, Belgium) on a screen (display dimensions 44 x 34 cm), viewed by the subject via a mirror placed at a distance of 11 cm from the face. The distance between the mirror and the screen was 64 cm and the stimuli covered approximately 18 degrees of the visual field. If necessary, visual acuity of the subject was corrected using MRI-compatible lenses. During presentation of the movie, a central fixation square changed colour with random intervals. Subjects had to report such change (to keep attention constant) by pressing a button with their right middle finger on an MR compatible response-box (forp, Current designs, Inc. U.S.A.). Per subject, the mean reaction time of the response to the colour change was calculated. Subjects were further instructed to press a button with their right index finger at the moment that they recognized the object or animal, i.e. at the moment of the perceptual pop-out. Before each session, this paradigm was practiced outside the scanner, while verbal responses were used to verify recognition of the images. In this way, we assessed whether subjects understood the task correctly and whether they indeed recognized the images. The mean reaction time on the colour change was subtracted from the image recognition times. Movie stimuli were presented in two runs, 25 per run. In between the two runs, an anatomical, T1 weighed scan was acquired.

57 48 CHAPTER 4. IMPAIRED VISUAL PROCESSING IN PD WITH VH MRI characteristics Data acquisition was performed using a 3 T Philips MR system (Best, The Netherlands) with a standard 6 channel SENSE head coil. Functional images were acquired with a gradient echo, i.e. echo planar imaging, T2* Blood Oxygen Dependent Level (BOLD) contrast technique in an ascending order with a TE of 35 ms, a TR of 2.3 seconds, 35 slices per TR, 450 volumes per run, isotropic voxels 3 x 3 x 3 mm and an axial orientation. A T1 weighted 3D anatomical scan was acquired to obtain high resolution anatomical information, isotropic voxels 1 x 1 x 1 mm, matrix size = 256 x 256 and an axial orientation Psychophysics The mean reaction time on the colour change of the fixation square was calculated per subject and averaged over groups. Reaction times were not normally distributed; differences between groups were investigated using the nonparametric Kruskal-Wallis test. The Mann-Whitney test was used to investigate between which groups differences exist (PDwithVH versus PDnonVH, PDwithVH versus healthy controls, PDnonVH versus healthy controls). The mean image recognition time over all movies was calculated per subject and averaged over groups. The percentage of unrecognized images was calculated per subject and averaged over groups as well. Unrecognized movies were considered as missing values. Mean image recognition times were normally distributed, differences between groups were investigated using ANOVA. A Helmert contrast was used to determine firstly, if healthy controls differed significantly from both PDnonVH and PDwithVH and secondly, if PDnonVH differed from PDwithVH fmri data analysis Image processing and statistical analysis were conducted with Statistical Parametric Mapping (SPM, Friston et al. (1995)) version 5 (2005, Wellcome Department of Cognitive Neurology, London, UK; http : // il.ion.ucl.ac.uk/spm). Pre-processing included slice time correction, realignment, coregistration of functional and anatomical scans and spatial normalization (to the template of the Montreal Neurological Institute, MNI). Images were smoothed using a Gaussian filter of 8 mm FWHM. Analyses were time-locked on the perceptual pop-out in an event-related design. In addition to the hemodynamic response

58 4.4. RESULTS 49 function (HRF), temporal and dispersion derivatives were modeled as well (Friston et al., 1998). Apart from the pop-out, a 30 seconds block of visual input and motor response as well as a block of the visual percept, lasting from the pop-out until the end of the movie, were modeled (figure 4.1B). Movement parameters were included as covariates. T-contrasts of the pop-out with respect to baseline (passive viewing of a fixation cross, projected on a dark background) were made for each subject. A Finite Impulse Response (FIR) analysis was used to investigate temporal dynamics. Again analyses were timelocked to the pop-out (minus 3TR). Basis functions (FIR) were generated using the following settings: duration 3.5TR, order 7. Thus, mean whole brain activations in seven time bins of 0.5 TR (=1.15 seconds) each, were assessed from 3TR (=6.9 seconds) before pop-out until 0.5TR (=1.15 seconds) after the pop-out. On a 2nd level (random effect analysis), the T-contrasts of the pop-out were analyzed per group, using an ANOVA (flexible factorial). The ANOVA was also used to make comparisons between PDwithVH, PDnonVH and healthy controls. FIR data on a 2nd level were also analyzed using an ANOVA (flexible factorial), to assess temporal dynamics on a group level and to make group comparisons. FIR results for the six time bins before pop-out, the seventh bin being non-informative due to excess activation at the given threshold, were rendered on 3D standard MNI brains. Regions of interest (ROIs) were defined based on the FIR group comparisons that showed significant activation differences in PDwithVH, compared to PDnonVH and healthy controls. Time courses of activation in these ROIs were plotted for all three groups. 4.4 Results Characteristics subjects The three groups were matched for age (ANOVA: F=2.13, p=0.13) and level of education (Kruskal-Wallis Test: χ 2 =0.82, p=0.70). PD groups were matched for cognition (MMSE; z=-1.09, p=0.27) and executive functioning (FAB; z=- 1.77, p=0.08). In the group of 9 PDwithVH, 1 reported having VH about once a week, 6 had VH several times per week and 2 reported having VH several times a day. Thirty-three percent of the hallucinating PD patients reported that they became upset during their VH and 44 percent considered their VH as a moderate to severe emotional burden. Fourty-four percent experienced VH of people and animals, 33 percent of only people, 11 percent of only animals and 11 percent of animals and objects. None of the PDwithVH experienced VH

59 50 CHAPTER 4. IMPAIRED VISUAL PROCESSING IN PD WITH VH during the testing. Neither did they report a relationship between their VH and the intake of dopaminergic medication. LEDD scores were also not significantly different in PDwithVH, compared to PDnonVH (t=0.33, p=0.74). Two subjects (PDwithVH) used clozapine for their VH, while none of the subjects used cholinesterase-inhibitors or anti-cholinergics. Of PDwithVH, 2 patients also experienced visual illusions and 2 other patients from this group occasionally experienced presence of a person hallucinations. With regard to visual abilities, the Kruskal-Wallis showed a significant difference on contrast sensitivity (χ 2 =9.19, p=0.01) between the three groups. Healthy controls had significantly higher contrast sensitivity compared to both PDnonVH (z=-2.88, p=0.004) and PDwithVH (z=-2.14, p=0.032), while no significant differences were found between the two PD patients groups (z=-0.20, p=0.84). Visual acuity was equal in all three groups (Kruskal-Wallis: χ 2 =1.83, p=0.40) Task performance All subjects were able to perform the task adequately. Mean reaction time on the colour change of the fixation square (standard deviation) was 761 (150) ms for PDwithVH, 814 (337) ms for PDnonVH and 575 (126) ms for healthy controls. The Kruskal-Wallis test showed significant differences between groups (χ 2 =11, p=0.004). The Mann-Whitney showed no significant difference between PDwithVH and PDnonVH (z=-0.25, p=0.80), while both PDwithVH and PDnonVH were significantly slower, compared to healthy controls (z=-2.6, p=0.009 and z=-3.0, p=0.003, respectively). Some pop-out movies were more difficult than others, leading to delayed recognition of the image. The mean percentage of images that were recognized in healthy controls was 96 percent, whereas the mean percentage of recognized images was 86 percent in PDnonVH and only 76 percent in PDwithVH. Mean image recognition time (standard deviation) over all movies was (1.82) sec in the healthy controls, (2.18) sec in PDnonVH and (2.23) sec in PDwithVH. The ANOVA showed significant differences between groups for the mean image recognition time (F=7.58, p=0.002). The Helmert contrast showed that healthy controls were significantly faster than PDnonVH and PDwithVH (p=0.001), while PDwithVH and PDnonVH had the same mean image recognition time (p=0.98).

60 4.4. RESULTS Activation at the moment of pop-out At the moment of pop-out, robust bilateral activations were seen in the fusiform gyrus and lingual gyrus in all groups (p<0.001, corrected, see figure 4.2 and table 4.2A). When groups were compared, no differences were observed Activation before pop-out, within group analysis The dynamic presentation in which images gradually appeared was associated with the temporal evolution of a changing pattern of cerebral activations that was different for each of the three groups (figure 4.3A). In the healthy controls, the fusiform gyrus was already activated before pop-out (p<0.001, corrected at -1.2 pop-out). PDnonVH did not show significant activation of the fusiform gyrus before pop-out, but showed activation of the middle occipital cortex and the inferior frontal gyrus at this stage (p<0.001, corrected and p<0.001, uncorrected, respectively). PDwithVH did not show significant activation of the fusiform gyrus before pop-out either, but showed activation of the parietal cortex bilaterally. At a lower threshold (p<0.05), the occipital and frontal cortices were activated as well in PDwithVH (data not shown) Activation before pop-out, group differences PDwithVH, compared to PDnonVH and healthy controls PDwithVH showed a significant reduction of cortical activation, compared to both PDnonVH and healthy controls, while no increases were seen in PDwithVH when compared to the other two groups. The decrease of activation in PDwithVH was already seen several seconds before pop-out. All changes in activation before pop-out at p<0.001 (cluster-level, uncorrected for the entire brain volume) are reported (table 4.2B and C). After volume correction, corrected statistical significance concerning these effects are partly influenced by the summed size of confluent clusters. At -5.8 sec -4.6 sec PDwithVH showed significantly decreased activation of bilateral occipital cortex compared to both PDnonVH (p<0.001, uncorrected) and healthy controls (p<0.001, corrected) In addition, PDwithVH showed significantly decreased activation of the left inferior parietal cortex, compared to healthy controls (p<0.001, uncorrected, see figure 4.3B and table 4.2B and C). At -3.5 sec -2.3 sec PDwithVH showed decreased activation of the superior frontal gyrus,

61 52 CHAPTER 4. IMPAIRED VISUAL PROCESSING IN PD WITH VH Figure 4.2: Activation during image recognition. Activation of the fusiform gyrus (1) and lingual gyrus (2) during image recognition in healthy controls (A), PDnonVH (B) and PDwithVH (C).

62 4.5. DISCUSSION 53 compared to PDnonVH (p<0.001, corrected), while a trend towards decreased activation in this region was seen in PDwithVH when compared to healthy controls (data not shown). At -1.2 sec pop-out, PDwithVH showed significantly decreased activation of the fusiform gyrus bilaterally and the left lingual gyrus, compared to both PDnonVH and healthy controls (p<0.001, corrected). In addition, PDwithVH showed decreased activation of the cingulate cortex and the right middle frontal gyrus, compared to both PDnonVH and healthy controls (p<0.001, uncorrected and corrected, respectively). No increased activations in PDwithVH were observed, compared to either PDnonVH or healthy controls. Time courses of pre-pop-out effects the left fusiform gyrus, which were similar to that of the right fusiform gyrus, the inferior- and middle frontal gyri, are plotted in figure Discussion The paradigm applied in this fmri study was designed to identify activation changes in circuitry particularly related to visual processing preceding image recognition. With the visual task involving the gradual revelation of complex images (animals and objects), we demonstrated that PDwithVH, compared to PDnonVH and healthy controls, had similar activation patterns at the moment of image recognition, but that marked differences occurred in the stage preceding image recognition. This underscored the importance to apply the dynamic paradigm with gradually revealed images, instead of clear static images Visual object processing The moment of pop-out, i.e. the moment at which the images of animals and objects were recognized after gradually appearing out of random noise, was related to marked activation in appropriate secondary visual regions such as the fusiform and lingual gyri in all groups. This provided support for the robustness of our paradigm. The mean time it took to recognize the image was similar in PDwithVH and PDnonVH, while both groups were significantly slower in recognizing the images than healthy controls. Because PDwithVH recognized only 76 percent of the images, compared to 86 percent in PDnonVH, these results are probably an underestimation. When images were not recognized in 30 seconds, a missing value was reported. If image recognition movies would have lasted

63 54 CHAPTER 4. IMPAIRED VISUAL PROCESSING IN PD WITH VH A. Activation at the moment of pop-out Contrast: Anatomical region MNI-coordinates T (voxel level) p-value (cluster level) HC fusiform R 30,-36, <0.001, corrected fusiform L -39,-69, <0.001, corrected lingual R 18,-42, <0.001, corrected lingual L -21,-54, <0.001, corrected PD fusiform R 30,-57, <0.001, corrected fusiform L -35,-43, <0.001, corrected lingual R 15,-45, <0.001, corrected lingual L -18,-57, <0.001, corrected VH fusiform R 30,-36, <0.001, corrected fusiform L -31,-73, <0.001, corrected lingual R 18,-45, <0.001, corrected lingual L -12,-69, <0.001, corrected B. Before pop-out: PDnonVH versus PDwithVH Contrast: Anatomical region MNI-coordinates T (voxel level) p-value (cluster level) PD2 - VH2 calcarine L -3, -81, <0.001, uncorrected occipital sup. R 18, -84, <0.001, uncorrected PD3 - VH3 occipital sup. R 24, -75, <0.001, uncorrected occipital mid. R 42,-81, <0.001, uncorrected PD4 - VH4 frontal sup. R 21, 30, <0.001, corrected PD6 - VH6 frontal mid. R 33, 39, <0.001, uncorrected cingulate ant. R 15, 33, <0.001, uncorrected cingulate mid. L -3, -15, <0.001, uncorrected lingual L -15, -84, <0.001, corrected fusiform gyrus L -33, -51, <0.001, corrected fusiform gyrus R 36, -63, <0.001, corrected C. Before pop-out: HC versus PDwithVH Contrast: Anatomical region MNI-coordinates T (voxel level) p-value (cluster level) HC2 - VH2 parietal inf. L -45, -51, <0.001, corrected occipital sup. L 18, -84, <0.001, corrected occipital sup. R 21, -78, <0.001, corrected occipital mid. R 33, -72, <0.001, corrected occipital mid. L -30, -78, <0.001, corrected HC6 - VH6 frontal mid. R 33, 42, <0.001, corrected lingual gyrus L -21, -57, <0.001, corrected cingulate mid. L -3, -12, <0.001, corrected fusiform gyrus R 36, -63, <0.001, corrected fusiform gyrus L -30, -36, <0.001, corrected Table 4.2: A. All three groups showed robust activation of bilateral fusiform and lingual gyrus (at p<0.001, k=20). In the group comparisons, no differences between groups were observed at the moment of pop-out (data not shown). Before pop-out, differences were observed between PD and VH (B) and between HC and VH (C) in the time frames 2-6 (2= sec, 3= sec, 4= sec, 5= sec, 6=-1.2 pop-out) at p<0.05, k=20. Relatively increased activation in PD, compared to VH (B) and also in HC, compared to VH (C) was seen (i.e. relatively decreased activation in VH, compared to both PD and HC). In 4.2B and 4.2C, statistical significance at cluster-level, after volume correction, may be influenced by the summed size of confluent clusters.

64 4.5. DISCUSSION 55 Figure 4.3: Cerebral activation patterns before image recognition. FIR results showing activations from 6.9 seconds before pop-out until pop-out of all groups separately at p<0.01 uncorrected, k=20 (A), group comparisons between PDwithVH and PDnonVH at p<0.05 uncorrected, k=20 (B) and group comparisons between PDwithVH and healthy controls at p<0.05 uncorrected, k=20 (C).

65 56 CHAPTER 4. IMPAIRED VISUAL PROCESSING IN PD WITH VH Figure 4.4: Time courses. Plots showing the time courses with standard error of the mean ( ) in healthy controls (HC), PDnonVH (PD) and PDwithVH (VH) from 6.9 sec before pop-out (-6.9 sec -5.8 sec, time bin 1) until 1.15 sec after pop-out (pop-out sec, time bin 7) longer than 30 seconds, and signal-to-noise consequently would have further increased, probably more patients would have been able to recognize the image, subsequently leading to a longer mean time until recognition [see also Meppelink et al. (2008)]. The time until image recognition was corrected for bradykinesia, by means of subtracting the mean reaction time per subject on the colour change from the recognition time. The fact that appropriate extrastriate visual areas were activated at pop-out in all three groups supported the adequacy of our strategy to define this event after correction for reaction time. Mean reaction times on the colour changing fixation square were similar in PDwithVH and PDnonVH, supporting our view that reduced activations were not explained by a generally reduced psycho-motor speed Impaired extrastriate visual processing By using FIR models, we were able to assess successive changes in the cerebral distribution of activations during the presentation of gradually revealed images. Just before and during pop-out, activation of occipito-temporal, inferior parietal and inferior prefrontal areas was seen in healthy controls. This is consistent with previous studies that have addressed visual perception and image recognition of gradually revealed images (James et al., 2000; Kleinschmidt

66 4.5. DISCUSSION 57 et al., 2002; Eger et al., 2006). We showed that specifically PDwithVH, as compared to both PDnonVH and healthy controls, had reduced activation of the lateral occipital cortex several seconds before pop-out and reduced activation of the ventral temporal cortex just before pop-out. This might indicate a disturbance at a processing stage beyond V1 in which the normal brain uses scant information to predict the structure of features in an impoverished scenery (Summerfield and Koechlin, 2008). Indeed cortical regions at the lateral and ventral occipito-temporal junctions are important for visual object recognition (Malach et al., 1995; Grill-Spector, 2003). The fusiform gyrus, lateral occipital complex and middle temporal gyrus are involved in the visual perception of a range of both living and non-living objects, while the parahippocampal gyrus is predominantly involved in the perception of scenes (Malach et al., 1995; Downing et al., 2006). Therefore our finding provides support for our first hypothesis that bottom-up visual processing is impaired in PDwithVH. It is important to notice that this impaired visual processing was independent of visual acuity, which was equal in all groups, or contrast sensitivity, which was similar in PD patients with and without VH. To what extend other variables, like the angle of view of objects or colour discrimination are involved, could be a topic for future research. Reduced activation of the lateral occipital cortex was already present from 5.8 seconds before pop-out in PDwithVH. Reduced activation of the inferior temporal cortex occurred just before pop-out in PDwithVH, while both occipital and temporal activation appeared to normalize (i.e. no differences between groups) when the image was fully perceived (see also figure 4.2). Following this early reduced activation of the lateral occipital cortex, significantly reduced activation before pop-out of frontal regions was seen, compared to PDnonVH, while a trend towards reduced frontal activation was seen when compared to healthy controls (data not shown). Just before pop-out, significantly decreased activation of the ventral temporal cortex was observed. Such a time-dependent defect underscores the importance to apply a dynamic paradigm in which images are gradually revealed, probably providing a more natural visual perceptual situation. Clinical studies have shown impaired object perception in PD patients compared to healthy controls (Laatu et al., 2004; Barnes et al., 2003). More interestingly, several studies have shown that non-demented PDwithVH had more severely impaired object perception than PDnonVH (Barnes et al., 2003; Ramirez-Ruiz et al., 2006; Meppelink et al., 2008). Impaired processing within cortical visual pathways has further been inferred from functional imaging with IMP-SPECT and FDG-PET that have demonstrated baseline hypoperfusion or hypometabolism of temporal gyri (Boecker et al., 2007; Matsui et al., 2006a;

67 58 CHAPTER 4. IMPAIRED VISUAL PROCESSING IN PD WITH VH Okada et al., 1999; Oishi et al., 2005) and the occipital cortex (Matsui et al., 2006a; Nagano-Saito et al., 2004; Boecker et al., 2007) in PDwithVH, compared to PDnonVH. One of these studies reported hypoperfusion of the fusiform gyrus, together with hyperperfusion of the middle and superior temporal lobe (Oishi et al., 2005). Structural indicators for impaired visual cortex function in PD are further provided by MRI that has revealed atrophy of the (occipital) lingual gyrus in PDwithVH, compared to PDnonVH and healthy controls (Ramirez-Ruiz et al., 2007b). Future studies might investigate whether cortical atrophy is confined to the occipital lobe or if the temporal cortex is also affected. Finally, neuropathological examination has shown that a higher density of Lewy bodies in the temporal lobe was associated with VH in PD (Harding et al., 2002). These findings lead to the conclusion that impairment of the ventral and lateral extrastriate visual cortex in PD patients indicates a risk for the occurrence of VH. Partial sensory deprivation is a long-appreciated risk factor for hallucinations in other clinical contexts like CBS. In CBS, complex VH occur in patients with acquired visual impairment without psychiatric disorders. The current view on its pathogenesis is that VH result from the release of visual cortical activity following the loss of visual inputs, also referred to as deafferentiation (Burke, 2002). Relative impairment of visual processing in the occipito-temporal cortex in PDwithVH might thus predispose to the occurrence of VH via a release of higher order visual cortices within attentional and visual perceptual networks like the prefrontal cortex (Collerton et al., 2005) Fronto-parietal visual processing Our second hypothesis was that reduced bottom-up visual cortical processing in PDwithVH might lead to an increased reliance on top-down visual processing, reflected by activation of the basal ganglia and/or prefrontal cortex (see e.g. Silbersweig et al. (1995)). No such increases of activation were observed, which implied that we did not gain arguments in favor of a compensatory role of these systems during visual processing in PDwithVH. In contrast, in the period before image recognition decreased activation of the right superior frontal gyrus was seen in PDwithVH, compared to PDnonVH, and decreased activation of the middle frontal gyrus was seen in PDwithVH compared to both PDnonVH and healthy controls. In addition, a decreased activation of the inferior parietal cortex was seen in PDwithVH, compared to healthy controls only. The ventral prefrontal- and inferior parietal regions have

68 4.5. DISCUSSION 59 been implicated in previous studies that investigated visual perception of gradually revealed images in healthy controls and seem to play an important role in the integration of sensory and mnemonic information. Like in our healthy control group, visual awareness (i.e. perception) in these studies activated a network of ventral visual cortex, inferior frontal gyrus and lateral/inferior parietal cortex (Kleinschmidt et al., 2002; Eger et al., 2006). These parietal and prefrontal cortex activations were interpreted as involvement of higher order areas in top-down facilitation of image recognition (Eger et al., 2006). So it seems that, apart from impaired bottom-up visual processing before image recognition, a broader network including especially frontal cortical areas involved in top-down processing is impaired as well in PDwithVH. A recent fmri study has also shown reduced activation of the right ventrolateral prefrontal cortex during face perception in cognitively impaired PDwith VH, compared to both PDnonVH and healthy controls (Ramirez-Ruiz et al., 2008). Dysfunction of the lateral prefrontal cortex was proposed to reflect a deficit in suppression of irrelevant stimuli, which might predispose to VH (Ramirez-Ruiz et al., 2008). Another implication of the reduced activation of ventro-lateral prefrontal cortex in PDwithVH is that these patients may have reduced tendency to address external stimuli. While lateral prefrontal regions are associated with externally cued behavior, medial prefrontal activation is associated with internally guided behavior (de Jong and Paans, 2007). The superior frontal gyrus plays a role in endogenous allocation and maintenance of visual attention and was shown to be involved in the inhibition of internally represented information (Corbetta et al., 2002; de Jong and Paans, 2007). In the present study, frontal hypoactivation was only seen in the stage preceding image recognition and not when the image was fully perceived. In contrast, Ramirez-Ruiz et al. found relative frontal hypoactivation in PDwithVH during image perception (Ramirez-Ruiz et al., 2008). One might speculate that the preserved prefrontal activation during image recognition in the present study reflected the relative cognitive preserved patient sample in our study. Generally, it seems that non-demented PD patients with VH indeed have more impairments in executive functioning, when compared to PD patients without VH, who again are worse than healthy control subjects [for example Barnes and Boubert (2008)]. The influence of executive dysfunction is widespread and might therefore explain a considerable part of behavioral and imaging results. Even in PD patients with dementia, MMSE scores can be relatively spared, while scores on the FAB are decreased and thus seem to better reflect cognitive impairments in PD patients. Executive dysfunction might indeed be related to lower performance in perceptual tasks, making it unclear whether visual

69 60 CHAPTER 4. IMPAIRED VISUAL PROCESSING IN PD WITH VH perception itself is impaired or whether lower scores (or activations) reflect executive dysfunction. Our PD patient groups were matched on both cognition and executive functioning, making it less likely that the observed decreased activations before image recognition are directly associated with one of these factors. Possibly, results would have been even stronger when patients were not matched on executive functioning, but in that case conclusions regarding the perceptual problems in PDwithVH would have been less strong. Because a trend towards lower scores on executive functioning was seen in PDwithVH, the FIR analysis was repeated with FAB scores as a covariate, which had no effect on the results (data not shown). Frontal regions might also be implicated in sustained attention (Johannsen et al., 1997), which was shown previously to be more impaired even in cognitively preserved PDwithVH, compared to PDnonVH, who again performed worse than healthy controls (Meppelink et al., 2008). Although we have corrected for attention, i.e. all subjects in all groups had adequate attention during performance of the task, this reduced activation of the middle frontal gyrus might still reflect subtle underlying attentional deficits in PDwithVH VH in Parkinson s disease, proposed mechanism Because none of the participating subjects experienced VH during scanning, our pop-out movies are an indirect way to measure functional cerebral impairments associated with VH. While in the present study we showed that the ventral/lateral temporal cortex and part of the prefrontal cortex were relatively impaired in PDwithVH, one may still assume that activation increases occur in these regions during VH in these patients. It was shown before that perfusion of the inferior frontal gyrus was increased during VH of a spider in one PD patient, together with increased perfusion of visual association areas (Kataoka et al., 2008). A comparable cerebral activation pattern was seen during hallucinations in patients with schizophrenia. Auditory and visual association cortices showed increased perfusion during auditory or visual hallucinations, respectively. Additionally, increased perfusion of the orbitofrontal cortex and the striatum was seen during hallucinations in these patients with schizophrenia (Silbersweig et al., 1995). It is unclear however, which cortical region initiates activation increases within the visual perceptual network of temporal, frontal and perhaps parietal cortical activation during VH. An intra-operative stimulation study in epilepsy patients showed that stimulation of the prefrontal cortex (inferior frontal gyrus) can evoke complex VH, probably by propagation of activity from the prefrontal

70 4.6. CONCLUSION 61 cortex along white matter tracts [uncinate fasciculus (Catani and Mesulam, 2008)] to the ventral occipito-temporal lobe (Blanke et al., 2000). Furthermore orbitofrontal seizures can present with complex VH, probably also by propagation of epileptic activity to temporal regions (La Vega-Talbot et al., 2006). 4.6 Conclusion Increased vulnerability for VH in PD is associated with impaired visual object processing in ventral/lateral visual association cortices, providing support for our hypothesis of impaired bottom-up visual processing in PDwithVH. Moreover, reduced activation in a wider network including lateral prefrontal in PDwithVH suggested that early stages at which top-down information is given are additionally impaired. We did not find arguments for compensatory increases of activation in PDwithVH, and thus no support for a link between vulnerability for VH and increased reliance on top-down processing during visual perception. 4.7 Acknowledgements We thank Richard Jacobs, MSc, for his help with the design of the presentation program. We thank Jan-Bernard Marsman, MSc, for his help with ROI analysis and group plots and Dave Langers, PhD, for his help with FIR analysis.

71

72 Chapter 5 Regional cortical grey matter loss in Parkinson s disease without dementia is independent from visual hallucinations A. M. Meppelink 1,2,3, B. M. de Jong 1,2,3, L. K. Teune 1,2,3, T. van Laar 1,2,3 (1) Department of Neurology, University Medical Center Groningen, The Netherlands (2) School of Behavioral and Cognitive Neurosciences, University of Groningen, the Netherlands (3) Neuro Imaging Center (NIC) Groningen, University of Groningen, the Netherlands Accepted for publication in Movement Disorders 63

73 64 CHAPTER 5. CORTICAL GREY MATTER CHANGES IN PD 5.1 Abstract In ourpreviousfmri study, Parkinson s disease (PD) patients with visual hallucinations (VH) showed reduced activations in ventral/lateral visual association cortices preceding image recognition, compared to both PD patients without VH and healthy controls. The primary aim of the current study was to investigate whether functional deficits are associated with grey matter volume changes. In addition, possible grey matter differences between all PD patients and healthy controls were assessed. Using 3 Tesla MRI and voxel-based morphometry (VBM), we found no differences between PD patients with (n=11) and without VH (n=13). However, grey matter decreases of the bilateral prefrontal and parietal cortex, left anterior superior temporal and left middle occipital gyrus were found in the total group of PD patients, compared to controls (n=14). This indicates that previously demonstrated functional deficits in PD patients with VH are not associated with grey matter loss. The strong left parietal reduction in both non-demented patient groups was hemisphere-specific and independent of the side of PD symptoms.

74 5.2. INTRODUCTION Introduction Parkinson s disease (PD) primarily affects the substantia nigra including its striatum projections (Lang and Lozano, 1998). Classically, cortical pathology has received little attention in PD (Selby, 1968). More recently, structural abnormalities have been described, although inconsistent and mainly focussed on cognitive impairment (Schneider et al., 1979; Dagher and Nagano-Saito, 2007). Functional imaging in non-demented PD patients, however, has revealed more consistent cortical impairment, both in motor and visual domains (Dagher and Nagano-Saito, 2007; Ma et al., 2007). Using fmri, we recently specified a relation between visual cortex function and VH in non-demented PD patients by demonstrating reduced extrastriate visual activations preceding image recognition (Meppelink et al., 2009). Now we aimed to gain further insight in possible occipito-temporal pathology associated with such VH using 3 Tesla MRI and voxel-based morphometry (VBM). VBM allows determination of cortex density and/or volume changes without a regional bias. It has been used before by one other group to address this topic (Ramirez-Ruiz et al., 2007b; Ibarretxe-Bilbao et al., 2009), but using 1.5 Tesla MRI. We also compared all PD patients with healthy controls to see whether regional cortex atrophy would match previously described reduced regional metabolism in PD. 5.3 Methods Subjects The thirty-eight included subjects were previously studied with fmri (Meppelink et al., 2009) and divided in three groups: PD patients with VH, experienced at least weekly during the last month (n=11), patients without VH (n=13) and healthy controls (n=14). Patients met the criteria of the UK PD Society Brain Bank. Cognition was assessed with the Mini Mental State Examination (MMSE) (Folstein et al., 1975) and the SCOPA-cog (SCales of Outcomes in PArkinson s disease, cognition)(marinus et al., 2003). Severity of motor symptoms was rated with the Unified Parkinson s Disease Rating Scale (UPDRS), part III. Severity of VH and executive functioning were assessed with the Neuropsychiatric Inventory (B: Hallucinations ) and the Frontal As-

75 66 CHAPTER 5. CORTICAL GREY MATTER CHANGES IN PD sessment Battery (FAB) (Dubois et al., 2000), respectively. Exclusion criteria were dementia (MMSE <24), neurological disorders other than PD, psychiatric disorders, visual acuity below 50 percent and visual field defects. The local Medical Ethical Committee approved the study. Participants signed an informed consent Voxel-based Morphometry MRI was performed with a 3 Tesla scanner (Philips, Best, NL) using a standard 6 channel SENSE head coil. T1 weighted 3D anatomical images were defined by isotropic voxels 1 x 1 x 1 mm, matrix 256 x 256 and axial orientation. Image processing and statistical analysis were conducted with Statistical Parametric Mapping (SPM, Friston et al. (1995), version 5 (Wellcome Dept. Cogn. Neurol. London, UK; Images were spatially normalised (T1 template Montreal Neurological Institute, MNI) and segmented into grey matter, white matter and cerebrospinal fluid. Grey matter images were modulated and smoothed (10 FWHM). We used modulated grey matter images, because modulation takes into account the deformation field generated during spatial normalization. In this way, the total amount of grey matter remains the same as it would be in the original images and grey matter volume changes rather than concentration differences can be assessed Statistical analyses MMSE, FAB and UPDRS-III scores for PD patients were not normally distributed and therefore compared using the Mann-Whitney test. Education levels, total SCOPA-cog and SCOPA-cog subscores were compared with the Kruskal-Wallis test. Normally distributed age differences were tested with ANOVA. Grey matter volume changes were assessed with ANOVA (flexible factorial, main effect factor group ). Total grey matter was calculated per subject and used as covariate to remove variance due to differences in head size. We compared PD with and without VH with each other and to healthy controls, and all PD versus healthy controls. Initial threshold was p<0.001, voxel-level, uncorrected. Clusters were considered statistically significant at brain-volume corrected cluster-level p<0.05. In addition, ROI s were defined

76 5.4. RESULTS 67 with Marsbar in SPM5, based on previously reported activation decreases in PD with VH in the left fusiform gyrus (LFG). 5.4 Results No differences existed between the three groups regarding age (F=0.62, p=0.54), gender (χ 2 =2.55, p=0.28) and education level (χ 2 =0.35, p=0.84). Mean (SD) PD disease duration was 8.0 (4.7) in PD with VH and 7.9 (2.4) in PD without VH. PD groups were similar regarding MMSE scores (z=-1.45, p=0.15). Total SCOPA-cog scores differed between groups (χ 2 =9.0, p=0.01), with verbal memory being the only significant subscore (χ 2 =8.37, p=0.02; attention: χ 2 =0.81, p=0.67; executive functioning: χ 2 =2.63, p=0.27; visuospatial: χ 2 =0.18, p=0.18). Mann-Whitney test revealed significant differences on verbal memory between PD with VH versus controls (z=-2.66, p=0.008). Differences were not significant between PD without VH versus either PD with VH (z=-1.66, p=0.10) or controls (z=-1.67, p=0.09). FAB-scores were lower in PD patients with VH, compared to patients without VH (z=-2.29, p=0.02). UPDRS-III scores did not differ (z=-0.70, p=0.48). Voxel-based comparison between grey matter images of PD patients with VH and without VH did not show any differences between the two groups, ROI analysis of the LFG showed no differences either (data not shown). In comparison with healthy controls, however, each of the two PD patient groups, i.e. with or without VH, showed grey matter decreases in prefrontal, parietal and temporal cortices (Fig 5.1B,C). The combined PD group (=PDtotal) showed grey matter reductions in prefrontal and parietal cortices (bilaterally), the left temporal lobe, left middle occipital gyrus and right (pre-) Supplementary Motor Area (SMA) (Fig.5.1A). Table 5.1 reports significant regions of grey matter decrease (p<0.05, clusterlevel, brain-volume corrected). Parietal grey matter reductions were most apparent in the left hemisphere (Fig.5.1, Table 5.1). In order to explore a possible relation with contralateral symptom dominance, a PD symptom lateralization index was calculated defined by negative values for left-sided dominance, positive values for right dominance and zero for absent lateralization. Adding this index as a covariate in the analysis had no effect on the results, indicating that the observed lateralization was hemisphere-specific, independent from the side of symptom dominance. Because PD patients with and without VH differed on the FAB-

77 68 CHAPTER 5. CORTICAL GREY MATTER CHANGES IN PD Figure 5.1: Regional cortical grey matter changes in PD patients. Grey matter reductions in PDtotal, compared to healthy controls (A), PD without VH, compared to healthy controls (B) and PD with VH, compared to healthy controls (C) at a threshold of p<0.001 (uncorrected), k=20. Regional grey matter reductions are rendered on a standard MNI brain. R = right, P = posterior. scores, these were also added as a covariate, without effect on SPM results either. 5.5 Discussion Equal grey matter in PD with and without VH Reduced cortical grey matter volume was a general PD characteristic, without differences between patients with or without VH. This indicates that the functional differences we previously found between these two patient groups, i.e. reduced activation of ventral/lateral extrastriate visual cortices in PD with VH

78 5.5. DISCUSSION 69 Pcorrected Cluster (k) MNI coordinates T Location PDtotal - HC ,-60, Parietal superior -48, -62, Angular gyrus L -4, -54, Precuneus L -40, -56, Parietal inferior L -30, -74, Occipital middle L , 18, Temporal superior pole L -50, 16, Frontal inferior L , -62, Angular gyrus R 18, -50, Parietal superior R 16, -70, Precuneus R , 32, Frontal orbicularis sup R 6, 36, Frontal orbicularis inf R , 6, (Pre-) SMA , -36, Fusiform gyrus L -46, -46, Temporal inferior L , 28, Frontal inferior R PDnonVH - HC , -74, Occipital middle L -20, -60, Parietal superior L -48, -62, Angular gyrus L -8, -66, Precuneus L -40, -56, Parietal inferior L , 16, Temporal superior pole L -54, 4, Temporal inferior L -48, 30, Front orbicularis inf L -38, -16, Temporal inferior L , 32, Frontal orbicularis sup R 20, 42, Frontal orbicularis mid R , -60, Angular gyrus R 24, -60, Parietal superior R PDwithVH-HC , -66, Precuneus L -46, -68, Temporal/occipital mid L -18, -60, Parietal superior L -48, -64, Angular gyrus L -40, -56, Parietal inferior L , -50, Parietal superior R 16, -72, Precuneus R 28, -72, Occipital superior R , 18, Temporal superior pole L -60, -22, Temporal middle L -50, 16, Frontal inferior L , 56, Frontal superior R 32, 56, Frontal middle R , 4, Frontal superior R 8, 4, (Pre-)SMA R 8, 30, Frontal sup med R , 28, Frontal inferior R Table 5.1: Demographic and illness characteristics of PD patients with visual hallucinations (PD + VH; n=14), PD patients without VH (PD - VH; n=14) and healthy controls (HC; n=14)

79 70 CHAPTER 5. CORTICAL GREY MATTER CHANGES IN PD (Meppelink et al., 2009), were not associated with local cortical atrophy. This seeming discrepancy with Ramirez-Ruiz and colleagues, who associated VH in PD with grey matter reductions in the lingual gyrus and superior parietal lobe (Ramirez-Ruiz et al., 2007b), is likely explained by the advanced disease stage in their study. Although they included Hoehn and Yahr stage as a covariate in their analysis, this only corrects for differences between their two PD groups and not between their patients and ours. VH-related functional impairment without anatomy changes in our patients suggests specific neurochemical deficits preceding structural changes. Cholinergic deficit might be considered in this respect, possibly causing impaired selection of subcortical information streams, subsequently predisposing to hallucinations (Perry and Perry, 1995). Higher density of Lewy bodies in the temporal lobe might also play a role (Harding et al., 2002). In non-demented PD patients, VH have been associated with cognitive impairment (Fenelon et al., 2000; Williams and Lees, 2005; Imamura et al., 2007; Ramirez-Ruiz et al., 2007a; Meppelink et al., 2008) and may predict dementia (Aarsland et al., 2003; Santangelo et al., 2007). In pathologically proven PD, VH were an initial milestone of advanced disease independent from disease duration (Kempster et al., 2007). The association between VH and cognitive decline in PD is consistent with enhanced brain atrophy in PD patients with VH, particularly when dementia follows (Ibarretxe-Bilbao et al., 2009). In this respect, our PD patients with VH might show cognitive impairment and atrophy in follow-up assessments Grey matter reductions in PD Although we saw no grey matter differences between the two non-demented PD patient groups, PDtotal showed grey matter reductions in specific parietal, temporal, occipital and frontal regions, compared to healthy controls. These reductions were more extensive than previously described, possibly explained by higher sensitivity of 3 Tesla imaging. Frontal and temporal cortex atrophy in non-demented PD patients has been described before with 1.5 Tesla MRI (Burton et al., 2004; Ramirez-Ruiz et al., 2007b; Pereira et al., 2009; Summerfield et al., 2005; Beyer et al., 2007; Tir et al., 2009), but not consistently (Nagano-Saito et al., 2005; Feldmann et al., 2008) and depending on cognitive impairment (Nagano-Saito et al., 2005) or depression (Feldmann et al., 2008). Medial frontal atrophy in our study particularly concerned the rostral (or pre-) SMA, which is consistent with functional cortical impairment in PD

80 5.5. DISCUSSION 71 due to loss of basal ganglia-thalamus output (Playford et al., 1992; Cunnington et al., 2001). Associated frontal and parietal grey matter reductions may further reflect impaired neuronal circuitry implicated in both motor- and cognitive functions (de Jong et al., 1996; Ma et al., 2007; Huang et al., 2007). To explain cortical volume reduction, a first consideration is disease-inflicted cell loss. This might e.g. be a consequence of α-synuclein pathology (Jellinger, 2009b), although subsequent cortical Lewy body deposition in non-demented PD remains an issue of debate (Jellinger, 2009a; Braak et al., 2004, 2005; Colosimo et al., 2003; Parkkinen et al., 2008). Tissue pathology, however, does not explain the left-sided predominance of parietal atrophy we found because it was not contralateral to the side of dominant symptoms. Volume reduction might alternatively be a dynamic consequence of reduced neuronal activity, leading to reduced dendritic spine volume or astroglial volume reduction (Draganski et al., 2004). The opposite effect, i.e. action-induced volume increase, has been demonstrated (Draganski and May, 2008). Bilateral parietal atrophy in PD has been described before, also with left-sided dominance (Pereira et al., 2009). To provide a functional explanation for leftsided parietal atrophy, possible associations between parietal motor functions (Binkofski et al., 1999) and PD symptoms need to be considered. Left parietal processing of body scheme- or self-referenced (motor) information subserves prehension (Binkofski et al., 1999; de Jong et al., 2001) and contributes to the initiation of new motor programs (de Jong et al., 1999, 2001), while deficit may result in ideomotor apraxia (Wheaton and Hallett, 2007). Although apraxia is not a key symptom of PD, reduced internally-driven performance would fit the hypothesis of de-learning skilled movements in PD. Atrophy would thus be secondary to reduced purposeful action, in which a general intentional drive is impaired due to basal ganglia disease. Such dynamic volume change has been demonstrated by the opposite effect: grey matter of visual motion area MT/V5 and the left parietal cortex thickens after learning a new skill such as juggling (Draganski et al., 2004); left frontoparietal cortex volume enlarges in skilled golfers (Jancke et al., 2009). Finally, although PD patients, especially those with VH, scored lower at the verbal memory subtask of the SCOPA-cog, we regard it unlikely that the left lateralized parietal atrophy reflects impairment of language-related function.

81 72 CHAPTER 5. CORTICAL GREY MATTER CHANGES IN PD 5.6 Conclusions Compared to healthy controls, gray matter was equally reduced in PD patients with and without VH, indicating that previously found VH-related functional deficits in these patients were not associated with detectable anatomical changes. Hemisphere-specific left parietal atrophy in PDtotal might reflect a secondary effect of basal ganglia disease, leading to impaired recruitment of internally guided motor programs. 5.7 Acknowledgements We would like to thank professor K.L. Leenders for critically reading previous versions of this manuscript.

82 Chapter 6 Cholinergic modulation of visual brain responses in Parkinson s disease with visual hallucinations and healthy controls; preliminary results A. M. Meppelink 1,2,3, B. M. de Jong 1,2,3,R.Renken 2, K. L. Leenders 1,2,3,T. van Laar 1,2,3 (1) Department of Neurology, University Medical Center Groningen, The Netherlands (2) Neuro Imaging Center (NIC) Groningen, University of Groningen, the Netherlands (3) School of Behavioral and Cognitive Neurosciences, University of Groningen, the Netherlands 73

83 74 CHAPTER 6. CHOLINERGIC MODULATION DURING FMRI 6.1 Abstract Previously described impairments of visual processing in Parkinson s disease (PD) patients with visual hallucinations (VH) were proposed to result from reduced cholinergic innervation. Acetylcholine plays a role in attention, which includes selection, while it was previously shown to enhance visual cortex activation in Alzheimer s disease. We hypothesized that activation in both frontal attention areas and visual association cortices might be enhanced by increasing cerebral levels of acetylcholine. Five non-demented PD patients with VH and 10 healthy elderly subjects were scanned on a 3Tesla MRI scanner. Static images of animals and objects gradually appearing out of random visual noise were used in an event-related design paradigm. Analyses were time-locked on the moment of image recognition, indicated by the subjects button-press. Subjects were scanned twice; once after administration of rivastigmine and once after placebo (both patch), in a balanced, pseudo-randomized manner (double-blind). Data pre-processing and statistical analysis were performed with SPM8. Healthy controls showed robust bilateral fusiform- and lingual gyri activation in both treatments (p<0.001), while rivastigmine (compared to placebo) resulted in activation increases of the anterior superior frontal gyrus, anterior cingulate, left insula and subthreshold putamen and pallidum. PD patients with VH showed less robust activation than controls, without significant differences between treatment conditions. However, non-significant activation increases after rivastigmine were seen in striate and extrastriate visual cortices, bilateral (para)hippocampus, posterior superior frontal gyrus and basal ganglia (pallidum and putamen). While healthy controls showed increased frontal and basal ganglia activation after cholinergic enhancement, PD patients with VH showed additional (nonsignificant) increased visual cortex activation. The latter may reflect normalization of impaired visual cortex activation in PD with VH by rivastigmine. Extension of these preliminary data is necessary to confirm this effect.

84 6.2. INTRODUCTION Introduction PParkinson s disease (PD) is a multisystem neurodegenerative disorder, in which deterioration of dopaminergic neurons in the substantia nigra, that project to the striatum, is a classical hallmark (Lang and Lozano 1998). Motor symptoms like bradykinesia, rigidity and tremor are dominant characteristics of PD, while non-motor symptoms such as cognitive impairment and visual hallucinations (VH) may additionally occur (Aarsland et al. 2005;Barnes and David 2001). The exact etiology of VH in PD is unknown, although a combination of impaired visual processing and attention seems to be involved (Diederich et al. 2005;Collerton et al. 2005;Meppelink et al. 2008;Koerts et al. 2010). Our previous functional MRI (fmri) study showed relatively impaired visual processing in non-demented PD patients with VH, compared to both non-demented PD patients without VH and healthy controls (Meppelink et al. 2009). These functional impairments were not associated with structural differences in grey matter volume (Meppelink et al. 2010). VH-related functional impairment without anatomy changes in these patients suggests specific neurochemical deficits preceding structural changes. Cholinergic deficit might be considered in this respect, possibly causing impaired selection of subcortical information streams, subsequently predisposing to hallucinations (Perry and Perry 1995). Post-mortem studies in PD have shown loss of cortical cholinergic neurons and associated degeneration of the Nucleus basalis of Meynert (NbM) (Mesulam 2004), in the basal forebrain (Perry et al. 1985). Recent in vivo cholinergic tracer studies have shown cholinergic denervation of the occipital and parietal cortex in non-demented PD patients, while others have shown an association of cholinergic denervation with impaired attention (Hilker et al. 2005;Shimada et al. 2009;Bohnen et al. 2006). Clinical evidence shows that VH can be induced by anti-cholinergics, while cholinesterase inhibitors ameliorate cognitive dysfunction and VH in PD (Burn et al. 2006;Wesnes et al. 2005). The aim of the current study was to investigate the effect of the cholinesterase inhibitor rivastigmine on cerebral activation during visual processing in PD patients with VH and healthy controls. To this end we assessed cerebral activations in these groups during presentation of images that were gradually revealed out of noise (Meppelink et al. 2008;Meppelink et al. 2009), in a double-blind, placebo controlled setup. We hypothesized that cholinergic enhancement would increase activation in prefrontal and parietal cortex in healthy controls (Bentley et al. 2008). In addition, activation enhancement of visual association cortices was hypothesized to occur in PD patients with VH.

85 76 CHAPTER 6. CHOLINERGIC MODULATION DURING FMRI 6.3 Methods Subjects Until now, fifteen subjects were eventually analyzed in this study, divided in 2 groups; 5 PD patients who experienced complex VH at least weekly during the last month and 10 healthy controls. The ultimate aim is to include another 5 PD patients (2x10 subjects design). PD was diagnosed according to the criteria of the UK Parkinson s Disease Society Brain Bank. These two groups were matched for age and level of education, which was rated with a Dutch education scale ranging from 1 (elementary school not finished) to 7 (university degree). All PD patients were on during the assessments. The levodopa-equivalent daily dose (LEDD) was calculated for all patients, according to the formula: LEDD=levodopa dose (mg) + (0.3 *levodopa dose if using entacapone with each dose) + (slow release levodopa *0.7) + (bromocriptine *10) + (ropinirole *20) + (pergolide *100) + (pramipexole *100) + (apomorphine *10) (Esselink et al. 2004). Visual acuity was assessed with the Snellen chart. Demographic and clinical characteristics are described in table 6.1. Exclusion criteria were dementia [Mini Mental State Examination (MMSE) score < 25], use of cholinesterase inhibitors or anticholinergics, neurological disorders other than PD, psychiatric disorders, visual acuity less than 50 percent (Snellen chart) and visual field defects. This study was approved by the Medical Ethical Committee of the University Medical Center Groningen. All participants signed an informed consent prior to study inclusion Clinical tests and statistics The severity of motor symptoms in PD patients was rated with the Unified Parkinson s Disease Rating Scale (UPDRS), part III. Severity of VH in PD patients was assessed with the University of Miami Parkinson s disease Hallucination Questionnaire (UM-PDHQ), a scale based on the Neuropsychiatric Inventory and phenomenology of VH in PD (Papapetropoulos et al. 2008). Cognition and executive functioning was assessed with the MMSE (Folstein et al. 1975), the SCOPA-cog (SCales of Outcomes in PArkinson s disease cognition) (Marinus et al. 2003) and the Frontal Assessment Battery (FAB) (Dubois et al. 2000). Severity of motor symptoms was rated with the Uni-

86 6.3. METHODS 77 HC PD + VH Mean (SD) range Mean (SD) range Age (years) Education Males: n (%) 7 (64 %) 4(80%) Females: n (%) 4 (36 %) 1(20%) Visual acuity MMSE FAB SCOPA-cog 1st session nd session placebo rivastigmine Table 6.1: Demographic and clinical characteristics of healthy controls (HC; n=10) and PD patients with VH (PD+VH, n=5). fied Parkinson s Disease Rating Scale (UPDRS), part III. Adverse events were documented. Not all variables were normally distributed. The non-parametric Mann-Whitney test was therefore used to investigate these non-normally distributed variables (visual acuity, level of education and age), while normally distributed mean image recognition times were compared using an ANOVA with repeated measures and post-hoc with paired t-tests Study design Assessments in each subject were done in two sessions, one after rivastigmine and one after placebo, double-blind, pseudo-randomized and balanced, with one week between sessions. A 10 cm 2 patch (rivastigmine or placebo) was applied the evening before the assessment, approximately 10 hours before the fmri scan, which was always performed at 9.30 a.m fmri paradigm and experimental procedure A total of 80 pictures of animals (40), well-known objects (40) were used to create a paradigm in which pictures gradually pop out of random uniform visual white noise (see figure 4.1A). Movie stimuli were generated in Matlab 5 on an Apple Macintosh computer (running Mac OS 9.2.1) using some of the rou-

87 78 CHAPTER 6. CHOLINERGIC MODULATION DURING FMRI tines of the Psychtoolbox (Brainard 1997;Pelli 1997). Noise contrast remained constant throughout the duration of the 30-second movie. Image contrast (and thus signal-to-noise) increased linearly over time causing the image to gradually appear out of the noise. Perceptual recognition ( pop-out ) occurred from 10 to 28 seconds after initial movie onset. All movies were created from grey-scale pictures with a resolution of 300 by 300 pixels. Movies were displayed at twice this size (600x600 pixels). The movies were presented using the Presentation program (Neuro Behavioral Systems, Inc. CA, USA). They were projected by a beamer (resolution 1024 x 768 pixels, Barco, Belgium) on a screen (display dimensions 44 x 34 cm), viewed by the subject via a mirror placed at a distance of 11 cm from the face. The distance between the mirror and the screen was 64 cm and the stimuli covered approximately 18 degrees of the visual field. If necessary, visual acuity of the subject was corrected using MRI-compatible lenses. During presentation of the movie, a central fixation square changed color with random intervals. Subjects had to report such change (thus urging to keep attention constant) by pressing a button with their right middle finger on an MR compatible response-box (forp, Current designs, Inc. U.S.A.). Per subject, the mean reaction time of the response to the color change was calculated. Subjects were further instructed to press a button with their right index finger at the moment that they recognized the object or animal. The mean reaction time on the color change was subtracted from the image recognition times. Movie stimuli were presented in two runs per session (total 2 sessions per subject), 20 per run. In between the two runs, an anatomical, T1 weighed scan was acquired MRI characteristics Data acquisition was performed using a 3 T Philips MR system (Best, The Netherlands) with a standard 8 channel SENSE head coil. Functional images were acquired with a gradient echo, i.e. echo planar imaging, T2* Blood Oxygen Dependent Level (BOLD) contrast technique in an ascending order with a TE of 35 ms, a TR of 2.3 seconds, 35 slices per TR, 450 volumes per run, isotropic voxels 3 x 3 x 3 mm and an axial orientation. A T1 weighted 3D anatomical scan was acquired to obtain high resolution anatomical information, isotropic voxels 1 x 1 x 1 mm, matrix size = 256 x 256 and an axial orientation.

88 6.4. RESULTS Psychophysics The mean image recognition time over all movies was calculated for each subject per condition and averaged over groups. The percentage of unrecognized images was similarly calculated per subject and averaged over groups as well. Unrecognized movies were considered as missing values. Mean image recognition times were normally distributed, differences between groups were investigated using an ANOVA with repeated measures (main factors: condition and group, interaction: condition x group) fmri data analysis Image processing and statistical analysis were conducted with Statistical Parametric Mapping (SPM, (Friston et al. 1995) version 8 (2008, Wellcome Department of Cognitive Neurology, London, UK; Pre-processing included slice time correction, realignment, coregistration of functional and anatomical scans and spatial normalization (to the template of the Montreal Neurological Institute, MNI). Images were smoothed using a Gaussian filter of 8 mm FWHM. Analyses were time-locked on the perceptual pop-out in an event-related design. In addition to the hemodynamic response function (HRF), temporal and dispersion derivatives were modeled as well (Friston et al. 1998). Apart from the pop-out, a 30 seconds block of visual input and motor response as well as a block of the visual percept, lasting from the pop-out until the end of the movie, were modeled (figure 4.1B). Movement parameters were included as covariates. T-contrasts of the pop-out with respect to baseline (passive viewing of a fixation cross, projected on a dark background) were made for each subject. On a 2nd level (random effect analysis), the T-contrasts of the pop-out were analyzed per group and condition, using an ANOVA (flexible factorial). 6.4 Results Characteristics subjects The two groups were matched for age (Mann-Whitney: z=-1.54, p=0.12) and level of education (Mann-Whitney: z=-0.97, p=0.33). Visual acuity was equal in both groups (Mann-Whitney: z=-0.70, p=0.49). Adverse events included

89 80 CHAPTER 6. CHOLINERGIC MODULATION DURING FMRI Figure 6.1: Scatter plots of mean image recognition in healthy controls (HC) and PD patients with VH (PD) after placebo (HC plac and PD plac) and after rivastigmine (HC riva and PD riva). dry mouth (n=1, healthy control) and slight feeling of hangover (n=1, healthy control). One patient who experienced nausea with vomiting did not complete the study and was therefore excluded. Two out of the 5 PD patients with VH reported having VH several times a week while 3 reported having VH several times a day. One patient reported that she became upset during VH and considered these experiences as a severe emotional burden. Two patients experienced VH of people, animals and objects, two patients experienced VH of people and animals and one patient of only people. None of the PD patients with VH experienced VH during the testing. Neither did they report a relationship between their VH and the intake of dopaminergic medication. One patient used clozapine treatment for her VH, while none of the subjects used cholinesterase-inhibitors or anti-cholinergics. Four of five patients also experienced visual illusions Task performance All subjects were able to perform the task adequately. The mean percentages of recognized images in healthy controls was 97.5 percent in the placebo

90 6.5. DISCUSSION 81 condition (HC plac) and 98 percent in the rivastigmine condition (HC riva), whereas the mean percentage of recognized images in PD patients with VH was 95 percent in the placebo condition (PD plac) and 94 percent in the rivastigmine condition (PD riva). Mean image recognition time (standard deviation) over all movies was sec in healthy controls, (2.2) sec in HC plac, (2.1) sec in HC riva, sec in PD patients with VH, (1.5) sec in PD plac and (1.8) in PD riva. ANOVA with repeated measures showed a significant effect of group (F=70.58, p<0.001), but no significant interaction between condition and group. Plotted results however, gave the impression that differences between groups after rivastigmine were smaller (figure 6.1), which was confirmed with two paired samples t-tests (HC plac versus PD plac: t=-2.27, p=0.04; HC riva versus PD riva: t=-1.25, p=0.24) fmri results Healthy controls showed robust activations in bilateral fusiform- and lingual gyri in both the rivastigmine and the placebo condition (p<0.001, see figure 6.2A). Rivastigmine (compared to placebo) resulted in an increased activation of the anterior superior frontal gyrus, anterior cingulate, left insula and subthreshold level in left pallidum and right putamen (figure 6.3A and table 6.2). PD patients with VH showed less robust activation than controls in each of the two conditions (see figure 6.2B). Rivastigmine compared to placebo showed no significant differences in this group. However, non-significant activation increases after rivastigmine were seen in striate and extrastriate visual cortices, bilateral (para)hippocampus, posterior superior frontal gyrus and basal ganglia (pallidum and putamen, see figure 6.3B). Plotting local activation for the two conditions in each PD patient at the right fusiform gyrus (MNI coordinates: 26,-46,-14) shows a clear distinction between the two conditions in all 5 patients (figure 6.4). 6.5 Discussion This study was designed to identify distribution of brain activation during visual processing in PD patients with VH and healthy controls after administration of the cholinesterase inhibitor rivastigmine. Although the patient data is still preliminary, the placebo-controlled, double blind, pseudo-randomized and balanced design of the study creates the possibility to draw some prelim-

91 82 CHAPTER 6. CHOLINERGIC MODULATION DURING FMRI Figure 6.2: Activation of the fusiform gyrus (1) and lingual gyrus (2) during image recognition in (A) healthy controls after placebo, (B) healthy controls after rivastigmine and (C) PD patients with VH after rivastigmine. Activations are projected on transverse sections of a standard brain (Montreal Neurological Institute) The sections traverse the anterior- and posterior commissures (AC-PC plane) and 10 mm inferior to it (z= -10 mm). R = right.

92 6.5. DISCUSSION 83 A. Activation at the moment of pop-out after administration of placebo Contrast: Anatomical region MNI-coordinates T (voxel level) p-value (cluster level) HC fusiform R 27, -54, <0.001, corrected fusiform L -45, -65, <0.001, corrected lingual R 16, -47, <0.001, corrected lingual L -12, -70, <0.001, corrected B. Activation at the moment of pop-out after administration of rivastigmine Contrast: Anatomical region MNI-coordinates T (voxel level) p-value (cluster level) HC fusiform R 26,-38, <0.001, corrected fusiform L -44,-58, <0.001, corrected lingual R 16,-46, <0.001, corrected lingual L -16,-60, <0.001, corrected PD+VH fusiform R 26,-46, <0.001, corrected lingual R 8,-70, <0.001, corrected lingual L -16,-46, <0.001, corrected C. Effects of rivastigmine compared to placebo in healthy controls Contrast: Anatomical region MNI-coordinates T (voxel level) p-value (cluster level) Riva > placebo Frontal superior 26, 48, <0.05, corrected Cingulate anterior -4, 46, <0.05, corrected Insula L -38,-22, <0.05, uncorrected Putamen R 30,-16, , uncorrected Pallidum L -22,-4, , uncorrected Table 6.2: Regions of cerebral activations in healthy controls (HC, n=10) and PD patients with VH (PD+VH, n=5) at the moment of image recognition after administration of placebo (A) and rivastigmine (B). Initial threshold in SPM was 0.001, uncorrected (k=9) for HC and 0.01, uncorrected (k=9) for PD+VH. C. Activation increases after rivastigmine (compared to placebo) in healthy controls (initial threshold in SPM 0.01, uncorrected, k=9). Figure 6.3: Cerebral activation increases after rivastigmine, compared to placebo A Increased frontal and basal ganglia activation in healthy controls (p<0.01, k=9) B Increased (para)hippocampal and visual cortex activation in PD with VH (p<0.05, k=9)

93 84 CHAPTER 6. CHOLINERGIC MODULATION DURING FMRI Figure 6.4: Fitted responses of PD patients with VH in the fusiform gyrus (MNI: 26,-46,-14) after rivastigmine (left side) or placebo (right side). inary conclusions. The moment at which the images of animals and objects were recognized after gradually appearing out of random noise, was related to marked activation in appropriate secondary visual regions such as the fusiform and lingual gyri in healthy controls. These activation patterns were similar to our previous study (Meppelink et al. 2009), again providing support for the robustness of our paradigm. The small group of PD patients with VH showed less robust activation of visual cortex during image recognition than healthy controls, but robustness of this effect will likely increase when more subjects are included (Meppelink et al. 2009). Like our previous study, the mean time it took to recognize the images was significantly longer in PD patients with VH, compared to healthy controls. After rivastigmine these group differences were no longer significant, mainly caused by a decrease in recognition time in PD patients after rivastigmine, compared to placebo. In healthy controls, rivastigmine enhanced cerebral activation in especially prefrontal (including anterior cingulate) cortex. Although this increased activation might reflect attention-related effects, a firm conclusion cannot be drawn because these effects accompanied by behavioral effects of rivastigmine

94 6.5. DISCUSSION 85 in controls. One previous study in healthy subjects showed cholinesterase inhibitor-induced increases of the left prefrontal cortex in a relatively simple visual task without behavioral improvement either (Bentley et al. 2008). Differences between conditions in PD patients with VH were not significant, but nevertheless potentially interesting. They provide an indication that in the studied PD patients with VH rivastigmine improves extrastriate visual processing. The finding that enhanced responses in the right fusiform gyrus in patients on rivastigmine did not overlap with those of patients on placebo, is promising in this regard. This increased activation of visual cortex and (para)hippocampal areas after rivastigmine might reflect compensation of impaired bottom-up processing (Meppelink et al. 2009), thus providing support for the hypothesis that acetylcholine favors bottom-up over top-down sensory processing (Bentley et al. 2004). Extension of these preliminary data is necessary to confirm this enhancing effect on visual cortex in PD patients with VH.

95

96 Chapter 7 The effects of apomorphine on visual perception in patients with Parkinson s disease and visual hallucinations; a pilot study L. Geerligs 1,2, A. M. Meppelink 1,2, W. H. Brouwer 1, T. van Laar 1,2 (1) Department of Neurology, University Medical Center Groningen, The Netherlands (2) School of Behavioral and Cognitive Neurosciences, University of Groningen, the Netherlands Accepted for publication in Clinical Neuropharmocology 87

97 88 CHAPTER 7. APOMORPHINE IN PD WITH VH 7.1 Abstract Visual hallucinations (VH) often occur in patients with advanced Parkinson s disease (PD). Overstimulation of dopamine receptors has been considered as one of the causes for VH in PD. However, several clinical studies suggested that apomorphine infusion did not worsen existing VH in PD, but could even improve VH in some PD patients. This pilot study included 4 PD patients with VH, who were examined before, during and after an intravenous infusion with apomorphine. The examinations included tests for lower and higher order visual functions, attention and motor functions. Apomorphine had a significantly positive effect on contrast sensitivity and showed a significantly negative effect on attention. These results may explain why apomorphine is able to improve VH in PD in some patients with mainly visual perceptive problems, but may also worsen VH in other patients, due to impaired attention.

98 7.2. INTRODUCTION Introduction Visual hallucinations (VH) often occur in patients with advanced PD and are associated with problems on behavioural and functional level and higher mortality (Holroyd et al., 2001). Although overstimulation of dopamine receptors has frequently been suggested as the cause for VH in PD (Diederich et al., 2005), there is contrasting evidence. Several clinical studies showed that subcutaneous infusion of apomorphine did not worsen existing VH (Morgante et al., 2004; Di Rosa et al., 2003), whereas another study even suggested that apomorphine could improve VH in PD (Ellis et al., 1997). A recent clinical study from our center, using subcutaneous infusion of apomorphine in PD patients with VH, showed that equipotent conversion from oral dopamine agonists to subcutaneous apomorphine infusion was able to improve the frequency and severity of VH in PD (submitted). Other clinical studies in which apomorphine was administered in PD patients, showed an improvement of contrast sensitivity (Buttner et al., 2000), which may be related to the improvement of VH. This study aims to confirm this finding of the effect of apomorphine on contrast sensitivity and to broaden the scope by investigating the effect of apomorphine on various visual, cognitive and motor functions in PD patients with VH. 7.3 Patients and Methods Design All patients were examined three times during this study. The first time at baseline, while they were using their standard medication, the second time during steady state intravenous apomorphine infusion and the third time 2 hours after having stopped the infusion. All patients were on dopaminergic medication during testing. The following tests were used to analyse possible changes in vision, cognition, motor- and executive functions at the pre-infusion, infusion and post-infusion examinations. Motor function: Unified Parkinson s Disease Rating Scale (UPDRS) part III Lower order visual functions: (1) Contrast sensitivity measured in log units (Mars Letter Contrast Sensitivity Test); (2) Visual acuity (Snellen chart); (3)

99 90 CHAPTER 7. APOMORPHINE IN PD WITH VH Colour Discrimination (Farnsworth D-15; we used the Total Colour Difference Score (TCDS) in which a lower score indicates better colour perception) (Farnsworth and Society, 1947); Higher order visual functions: (VOSP) (Warrington, 1991) Visual object and space perception battery Attention: Reaction time with and without auditory cue corrected for motor times (Schuhfried S7, (Schufried and Mîdling, 1992) Executive functions and cognition: (1) Frontal Assessment Battery (FAB) (Dubois et al., 2000); (2) Scales for Outcomes in Parkinson s disease, cognition (SCOPA-Cog) (Marinus et al., 2003) Control task: Prosody test (Bos et al., 2005), because the performance on this task is expected not to change with apomorphine infusion. The intravenous apomorphine infusion (30µg/kg/hr) lasted for 3 hours in total, after an initial bolus of 20µg/kg. All patients were pre-treated during 24 hours with 20 mg domperidon t.i.d., in order to prevent nausea and vomiting, which may be caused by apomorphine. One hour after the bolus injection and the infusion of apomorphine, a steady state plasma concentration was expected to be present. At that time the whole test battery was repeated. Two hours after having stopped the apomorphine infusion (4 times the elimination half-life of apomorphine (30 min), a selection of tests, not influenced by retest effects, were performed for the third time to control intra-individually for the effect during steady state infusion. Because of the short duration of the apomorphine infusion and the infrequent hallucinations in our patient group (which is usually the case in PD patients with VH) it was not possible to score the effects of the apomorphine on the VH Subjects Four patients were selected according to the following in- and exclusion criteria. Inclusion criteria: (1) Diagnosed Parkinson s disease according to the UK Brain Bank Criteria; (2) At least weekly hallucinations during the past month; (3) Mini Mental State Examination>24; (4) FAB>10; (5) Patients must be

100 7.3. PATIENTS AND METHODS 91 able to understand the procedures; (6) Medication must be stable for at least one month. Exclusion criteria: (1) Severe visual disorders (cataract, macula degeneration, severe retinal pathology, visual acuity< 0.5); (2) Patients with cerebral electrodes for deep brain stimulation; (3) Presence of other neurological or psychiatric disorders; (4) Unstable internal disease. All patients in this study had insight into their hallucinations and none of them had delusions. The levodopa equivalent dose of medication the patients were taking at the time of the study varied from 625 mg to 1950 mg (mean 1256). Three patients were on antipsychotic medication at the time of the study. Of these patients, one used clozapine (25 mg) and galantamine (24 mg), another used clozapine (25 mg) and rivastigmine (6 mg) and one used only rivastigmine (9 mg) Statistics To model the data, multilevel models for change were used. Multilevel modeling with repeated measures is preferable instead of analysis of variance procedures because fewer assumptions are made, which guarantees a more open analysis (Quene and van den Bergh, 2004). In the multilevel models, we used a random intercept parameter to model the pre-test data, since there were clear differences between patients in their pre-test scores. Because the effect of apomorphine was similar over all participants, no random slope parameter was added. Multilevel analysis can only estimate linear effects, whereas we were interested in quadratic effects over time; we expected a different score during apomorphine infusion compared to the pre- or post-test scores. Therefore two dummy variables were used in the analyses; both dummies used the pre-infusion measurement as the reference. The first dummy indicates whether there was a significant difference between the pre- and during infusion tests. The second dummy indicates whether there was a significant difference between the pre and post infusion measurements Ethics This study was approved by the Medical Ethical Committee of the University Medical Centre Groningen. All participants signed an informed consent prior

101 92 CHAPTER 7. APOMORPHINE IN PD WITH VH to study inclusion. 7.4 Results Apomorphine improved the contrast sensitivity during steady state infusion, as compared to the pre-test situation [t(8)=2,51; p=0,036]. While the mean scores on the pre- and post-tests were equal (1,55) [t(8)=-0,14; p=0,893], mean scores during the infusion increased with log 0,09. However, apomorphine did not show a significant effect on the UPDRS motor scores, on cognitive- and executive functions, and also not on visual acuity, colour perception and object and space perception (VOSP). On the contrary the reaction times were significantly higher during apomorphine infusion, as compared to pre-test data [t(6)=4,20; p=0,006]. The mean reaction times on the pre- and post-test were 345 ms and 371 ms respectively, while the mean reaction time during infusion was 430 ms. The difference between the pre- and post-test were not significant [t(6)=1,29; p=0,244]. A similar result was found for reaction times with auditory cue, which showed a significant difference between the pre- and during-infusion tests [t(6)=2,83; p=0,030] and no difference between the pre- and post tests [t(6)=1,19; p=0,281]. 7.5 Discussion The main hypothesis of this study was that apomorphine infusion would increase both lower and higher order visual functions. This hypothesis was partially confirmed, because only contrast sensitivity improved significantly, but not colour perception, visual acuity or visual object and space perception. Deficits in contrast sensitivity have been shown to be larger in PD patients with VH compared to patients without VH (Diederich et al., 1998). While a positive effect of apomorphine on contrast sensitivity in PD patients has been shown earlier (Buttner et al., 2000), no study so far has examined the effect of apomorphine in PD patients with VH. Retest effects are unlikely to explain the differences with the interval of 3 months between measurement 1 and 2, because own data on file did not show any retest effects 3 months after baseline, in patients with PD and VH. The UPDRS motor scores could be expected to improve with the infusion of apomorphine (Frankel et al., 1990). However, the UPDRS motor scores showed

102 7.5. DISCUSSION 93 a trend toward an increase instead of decrease during the infusion with apomorphine, as compared to the pre-test data. The scores during the infusion and at the post-infusion test were similar. The most likely explanation for this result is that we were looking at a ceiling effect, because the apomorphine was infused on top of the subjects oral medication, which already relieved most of their motor symptoms. This suggests that apomorphine infusion in this study specifically influenced non-motor functions, i.e. attention and contrast sensitivity, independent of an effect on motor symptoms. Apomorphine infusion increased reaction times, indicating a worsened attention. This finding is consistent with other data (Muller et al., 2002), indicating that apomorphine has a similar effect on attention in hallucinating and nonhallucinating PD patients. Apomorphine infusion did not improve or worsen any of the frontal and cognitive functions, which is also in accordance with previous data (Alegret et al., 2004). However, it might be the case that there is a positive effect on cognition, which could be masked by the decrease in attention. Our results have to be interpreted with caution, because of the open design and the small group size, which however seems to be enough to show significant differences. On the other hand, patients were their own controls and did not know anything about the value of the possible outcomes of the tests. We also included a control test (prosody test) which did not show any difference, indicating that patients did not increase their effort during the infusion. Finally, no difference was found in the VOSP scores, while the VOSP is likely to be influenced by both the effort of the patient and expectations of the person administering the tests. It was expected that the scores on the VOSP would increase during infusion. However, no differences were found, suggesting absence of an effect of apomorphine on higher order visual functions, but also absence of placebo effects. Our data fit nicely in the perception and attention deficit model, which suggests that VH may result from a combination of reduced perceptual input and lowered attention (Collerton et al., 2005). Apomorphine causes an increase of lower order visual perception and a decrease of attention, the two main factors causing VH according to this model. If an attention deficit is the main cause of VH in a patient, apomorphine will most likely increase the frequency of VH, as has been reported in the literature (Frankel et al., 1990). However, if the main cause of VH can be attributed to reduced perceptual input, apomorphine may reduce the frequency of VH, as has been reported previously as well, despite the negative effect on attention (Ellis et al., 1997). Our data suggest an impor-

103 94 CHAPTER 7. APOMORPHINE IN PD WITH VH tant mechanism by which apomorphine might act as an effective anti-psychotic in selected cases, consisting of the improvement of predominantly visual perceptive disorders. If severe attentional deficits are present as well, one should prescribe firstly a cholinesterase inhibitor, which may lead to improvement of the attentional performance and perhaps also of the VH. Once the attention (and/or VH) have improved, apomorphine can be reconsidered to control the existing motor fluctuations. Future studies should address this issue in more detail. Especially the interaction between cholinesterase inhibitors and apomorphine might be relevant for this group of PD patients with VH.

104 Chapter 8 Lasting visual hallucinations in visual deprivation; fmri correlates and the influence of rtms A. M. Meppelink 1,2, J. Koerts 1,2, M. A. Borg 1, K. L. Leenders 1,2, T. van Laar 1,2 (1) Department of Neurology, University Medical Center Groningen, The Netherlands (2) School of Behavioral and Cognitive Neurosciences, University of Groningen, the Netherlands Accepted for publication in the Journal of Neurology, Neurosurgery and Psychiatry 95

105 96 CHAPTER 8. HALLUCINATIONS IN VISUAL DEPRIVATION 8.1 Abstract Charles Bonnet s Syndrome (CBS) is characterized by complex visual hallucinations (VH) following visual impairment in psychologically normal people (Teunisse et al., 1996). We report a blind patient with a CBS-like syndrome, experiencing simple VH of colour and visual motion patterns for more than 20 years after bilateral eye disease. Functional MRI (fmri) revealed activations of visual colour- and motion areas. Repetitive transcranial magnetic stimulation (rtms) at these activation areas resulted in transient changes of the perceived visual pattern. 8.2 Case Report Clinical symptomatology A blind, otherwise healthy female (50y old) was referred to our hospital with visual complaints concerning ongoing sensations of colour and movement. She had suffered from bilateral eye disease (retinopathy), resulting in irreversible blindness 22 years ago. Ever since, she perceived visual sensations in the entire visual field, consisting of changing colors and a semi-transparent flow. The movement sensations showed a regular cyclic pattern; changing direction every 2 days, being slow when directed to the right and fast when directed to the left. Especially the first day with flow to the left was very disturbing. One year after the start of the visual sensations, both eyes were removed and replaced by prostheses, which of course had no effect on the visual sensations. Anti-epileptic and anti-psychotic drugs had no effect either FDG-PET and fmri F18-fluorodeoxyglucose Positron Emission Tomography (FDG-PET) and fmri were performed to get further insight in mechanisms underlying these visual sensations and to define a possible focus for therapeutic rtms. Cerebral FDG- PET showed a bilaterally reduced occipital and thalamic metabolism (Figure 8.1A). Using fmri, we localized brain regions specifically involved in either visual motion or color perception, by instructing the patient to focus attention to either color or movement. In a control condition, she counted internally. The

106 8.3. DISCUSSION 97 three conditions and a resting condition were balanced, pseudo-randomized and presented in a block design, during 3 Tesla fmri in two separate runs, lasting 15 minutes together. T1-weighted anatomical images were also made. Image processing and statistical analysis were conducted with Statistical Parametric Mapping (SPM) (Friston et al., 1995) version 5. Each condition was contrasted with the two others, resulting in 6 contrasts. Increased activations related to the visual motion- and color conditions were found in respectively extrastriate visual area V5/MT and the fusiform gyrus (Fig.8.1B). The observed extrastriate activations were accompanied by increased activation along the anterior calcarine sulcus, containing the peripheral visual field representation of V1 (Fig.8.1B) during the two conditions with attention to her visual sensations rtms Our patient was stimulated at 1 Hz at V5/MT in order to reduce the simple VH by rtms. Suppression of complex VH was previously shown by stimulating the occipital pole (Merabet et al., 2003). The fmri contrast motion counting was loaded into Brainlab. The nearest skull point to the left V5/MT was determined using neuronavigation. Firstly a 10 min sham rtms session on V5 was performed, single-blinded. After two weeks, a second rtms session was performed at 1 Hz during 10 minutes at 80 percent of the maximal output. After this second session the patient reported an almost complete disappearance of her slow phase, while a shaking visual motion sensation emerged. After extra rtms sessions at three consecutive days she reported mild fast phase reduction and a slight shortening of the cycle duration. 8.3 Discussion Baseline occipital activation was low, as demonstrated by reduced FDG uptake, probably reflecting deprivation of external visual input. fmri showed activations of both the visual motion area V5/MT and the fusiform gyrus by attending to either motion or color features of the visual sensations. Therefore it can be concluded that these two areas dedicated to visual motion and color processing (Zeki, 2001) were activated in a top-down fashion. Moreover, activation of the peripheral field representation in V1 was similarly enhanced,

107 98 CHAPTER 8. HALLUCINATIONS IN VISUAL DEPRIVATION Figure 8.1: (A) FDG-PET showing bilateral hypometabolism in the occipital cortex and the thalamus. (8) Cerebral fmri activation patterns projected on slices of the patient s own anatomy T1 scans. Attention to motion (upper panel) activated V5/MT (1) when compared to counting and, to a lesser extent, compared to attention to color. Attention to color (lower panel) activated the fusiform gyrus (2), compared to counting and, to a lesser extent, to attention to motion. Activation of the anterior calcarine sulcus (3) was seen in both motion versus counting and color versus counting, with stronger activation in the motion condition.

108 8.4. ACKNOWLEDGEMENTS 99 emphasizing the interactions along the segregated magno- and parvocellular pathways within striate- and extrastriate cortex. One should however remark that only the peripheral- and not central representation in V1 showed modulation of activity. Top-down evoked increases of visual cortex activation is consistent with the concept that central generators are the primary source of hallucinations, while interaction with specific sensory cortical regions defines the content of such hallucinations (Silbersweig et al., 1995; Ffytche et al., 1998). A possible generator of VH might include the left superior frontal gyrus, which was activated in both motion>counting and color>counting (not shown). However, an alternative explanation for this frontal activation might be the attentional effort, needed to perform the task. Although rtms at V5/MT slightly suppressed perceived visual motion, no lasting effect was induced. No effect was obtained either by rtms at V1. Maybe rtms has to be repeated more frequently, because such long-lasting VH may require a big change in neuronal function Future rtms strategies perhaps should also combine V5/MT with the left superior frontal gyrus. Our data nevertheless demonstrated logical patterns of increased regional activations associated with the nature of the hallucinations. This procedure may help targeting for rtms. 8.4 Acknowledgements We thank Martijn Beudel for his support with the fmri design and Arjen v.d. Hulzen for the neuronavigation.

109

110 Chapter 9 General Discussion 101

111 102 CHAPTER 9. GENERAL DISCUSSION 9.1 Introduction The hypothesis that VH in PD are associated with impaired visual perception and attention was investigated in this thesis. Below, the results of our clinical and imaging studies will be discussed in a broader perspective, related to visual processing and attention in PD. In addition, possible mechanisms of impaired visual perception and attention leading to VH will be considered, using a functional network approach. We will focus on two important aspects; 1) bottom-up and top-down processing and 2) modulation and selection. Interactions between and impairments in PD of processing, modulation and selection of visual stimuli will be described on a functional anatomical level, leading to a model on VH in PD that reveals future research areas of interest on pathogenesis and treatment of VH in PD. 9.2 Bottom-up processing Early visual processing Visual processing begins in the retina, where light is absorbed and transduced into electrical signals by the photoreceptors. The retina is part of the central nervous system and contains three functional classes of neurons; photoreceptors, interneurons (bipolar, horizontal and amacrine cells) and ganglion cells. Retinal signaling occurs in vertical and horizontal directions. Vertical neurotransmission takes place from photoreceptor to bipolar cell to ganglion cell. Axons of the ganglion cells form the optic nerve and fibers of each eye partly cross at the chiasm to form the optic tract, carrying the representation of the contralateral visual field. The human retina contains two types of photoreceptors; the cones, mediating bright-light, chromatic vision, and the rods, mediating low-light, achromatic vision. Cones are mainly present in the central part (fovea) of the retina and project to parvocellular and koniocellular ganglion cells, while rods are distributed in the peripheral retina and project to magnocellular ganglion cells. In this way already at the level of the retina, a functional segregation into a parvocellular processing stream, involved mainly in color vision, and a magnocellular stream, involved in motion perception, exists (Kandel, 2000). Horizontal transmission, through amacrine and horizontal cells, involves the combining of signals from several photoreceptors to

112 9.2. BOTTOM-UP PROCESSING 103 the ganglion cells, thus shaping the temporal and spatial qualities of chromatic and achromatic vision. Glutamate is the principal neurotransmitter of vertical transmission, while GABA and glycine mainly mediate horizontal transmission. In addition, dopaminergic (DA) neurons have been identified in the human retina (Frederick et al., 1982). These DA amacrine cells exhibit widespread dendritic arborizations and long axons ensuring overlap with other amacrine cells and bipolar cells. The density of DA amacrine cells is highest in the peripheral retina, paralleling the distribution of rods (Harris et al., 1992). In line with this, DA amacrine cells amplify contrast sensitivity through modulation of vertical transmission (Bodis-Wollner, 2009; Archibald et al., 2009) Cortical visual processing The optic tract projects to the lateral geniculate nucleus (LGN) of the thalamus, which is the main relay between the retina and the visual cortex. About 10 percent of retinal axons project to the superior colliculus (SC), involved in head movements and saccadic eye movements. In addition, some retinal axons project to the suprachiasmatic nucleus of the hypothalamus, involved in synchronizing biological rhythms, the pretectum in the midbrain, controlling pupillary reflexes, and the pulvinar nucleus of the thalamus. From the LGN, optic radiations project to the primary visual cortex (V1) in the occipital lobe. V1 is also called the striate cortex, because of the presence of a stripe of white matter, the stria of Gennari, in layer 4. Magno- and parvocellular axons project to different sublayers of LGN and V1, thus maintaining the segregation of these cellular pathways at successive levels of neuronal processing. Each half of the visual field is represented upside-down in the striate cortex around the contralateral calcarine sulcus, the fovea being represented in the most posterior half of V1. After processing of visual stimuli in V1, information is conveyed bottomup via area V2 to the occipito-parietal (containing the motion sensitive area V5) and occipito-temporal (with color- and form-sensitive V4) cortex, also called the dorsal and ventral visual stream, respectively. Input to the occipitoparietal pathway derives mainly from the magnocellular cells, while input to the occipito-temporal pathway derives from cells in both the magnocellular and parvocellular layers of the LGN (Ungerleider and Haxby, 1994). Retinogeniculate signals also project directly to V5, mostly via koniocellular neurons (Sincich et al., 2004). Area V6, located at the medial parieto-occipital region, is also involved in motion perception (Pitzalis et al., 2010). The dorsal route is

113 104 CHAPTER 9. GENERAL DISCUSSION also called vision-for-action, because of its important role in real-time actions to visual targets (Goodale et al., 2004). In accordance with this is the functional connectivity between parietal cortex and dorsolateral prefrontal cortex related to grasping (Hattori et al., 2009). The ventral and lateral occipitotemporal cortical areas are important in perceiving and recognizing visual objects (Grill-Spector, 2003; Downing et al., 2006). Several subregions in the occipito-temporal cortex exist that respond more strongly to specific object categories, such as the fusiform face area for faces and the parahippocampal place area for scenes (Kanwisher et al., 1997; Epstein et al., 1999b). Other regions that are important in visual object recognition are the fusiform gyrus (including the fusiform face area), the lingual gyrus, the lateral occipital complex and the middle temporal gyrus (Malach et al., 1995; Downing et al., 2006). The ventral route has been logically called vision-for-perception and was suggested to mediate memory-driven actions (Goodale et al., 2004) Visual impairments in Parkinson s disease Abnormalities in early visual processing have been shown to occur in PD patients studying Visual Evoked Potentials (VEPs). Furthermore, contrast sensitivity was especially reduced in the peripheral retina in PD, possibly reflecting malfunction of the DA amacrine cells (Bodis-Wollner, 1990; Harris et al., 1992). Direct functional evidence of retinal involvement in PD was shown using pattern electroretinography, assessing spatial tuning (Tagliati et al., 1996). Autopsy studies have shown reduced retinal dopamine levels of unmedicated PD patients, but normal levels in PD patients that received levodopa until death (Harnois and Di Paolo, 1990). Apart from DA amacrine cell dysfunction, multiple early visual pathways are affected in PD as well, including independent damage of the magno- and parvocellular pathways (Silva et al., 2005). Furthermore, a recent study showed inner retinal layer thinning in PD (Hajee et al., 2009). Color discrimination can be reduced in PD, but probably only a subset of PD patients may show chromatic deficits (Buttner et al., 1995; Pieri et al., 2000; Silva et al., 2005). Magnocellular impairment, leading to reduced achromatic contrast sensitivity, seems more generally associated with PD disease duration (Silva et al., 2005). Levodopa as well as apomorphine have been shown to improve both achromatic and chromatic contrast discrimination, although it is unclear if this is a retinal or a cortical effect (Buttner et al., 1994, 2000; Geerligs et al., 2009). For a comprehensive review on the retina in PD, see Archibald et al. (2009).

114 9.3. TOP-DOWN VISUAL PROCESSING AND ATTENTION 105 In addition to retinal dysfunction, higher visual function deficits occur in PD. Pre-attentive visual processing of orientation differences, without need for attentive visual search to perform the task, was shown to be impaired in nondemented PD patients, while attentive visual search was similar compared to healthy controls (Lieb et al., 1999). Another, more recent study also showed that visual perceptual impairments in PD patients are mainly pre-attentive or bottom-up. When saliency of stimuli in a visual search paradigm was decreased and no prior information (top-down, see below) about the identity of the object of interest was provided, PD patients scored worse than healthy controls (Horowitz et al., 2006). These data suggest pathological involvement of the striate and extrastriate visual cortex in PD, which is in accordance with neuroimaging studies showing hypoperfusion in the occipital and parietal cortex in non-demented PD patients (Abe et al., 2003). Given the fact that visual deprivation may lead to VH (Charles Bonnet Syndrome, CBS), both retinal and early visual cortex dysfunction might play a role in the generation of VH. 9.3 Top-down visual processing and attention Top-down visual processing Success in visual object recognition also depends on attentive, top-down influences that predict likely object identities. Top-down factors become increasingly important in circumstances of impoverished visual input, for example when images are noisy or partially degraded (Eger et al., 2006; Bar et al., 2006). It is thought that a partially analyzed version of the input image is rapidly projected from early visual areas to the prefrontal cortex, where it activates an initial guess, which is projected back to the temporal cortex (Bar et al., 2006). There, it is integrated in bottom-up visual processing. It is largely unknown through which pathways the feedforward information to the prefrontal cortex is conveyed. A magnocellular pathway via the superior longitudinal fasciculus of the dorsal stream was suggested (Kveraga et al., 2007), but also occipito-pulvinar-frontal projections and SC-amygdala-frontal projections might be involved. The latter route is primarily involved in the processing of threatening, fearful stimuli (Reinders et al., 2006). Another possibility is the inferior fronto-occipital fasciculus (IFOF), which projects from the inferior/lateral occipital lobe to inferior- and dorsolateral regions of the frontal lobe. Recently demonstrated reductions in white matter connectivity of the IFOF associated with an age-related decline in face perception were

115 106 CHAPTER 9. GENERAL DISCUSSION proposed to reflect top-down visual processing deficits (Thomas et al., 2008) Attention When multiple stimuli appear simultaneously in the visual field, they interact in a mutually suppressive way, competing for representation in the visual cortex. Selective attention biases competition among stimuli in the visual scene for representation in the visual cortex by coordinating selective visual information processing (Desimone and Duncan, 1995; Serences and Yantis, 2006). Cortical and subcortical areas exerting this bias in competition are source regions of attentional control, while visual cortex areas in which activity is modulated are target regions of attentional control. Attentional influence can be implemented by both top-down signals, depending on goals and expectations, and bottom-up signals that depend on the physical salience of the stimulus. Neurophysiological data confirm that both voluntary and stimulus-driven factors influence neuronal activity of target regions, the relative impact of the latter decreasing as incoming information ascends the cortical hierarchy from V1 to prefrontal cortex (Treue, 2003). Stimulus-driven bias, for example pop-out of a distinguishable stimulus from other stimuli, leads to increased activation in striate and extrastriate cortices (Beck and Kastner, 2005, 2009). The same effect is seen when attention is directed selectively to a specific feature or location of an image in target brain regions that are involved in the processing of the feature, for example activation of the fusiform gyrus when attending to changes in shape (Corbetta et al., 1991; Kastner et al., 1998). Using fmri, we have shown similar activations in a blind patient with CBS of both the visual motion area V5/MT and the fusiform gyrus by attending to either motion or color features of her visual hallucinations (Meppelink et al., 2010). So, attention can increase the signal to noise ratio in specific cortical target areas both in a bottom-up and a topdown manner. Cortical sources of attention include the prefrontal cortex (frontal eye fields, cingulate cortex) and the posterior parietal cortex. During selective attention, increased perfusion of posterior parietal and prefrontal cortices was seen (Fink et al., 1997). Right inferior parietal gyrus and right prefrontal activation were shown during a sustained attention task, with relative deactivations over time, that were associated with increased reaction times (Coull et al., 1998). Several studies showed transient activation of the superior parietal cortex when shifts of attention were made, regardless of the type of the attentional deploy-

116 9.3. TOP-DOWN VISUAL PROCESSING AND ATTENTION 107 ment (Liu et al., 2005; Serences and Yantis, 2006). This transient, domainindependent switch signal might enable a new attentional state, without carrying information about the parameters of the new state (Serences and Yantis, 2006). Thus, regions in both frontal and parietal cortex are likely candidates for the source of the biasing signal that, according to biased competition theory (Desimone and Duncan, 1995), resolves competition in the visual cortex (Beck and Kastner, 2009). Subcortical structures, including the basal ganglia and the thalamus, also play a role in selective attention and higher order visual processing. In the thalamus, two nuclei apart from the LGN are important in this respect; the pulvinar and the thalamic reticular nucleus (TRN). The pulvinar is highly connected with visual and attentional areas (Leh et al., 2008). Cortico-pulvino-cortical connections exist between parietal and frontal cortices and between occipital and inferior temporal areas, while fronto-parietal areas also connect to occipito-temporal areas via a cortico-colliculo-pulvino-cortical pathway (Saalmann and Kastner, 2009). In addition, the pulvinar receives input from the TRN, which is reciprocally connected to both LGN and pulvinar and provides inhibitory input to both. Considering these extensive connections, the pulvinar can be considered as a subcortical component of the attention network in the brain. Selective attention has been shown to increase activation or metabolism in the intact human pulvinar (LaBerge and Buchsbaum, 1990; Kastner et al., 2004; Smith et al., 2009), while pulvinar lesions in humans have been associated with deficits in selective attention (Snow et al., 2009). The described increase of activation or metabolism in the pulvinar during tasks requiring selective attention was proposed to reflect inhibition of irrelevant stimuli and facilitation of behaviorally relevant stimuli (Robinson and Petersen, 1992). The pulvinar, as well as the LGN, are regulated by the TRN, which also receives input from the SC and the striate, extrastriate and prefrontal cortices. After integration of these inputs, the TRN is in a position to regulate information transmission in the LGN and pulvinar, according to the behavioral context (Saalmann and Kastner, 2009). Enhanced visual attention inhibits TRN activity, releasing inhibition of TRN to the LGN (McAlonan et al., 2008), providing support for its guardian role of the thalamic gateway tot the cortex (Crick, 1984; McAlonan et al., 2008; Mayo, 2009). Temporal and occipital extrastriate cortices, as well as before-mentioned visually related ventrolateral prefrontal regions project to the body and tail of the caudate nucleus of the striatum. The striatum, encompassing the caudate nucleus, putamen and nucleus accumbens, is the main input structure of the

117 108 CHAPTER 9. GENERAL DISCUSSION basal ganglia (BG). The striatum receives input from the entire cortex as well as modulatory dopaminergic input from the substantia nigra pars compacta (SNc; caudate nucleus and putamen) and the VTA (nucleus accumbens), both localized in the ventral mesencephalon. The striatum projects to the output nuclei of the BG, the internal globus pallidus and the substantia nigra pars reticulata (SNr), via a direct pathway and via an indirect pathway. This indirect pathway also encompasses intrinsic nuclei of the BG; the external globus pallidus and the subthalamic nucleus (Mink, 1996). Both the internal globus pallidus and the SNr have an inhibitory connection to the thalamus, which in turn has excitatory projections back to the cortex, mainly the frontal lobe (Wolters et al., 2007). The BG play an important role in motor, cognitive and affective behavioral functions. The mechanism by which the BG contribute to these functions, seems to be through the selection of an appropriate response in a particular context and, in parallel, the suppression of inadequate responses (Redgrave et al., 1999; de Jong and Paans, 2007). They form a complex network of parallel, functionally segregated cortico-basal ganglia-thalamo-cortical loops. In the visual corticostriatal loop, the SNr receives input from ventral extrastriate cortices by way of the visual striatum (body and tail of the caudate nucleus) that projects back to visual extrastriate cortices via the thalamus (Middleton and Strick, 1996). This visual loop may contribute to selection of a particular interpretation of an ambiguous visual scene or updating visual working memory (Groenewegen, 2009). Human fmri studies show activation of the body and tail of the caudate during visual categorization (Seger and Cincotta, 2005). In addition, the visual loop may enable selection of appropriate motor programs on the basis of current visual processing, via output projections from the visual to the motor loop (Ashby et al., 2007). Specific BG structures and distinct thalamic nuclei thus play an important role in both selecting visual information to be brought in attentional focus and movements to be prepared for distinct action. 9.4 Parkinson s disease and visual hallucinations Impaired visual processing Reduced contrast sensitivity, impaired color discrimination as well as visual acuity, factors that are at least partly caused by retinal defects, have been associated with VH in PD (Diederich et al., 1998; Holroyd et al., 2001; Matsui

118 9.4. PARKINSON S DISEASE AND VISUAL HALLUCINATIONS 109 et al., 2006b). Apart from impaired early visual processing, several studies have shown that VH in PD are associated with impaired visual perception, due to deficits at later processing stages (Barnes et al., 2003; Ramirez-Ruiz et al., 2006, 2007a). We and others have shown that object and space perception in PD patients with VH is more impaired compared to PD patients without VH (Koerts et al., 2010). In addition, we have shown that the speed of identifying distinct images emerging from visual noise is decreased in PD patients with VH, compared to both PD patients without VH and healthy controls (Meppelink et al., 2008). Functional imaging studies have shown that the primary visual cortex is equally affected in PD patients with VH as compared to PD patients without VH (Boecker et al., 2007; Oishi et al., 2005). Visual association cortices, on the other hand, showed reduction of either activation, perfusion or metabolism during rest or simple visual stimulation in PD patients with VH, as compared to PD patients without VH (Okada et al., 1999; Oishi et al., 2005; Matsui et al., 2006a; Stebbins et al., 2004; Boecker et al., 2007). Using similar images emerging from visual noise as described above (Meppelink et al., 2008) during fmri, we have shown decreased activation of occipital and temporal extrastriate visual cortices, including the fusiform gyrus, before image recognition in PD patients with VH, compared to both PD patients without VH and healthy controls (Meppelink et al., 2009). This is in line with the previously described perfusion, activation and metabolism reductions in visual association cortices in these patients, as was shown by several groups (Okada et al., 1999; Oishi et al., 2005; Matsui et al., 2006a; Stebbins et al., 2004; Boecker et al., 2007) Impaired attentive top-down processing During visual search, PD patients relied more on top-down processing to compensate for their bottom-up visual processing deficits, when compared to healthy controls (Horowitz et al., 2006). PD patients with VH, with relatively more visual impairments compared to PD without VH, might therefore rely even more on top-down processing. Hypothetically, this could lead to excessive compensatory top-down visual processing and internal image generation, giving rise to VH. Involvement of top-down areas during VH was shown by several groups. Increased perfusion of the cingulate cortex and striatum was observed as a common feature during hallucinations in patients with schizophrenia, while the content of the hallucinations, being visual or auditory, was related to specific cortical activation (Silbersweig et al., 1995). Similarly, increased

119 110 CHAPTER 9. GENERAL DISCUSSION prefrontal perfusion or activation was seen in one PD patient and in some CBS patients during VH (Kataoka et al., 2008; Ffytche et al., 1998). Apart from activation of these top-down areas, all three studies also showed activation of visual association cortices during VH. Interestingly, it was shown in CBS that the content of VH reflected the functional specialization of regions in the extrastriate visual cortex (Ffytche et al., 1998). Frontal and parietal cortices may show compensatory increased activation during visual processing or, on the contrary, less activation, due to involvement of these top-down attentive areas in addition to bottom-up processing areas. One study showed increased activation of the inferior frontal gyrus and the caudate nucleus during simple visual stimulation in PD with VH, compared to PD without VH (Stebbins et al., 2004). Although this may reflect increased topdown involvement, the simple nature of the task makes this less likely. During visual processing of more complex, gradually revealed images we did not find support for the hypothesis of compensatory increased top-down activations during or before image recognition in PD patients with VH. In contrast, in the period before image recognition decreased activation of the right superior and middle frontal gyrus was seen in PD patients with VH, compared to PD patients without VH. In addition, a decreased activation of the inferior parietal cortex was seen in PD patients with VH, compared to healthy controls only (Meppelink et al., 2009). A recent fmri study has also shown reduced activation of the right ventrolateral prefrontal cortex during face perception in cognitively impaired PD patients with VH, compared to both PD without VH and healthy controls (Ramirez-Ruiz et al., 2008). Dysfunction of the lateral prefrontal cortex was proposed to reflect a deficit in suppression of irrelevant stimuli, which might predispose to VH. Another implication of the reduced activation of ventrolateral prefrontal cortex in PD patients with VH is that these patients may have reduced tendency to address external stimuli. While lateral prefrontal regions are associated with externally cued behavior, medial prefrontal activation is associated with internally guided behavior (de Jong and Paans, 2007). The superior frontal gyrus plays a role in endogenous allocation and maintenance of visual attention and was shown to be involved in the inhibition of internally represented information (Corbetta et al., 2002; de Jong and Paans, 2007). The process of distinguishing between internally- and externally- generated information is also called reality monitoring. Reduced activations of the anterior medial part of the superior frontal gyrus (i.e. medial anterior prefrontal cortex) during a reality monitoring task was associated with proneness to psychotic symptoms in healthy volunteers (Simons et al., 2008). Barnes

120 9.5. MODULATION 111 and colleagues have shown that PD patients with VH, when compared to PD patients without VH, had a greater propensity to report imaged stimuli as real percepts, which was interpreted as a reality-monitoring deficit (Barnes et al., 2003). In line with this, reduced activation or metabolism of the pulvinar was associated with an increased proneness to hallucinations in healthy subjects and in patients with schizophrenia (Ku et al., 2008; Hazlett et al., 2004). This suggests that the process of inhibiting irrelevant information of the pulvinar is less activated in an individual that is prone to VH, although no data on the pulvinar in PD exist. Relatively reduced activation of prefrontal and/or posterior parietal cortex in PD patients with VH might also reflect attentional impairments. This is in line with recent publications describing a decrease in selective attention as well as sustained attention in PD patients with VH, when compared to PD patients without VH, who again performed worse than healthy controls (Barnes and Boubert, 2008; Meppelink et al., 2008). Concluding, VH in PD are associated with impairment of both bottom-up and top-down visual processing and attention. The influence of modulatory neurotransmitter systems on these deficits and on what level they might occur in PD are discussed below. 9.5 Modulation of cortical and subcortical processing streams Connections between or within cortical areas use mainly glutamate (excitatory) and GABA (inhibitory) in their synapses, while thalamic projections are also glutaminergic. In contrast, ascending modulatory systems use monoaminergic [dopamine, (nor)epinephrin, serotonin] and cholinergic (acetylcholine) neurotransmitters, amongst others. Modulatory projections to the cortex and subcortical areas are widespread and play a role in arousal, attention and selection. We will discuss the role of ascending cholinergic projections on attention, sleep and dreaming and the role of dopaminergic projections on selection and visual processing, with the focus on VH in PD Dopaminergic projections Dopamine from neurons in the mesencephalon, which is released in the striatum, stimulates the direct pathway via D1 dopamine receptors and inhibits

121 112 CHAPTER 9. GENERAL DISCUSSION the indirect pathway via the D2 dopamine receptors. The net result is a disinhibition of the thalamus, leading to activation of the cortex and thus selection of (motor or behavioral) programs. In PD, less dopamine is delivered to the striatum, resulting in reduced stimulation of the direct pathway and reduced inhibition of the indirect pathway. The consequence of these effects is an increased inhibition of the thalamus and thus reduced motor or behavioral output. Classically, VH in PD have been viewed as an adverse effect of dopaminergic treatment for PD, causing a relative overstimulation of the mesolimbic dopaminergic receptors (Bosboom et al., 2004). The exact mechanism of this overstimulation in PD is unknown. In schizophrenia, increased ventral striatal and cingulate activity was shown, together with activations in distinct auditory- and visual cortical regions during respectively auditory- and visual hallucinations (Silbersweig et al., 1995). With fmri, increased caudate activation has been demonstrated during visual stimulation in PD patients with VH, compared to PD patients without VH, thus showing some resemblance with the above described functional imaging findings in schizophrenia (Stebbins et al., 2004). Limbic and paralimbic structures in the temporal lobe project to the ventral striatum. Together with dopaminergic input from the VTA, these projections seem to modulate the responsiveness of the ventral striatum to stimulation of other, prefrontal afferents (Epstein et al., 1999a). Antipsychotic drugs exert their effect mainly via blockade of the D2 subtype dopamine receptor, located at dopaminergic neurons in the midbrain, including the VTA, the stiatum and the prefrontal cortex (Westerink, 2002). The opposite effect, i.e. stimulation of dopamine receptors, might increase activation in VTA, ventral striatum, prefrontal and (para)limbic areas and might induce psychotic symptoms in schizophrenia (with an overactive dopamine system) or PD (due to extrinsic factors as levodopa or dopamine agonists). Alternatively, excessive stimulation of DA receptors in the visual striatum may lead to a net decrease of SNr activity and an abnormal increase in thalamic input to the temporal cortex (Middleton and Strick, 1996). VH in patients with dementia with Lewy bodies were associated with decreased dopamine transporter (DAT) binding in the caudate nucleus (Roselli et al., 2009). Reduced DAT binding in PD was shown to be associated with higher DA turnover which may lead to higher oscillations in synaptic DA, leading to transient states of intrinsic overstimulation (Sossi et al., 2007; Roselli et al., 2009). Extrinsic overstimulation may occur after administration of dopaminergic drugs. All types of dopaminergic drugs are associated with the induction or exac-

122 9.5. MODULATION 113 erbation of VH, although the evidence is stronger for dopaminergic agonists than for levodopa (Baker et al., 2009; Diederich et al., 2009). However, the hypothesis that VH in PD are simply caused by dopaminergic overstimulation has been challenged by several observations. First, a majority of PD patients on dopaminergic treatment do not report VH, while several studies report that the mean levodopa-equivalent dose is equal in PD patients with and without VH (Fenelon et al., 2000; Merims et al., 2004). Second, high-dose challenge with levodopa in non-demented PD patients with daily VH does not precipitate hallucinations (Goetz et al., 1998). Moreover, VH have already been reported in the pre-levodopa era (Fenelon et al., 2006) and have also been reported in typical PD patients in areas where levodopa treatment was not available (Dotchin et al., 2009). So, although the striatum and the dopaminergic system seem to be involved in the pathophysiology of VH in PD, evidence points to a broader involvement of several systems interacting Cholinergic modulation from the brainstem: PPN Apart from involvement of DA neurons in the SN and later also the VTA, several other brainstem nuclei, including the dorsal vagal nucleus, raphe nucleus, locus coeruleus and pedunculopontine nucleus (PPN) (Zweig et al., 1989) are affected in PD. The PPN is connected with the SC, SN, BG, thalamus, hypothalamus, cortex and other brainstem nuclei (raphe nucleus and locus coeruleus, a.o.) and exerts enhancing influence on many processes, including locomotor activity, sleep, attention and visual processing (Kobayashi and Isa, 2002). Apart from cholinergic neurons, the PPN contains also a substantial amount of glutaminergic neurons, that project to and activate the nucleus basalis of Meynert (NbM, see below). Regarding the enhancing role of the PPN on the thalamus, degeneration of the PPN was hypothesized to contribute to the presence of cognitive fluctuations, VH and/or sleep disturbances in α-synuclein pathology related neurodegenerative diseases, like PD. PPN cholinergic activity increases during REM sleep, together with attenuation of raphe serotonergic activity, leads to ponto-geniculate-occipital (PGO) wave production (Rye, 1997; Steriade, 2004). REM sleep behavioral disorder (RBD), consisting of loss of muscle atonia during sleep and vivid dreaming, frequently occurs in PD and might be related to the occurrence of VH. Although several studies have shown an association between RBD and VH in PD, the largest prospective longitudinal study in PD did not confirm this and only showed a relation of VH with vivid dreaming

123 114 CHAPTER 9. GENERAL DISCUSSION (Goetz et al., 2005). Moreover, PPN cell loss in DLB and multi system atrophy (MSA) was not associated with either VH, cognitive impairment or RBD (Schmeichel et al., 2008). This suggests that the degeneration of the PPN in PD may only partially explain alterations in cortical arousal and visual processing and is probably not directly involved in emergence of either RBD or VH Cholinergic modulation from the forebrain: NbM Attention is closely related to the cholinergic system (Sarter et al., 2001). Post-mortem studies in PD have shown loss of cortical cholinergic neurons and associated degeneration of the NbM, or Ch4 group (Mesulam, 2004), in the basal forebrain (Perry et al., 1985). Recent in vivo cholinergic tracer studies have shown cholinergic denervation of the occipital and parietal cortex in non-demented PD patients, while others have shown an association of cholinergic denervation with impaired attention (Hilker et al., 2005; Shimada et al., 2009; Bohnen et al., 2006). The cholinergic system plays an important role in awareness. A decrease in the cortical acetylcholine levels impairs the selection of subcortical information streams, causing unselected and chaotic cortical activation, which may predispose to hallucinations (Perry and Perry, 1995). Clinical evidence shows that VH can be induced by anti-cholinergics, while cholinesterase inhibitors ameliorate cognitive dysfunction and VH in PD (Burn et al., 2006; Wesnes et al., 2005). A recent study indirectly showed lower cortical acetylcholine in PD patients withvh, compared to PD patients without VH, using short-latency afferent inhibition (SAI) (Manganelli et al., 2009) Other modulatory systems A recent study showed that PD patients with VH had increased serotonin 2A receptor binding. In combination with the high affinity for serotonin 2A receptors of clozapine, regularly used to effectively treat VH in PD, suggests a role for this system as well (Schotte et al., 1993; Ballanger et al., 2010). The role of glutaminergic neurotransmission on VH in PD is still unclear, although the glutamate receptor antagonist memantine may improve cognition and VH in patients with PDD (Litvinenko et al., 2010).

124 9.6. VH IN PD: FROM PHENOMENOLOGY TO FUNCTIONAL ANATOMY VH in PD: from phenomenology to functional anatomy Regarding the phenomenology of VH in PD and the above summarized impairments in visual processing, attention and modulation, possible mechanisms on its pathophysiology and directions for future research are discussed here Visual processing and attention VH in PD typically consist of complex visual images, implying involvement, i.e. activation, of extrastriate visual cortices during VH. Decreased metabolism or activation during rest or visual perception, respectively, does not contradict with this. On the contrary, basal reduced activity in visual cortices could lead to release of higher order visual cortices, like in CBS, although the cause of underlying visual dysfunction is different. It was shown before that perfusion of the inferior frontal gyrus was increased during VH of a spider in one Parkinson s disease patient, together with increased perfusion of visual association areas (Kataoka et al., 2008). A comparable cerebral activation pattern was seen during hallucinations in patients with schizophrenia (Silbersweig et al., 1995). It is unclear however, what cortical region initiates activation increases within the visual perceptual network of temporal, frontal and perhaps parietal cortical activation during VH. An intra-operative stimulation study in epilepsy patients showed that stimulation of the prefrontal cortex (inferior frontal gyrus) can evoke complex VH, probably by propagation of activity from the prefrontal cortex along white matter tracts [uncinate fasciculus (Catani and Mesulam, 2008)] to the ventral occipito-temporal lobe (Blanke et al., 2000). Furthermore orbitofrontal seizures can present with complex VH, probably also by propagation of epileptic activity to temporal regions (La Vega-Talbot et al., 2006). Many patients hallucinate animals, people or objects in a behaviorally correct context. The often stereotypical environment where these VH occur (for example only in their own home) suggests that veridical visual information (i.e. the environment) elicits or at least facilitates a certain false perception (the VH). This strongly suggests involvement of the prefrontal cortex, where an initial guess of objects identities, based on a coarse representation of the object or the scene, can be projected to temporal association cortices (Bar et al., 2006). This frontal activation mainly occurs when visual input is suboptimal,

125 116 CHAPTER 9. GENERAL DISCUSSION which is again in accordance with visual perceptual impairments in PD with VH. Also, VH in PD mainly occur in the evening or night, when visual input is often reduced because of dim light. However, this typical circadian pattern is only present in part of the patients and does not account for VH during the day, suggesting that other mechanism additionally play a role (see below). VH in PD are mostly very realistic, unlike hallucinations in CBS, often consisting of fairy tale figures or Lilliputian figures (Teunisse et al., 1996). Distinguishing between externally and internally generated images, i.e. reality monitoring, was shown to be impaired in PD patients with VH (Barnes et al., 2003). The superior frontal gyrus, especially the medial part, plays an important role in this process and might be (partly) dysfunctional in PD with VH. A subcortical component of reality monitoring might be the pulvinar nucleus of the thalamus, which was shown to be activated when healthy subjects realized that an imposed illusion was not real. Future research might involve reality monitoring tasks during functional imaging in PD patients with VH, focusing on medial prefrontal and pulvinar (de)activation. Imaging during VH might reveal the time course of activation in cortical and subcortical structures, but is notoriously difficult, because VH in PD are usually infrequent and transient. Alternatively, if VH always occur in a specific visual scene, photographs of that scene might evoke VH and can be used during functional imaging Modulation and selection Selection of an appropriate visual image also involves the basal ganglia, projecting to extrastriate cortices via the thalamus. Dopaminergic modulation facilitates processing in cortico-basal ganglia-thalamo-cortical loops. Overstimulation, either intrinsic or extrinsic, might activate either the visual or the limbic (projecting to the prefrontal cortex) loop, which may lead to VH. Concurrent reduction of cholinergic modulation might lead to inappropriate activation of visual networks that are normally (with normal cholinergic input) inhibited through attention, focusing only at behaviorally relevant information. Attentional impairments in PD are associated with VH and might fluctuate during the day, like typically seen in dementia with Lewy bodies. Reduced NbM cholinergic output to the cortex, either secondary to reduced glutaminergic input from the PPN or due to cell loss in the NbM, is associated with attentional impairments in PD. In combination with other pontine nuclei, like the raphe (serotonin) and the coeruleus (noradrenergic), and hypothalamic nuclei, the PPN plays an impor-

126 9.7. CONCLUSIONS 117 tant role in the regulation of circadian rhythms like sleep and wakefulness. During REM sleep, levels of cortical ACh are similar to those during waking and twice or more the levels observed during slow-wave sleep (Rye, 1997; Steriade, 2004). No direct association between PPN cell loss, RBD and VH seems to exist. Possibly however, consciousness changes at sleep onset and during NREM sleep are more comparable with states in which VH in PD occur. In line with generation of hypnogogic hallucinations in healthy subjects, dysregulation of circadian rhythms in PD might predispose to VH. Narcolepsy, characterized by sleep attacks, VH and abnormalities of the sleep-wake cycle, is caused by a loss of hypocretin neurons in the hypothalamus (Thannickal et al., 2000). Recently it was shown that PD is also characterized by a massive loss of hypocretin neurons, which was associated with the clinical stage of PD (Thannickal et al., 2007). Because VH in PD also occur in more advanced disease stages, future research should explore a possible relation with a more severe hypocretin cell loss in PD patients experiencing VH. Increased inhibition of thalamocortical processing by the TRN in combination with reduced cholinergic input to the cortex might lead to increased intrinsic excitation of the cortex. 9.7 Conclusions Impairments of visual processing and attention in PD occur at different levels of the visual information processing system and are associated with the occurrence of VH. Visual impairments may be bottom-up, from retina to extrastriate visual cortex, and top-down, involving prefrontal and parietal cortices, and might be caused by a combination of reduced retinal dopamine, damage to magno- and parvocellular pathways, cortical Lewy bodies and/or atrophy. In addition, changed modulation of the cortex and the thalamus by brainstem and basal forebrain neurotransmitter systems can lead to dysregulation of circadian rhythms and faulty selection in cortico-basal-ganglia-thalamo-cortical circuits. The combination of impaired visual processing, fluctuating attention accompanying dysregulated circadian rhythms and decreased reality monitoring can give rise to internally generated images that are perceived as real perceptions, i.e. visual hallucinations. Obviously, not all these deficits have to be present in one individual with PD for VH to occur. It is of clinical relevance to investigate which of the aforementioned domains is impaired in a patient, to decide what treatment will be most beneficial. Future research in PD with VH should direct 1) the role of cortical choliner-

127 118 CHAPTER 9. GENERAL DISCUSSION gic deficits on impaired visual processing and attention, 2) reality monitoring focussing on prefrontal and pulvinar activations and 3) the influence of brainstem pathology, attention and circadian rhythms.

128 Summary and conclusion 119

129 120 SUMMARY AND CONCLUSION This thesis topic addresses underlying mechanisms of visual hallucinations (VH) in Parkinson s disease (PD). The main objectives of this thesis were to investigate: 1) the association between VH in PD and impairments of visual processing and attention and 2) the underlying functional cerebral architecture explored by fmri using visual activation paradigms in PD patients with VH. In addition, we explored the influence of pharmacological and nonpharmacological interventions on visual processing in PD and on VH in CBS, respectively. The results of these clinical and imaging studies were discussed in a wider perspective in the previous chapter. In the following section the results of each objective will be summarized and briefly discussed. Impaired visual processing and attention in PD with VH In chapter 2 and 3 we investigated the hypothesis that VH in PD are associated with impaired visual perception and attention. Importantly, the patients that participated in these studies were matched for education level and executive functioning, which was an important strategy to exclude such confounding factors that might otherwise influence performance on tests of visual perception and attention. In chapter 2 visual processing of gradually revealed images of animals, objects and people and sustained attention was demonstrated to be significantly slower in PD patients with VH than both PD patients without VH and healthy controls. Although recognition was slower, all images were correctly recognized. In addition, PD patients with VH showed decreased sustained attention compared to PD patients without VH, while the latter performed worse than healthy controls. In chapter 3 visual object and space perception was investigated in the same subjects. PD patients with VH showed impairment of both object and space perception, although such impairment was only statistically significant for some subtests, when compared to PD patients without VH and healthy controls. Impaired object and space perception in PD patients with VH was associated with a decrease of sustained visual attention, while slower image recognition in PD patients with VH was not. These clinical studies thus confirmed our hypothesis that VH in PD are indeed associated with deficits in visual processing and attention. Impairment of both object- and space perception suggests involvement of the parvo- as well as the magnocellular pathways to the ventral and dorsal visual streams, respectively (see chapter 9). We proposed that reduced speed of identifying distinct images emerging from visual noise in specifically PD patients with VH reflects impaired bottom-up processing in these patients. Hypothetically, this may lead to a higher demand on the top-down system, resulting in activation of visual images by a kind of over-

130 121 compensation, causing VH in PD. Functional and anatomical imaging in PD with VH To gain further insight in underlying mechanisms of VH in PD, we investigated cerebral activation patterns with fmri before and during recognition of the described gradually revealed images in these patients, compared to PD patients without VH and controls (chapter 4). We started with the hypothesis that PD patients with VH would show reduced activations of ventral visual association cortices before image recognition complemented by compensatory frontal or parietal activations, reflecting increased top-down processing. We indeed demonstrated that PD patients with VH were characterized by a pattern of decreased activation of lateral occipital cortex and extrastriate temporal visual cortices before image recognition. Contrary to our initial ideas, reduced (instead of increased) activation in a wider network included the lateral prefrontal cortex which suggested that cortical regions involved in top-down processing are additionally impaired. No differences were seen between PD patients without VH and healthy controls, implying a specific association of activation changes with VH and not with PD in general. No arguments for compensatory increases of activation in PD patients with VH were found, and thus no support for a link between vulnerability for VH and increased reliance on top-down processing during visual perception. Because none of the participating subjects experienced VH during scanning, the gradually revealed images were an indirect way to measure functional cerebral impairments associated with VH. Although the ventral/lateral temporal cortex and part of the prefrontal cortex were relatively impaired in PD patients with VH, one may still assume that activation increases occur in these regions during VH in these patients. Joint activation of frontal and visual association cortices during VH was shown in one patient with PD and in schizophrenia (Kataoka et al. 2008, Silbersweig et al. 1995), while activation of ventral visual association cortices during VH was demonstrated in CBS (ffytche 1998). Thus, diseaserelated factors like impaired visual perception, reflected in cortical activation reductions are associated with VH in PD. Underlying these functional deficits may be grey matter volume changes (see below), deposition of Lewy bodies or decreased cholinergic innervation (see below). In chapter 5 we used Voxel Based Morphometry (VBM) to investigate whether the functional differences in PD patients with VH (from chapter 4) were associated with structural, i.e. grey matter volume, changes. In addition, we assessed possible grey matter differences between all PD patients and healthy

131 122 SUMMARY AND CONCLUSION controls. In this study we have found no differences between PD patients with and without VH. However, grey matter decreases of bilateral prefrontal and parietal cortex, left anterior superior temporal and left middle occipital gyrus were found in the total group of PD patients, compared to controls. Most extensive grey matter volume reductions were found in the left parietal cortex in both non-demented patient groups, which was hemisphere-specific and independent of the side of PD symptoms. These results indicate that the functional deficits that we were able to identify in PD patients with VH are not associated with grey matter loss. Given the early stage of non-motor functional deterioration in our PD group with VH, functional deficit was regarded to possibly precede structural changes. We hypothesized that the strong left parietal reduction in both PD patient groups might reflect a secondary effect of basal ganglia disease, leading to impaired recruitment of internally guided motor programs and subsequent reduction of sustained skilled purposeful movements. Pharmacological and non-pharmacological interventions The second part of this thesis (chapters 6, 7 and 8) focused on therapeutic interventions in patients with VH. Chapter 6 described preliminary data of our follow-up study on cerebral activation during visual object processing from chapter 4. We hypothesized that the reduced activations in ventral/lateral visual association cortices in PD patients with VH, compared to both PD patients without VH and healthy controls, might result from decreased cholinergic input to these regions. This idea is consistent with the beneficial effects of cholinesterase-inhibitors, like rivastigmine, on VH in PD and PD dementia. Therefore, administration of rivastigmine was expected to normalize occipital and temporal cortex activation during image perception in PD patients with VH. We assessed cerebral activation patterns with fmri during image recognition in healthy controls and PD patients with VH in two sessions; one after administration of placebo and one after rivastigmine (double-blind, pseudorandomized design). Healthy controls showed robust bilateral fusiform- and lingual gyri activation in both treatments, while rivastigmine (compared to placebo) resulted in activation increases of the superior frontal gyrus, anterior cingulate and insula. PD patients with VH showed less robust activation than controls, without significant differences between treatment conditions. However, non-significant activation increases after rivastigmine were seen in bilateral (para)hippocampus, posterior superior frontal gyrus and striate and extrastriate visual cortices. The latter may reflect normalization of impaired visual cortex activation in PD with VH by rivastigmine. Extension of these preliminary data is necessary to confirm this effect. This study is ongoing.

132 123 Chapter 7 describes a small study investigating the effect of apomorphine on visual perception and attention in PD patients with VH. In our study, administration of apomorphine resulted in an increase of contrast sensitivity, possibly by stimulating retinal dopamine receptors, but prolonged reaction times in a selective attention task. These data may lead to the hypothesis that apomorphine improves VH in PD in some patients with mainly visual perceptive problems, but may also worsen VH in other patients, with mainly attentional impairments. In chapter 8, we describe a patient with a CBS-like syndrome due to retinal impairment, experiencing continuously present visual sensations of motion and colour. By instructing her to focus attention to 1) either one of these visual sensations or 2) a control condition during fmri, we were able to discriminate cerebral activation patterns related to the two types of visual sensations. Activation patterns in motion-sensitive area V5 were subsequently used for neuronavigation to localize the position for rtms. No clear effects were seen using inhibitory frequency rtms, neither on V5 nor V1. Nevertheless, this study provided clear support for the concept that areas dedicated to visual motion and color processing were activated in a top-down fashion and that this procedure may help targeting for rtms. In chapter 9 the results of our clinical and imaging studies were discussed in a broader perspective, focusing on interacting mechanisms of impaired visual perception and attention possibly leading to VH in PD. In this respect, we followed a functional network approach, focussing on two basic principles of cerebral organization; 1) bottom-up and top-down processing and 2) modulation and selection within such processing streams, including the description of cortical and subcortical interactions. Interactions between and impairments in PD of such processing, modulation and selection of visual stimuli were described on a functional anatomical level, leading to a model on VH in PD that provides a starting point for future research topics concerning pathogenesis and treatment of VH in PD. Final conclusions VH in PD are associated with impaired visual processing and reduced attention. Reduced activation of visual association cortices before image recognition reflects vulnerability for VH in PD. The demonstrated functional deficits in PD patients with VH were not a result of grey matter volume loss, but might precede such anatomical change.

133

134 Samenvatting 125

135 126 SAMENVATTING De ziekte van Parkinson (ZvP) is een progressieve aandoening van de hersenen waarbij zenuwcellen in de middenhersenen, die de neurotransmitter dopamine produceren, afsterven. Motore symptomen zoals tremor, bradykinesie, akinesie, rigiditeit en loopstoornissen staan op de voorgrond, terwijl ook nonmotorsymptomen zoals depressie, autonome dysfunctie en dementie, veelvuldig voorkomen. Visuele hallucinaties (VH) komen voor bij 30-50% van patiënten met de ZvP. VH worden gedefinieerd als onwillekeurige visuele ervaringen tijdens waken, zonder visuele stimulatie van buitenaf. VH kunnen simpel zijn (bijv. geometrische figuren) of complex (objecten, personen). Patiënten met de ZvP ervaren hoofdzakelijk complexe VH, waarbij mensen, dieren en objecten worden gezien. Het pathofysiologische mechanisme waardoor VH in de ZvP ontstaan is nog onbekend. Voorheen werden VH vooral gezien als bijwerking van dopaminerge medicatie, maar inmiddels is duidelijk dat ziektegerelateerde processen ook een belangrijke rol spelen in de pathofysiologie. Het kernthema van dit proefschrift betreft onderliggende mechanismen van VH in de ZvP. Dit onderwerp werd onderzocht door te kijken naar de mogelijke associatie tussen gestoorde visuele perceptie, verminderde aandacht en VH. Met behulp van klinische tests onderzochten we of VH in de ZvP geassocieerd zijn met stoornissen in visuele perceptie en aandacht. Verder werd functionele MRI (fmri) gebruikt om meer inzicht te krijgen in de onderliggende functionele cerebrale mechanismen in de ZvP met VH, gebruik makend van een visueel activatie paradigma. Tenslotte onderzochten we wat de invloed van farmacologische en niet-farmacologische interventies is op respectievelijk visuele verwerking in MP en VH in oogziekte. Gestoorde visuele perceptie en aandacht In hoofdstuk 2 en 3 was onze hypothese dat VH in de ZvP sneller optreden bij patiënten met gestoorde visuele perceptie en aandacht. De patiënten die hebben deelgenomen aan deze studies hadden een vergelijkbaar opleidingsniveau en niveau van executief functioneren. Dit is van belang, omdat deze factoren de scores op testen van visuele attentie en perceptie kunnen benvloeden. In hoofdstuk 2 onderzochten we de snelheid waarmee gezonde vrijwilligers, Parkinsonpatiënten zonder VH en Parkinsonpatiënten met VH een opkomend beeld herkenden. Hiervoor keken de deelnemers keken naar plaatjes van dieren, objecten en mensen die langzaam tevoorschijn kwamen uit visuele ruis, waarbij ze aan moesten geven wanneer zij het plaatje herkenden. Alle deelnemers herkenden de plaatjes, maar Parkinsonpatiënten met VH waren significant langzamer in het herkennen van deze plaatjes, vergeleken met zowel Parkin-

136 127 sonpatiënten zonder VH als gezonde vrijwilligers. Bovendien hadden Parkinsonpatiënten met VH meer moeite om aandacht vast te houden, vergeleken met Parkinsonpatiënten zonder VH. Deze laatsten presteerden in deze aandachtstaak overigens weer minder goed dan gezonde vrijwilligers. In hoofdstuk 3 werden visuele waarneming van voorwerpen (objectperceptie) en ruimtelijke waarneming (spatiële perceptie) onderzocht in dezelfde patiënten. Parkinsonpatiënten met VH presteerden minder goed op testen van zowel object- als spatiële perceptie, hoewel de resultaten van slechts enkele testen significant afweken van Parkinsonpatiënten zonder VH en gezonde vrijwilligers. Gestoorde visuele objectperceptie en perceptie van ruimtelijke verhoudingen in Parkinsonpatiënten met VH was geassocieerd met een afname in het vasthouden van de aandacht. Deze klinische studies ondersteunden onze hypothese dat VH in de ZvP geassocieerd zijn met stoornissen in visuele perceptie en aandacht. T.a.v. een mogelijke verklaring voor VH stelden we voor dat de verminderde snelheid waarmee Parkinsonpatiënten met VH de plaatjes die uit ruis tevoorschijn kwamen een afspiegeling is van een verstoring van zogenaamde bottom-up visuele verwerking. Dit zou, hypothetisch, kunnen leiden tot een toegenomen vraag naar top-down processen, resulterend in activatie van visuele representaties in het brein d.m.v. een soort overcompensatie, lijdend tot VH in de ZvP. Functionele en anatomische beeldvorming in Parkinsonpatiënten met VH Functionele MRI (fmri) werd gebruikt om meer inzicht te krijgen in welke hersengebieden betrokken zijn bij het ontstaan van VH in de ZvP. Wij analyseerden cerebrale activatiepatronen voor en tijdens herkenning van de hierboven beschreven plaatjes die tevoorschijn komen uit ruis. Wij zochten naar verschillen in deze patronen tussen Parkinsonpatiënten met VH vergeleken met Parkinsonpatiënten zonder VH en gezonde vrijwilligers (hoofdstuk 4). Onze hypothese was dat Parkinsonpatiënten met VH verminderde activatie zouden hebben in ventrale visuele associatiecortices in de fase voordat de plaatjes werden herkend en toegenomen frontale en/of pariëtale activatie, wijzend op verhoogde top-down activatie bij onvolledige bottom-up informatie. Wij vonden inderdaad dat Parkinsonpatiënten met VH gekarakteriseerd werden door een patroon van verminderde activatie van de laterale occipitale cortex en de temporale visuele cortices voor herkenning van de plaatjes. In tegenstelling tot onze verwachting van verhoogde activatie bij Parkinsonpatiënten met VH vonden wij verlaagde activatie van een uitgebreider netwerk, inclusief de laterale prefrontale cortex. Dit betekent dat niet alleen corticale gebieden

137 128 SAMENVATTING betrokken bij relatief vroege stadia van visuele verwerking zijn aangedaan in Parkinsonpatiënten met VH, maar ook in latere verwerkingsstations. Het zijn deze latere verwerkingsstations die in andere studies verhoogde activatie tijdens hallucinaties lieten zien. Wel is het van belang om op te merken dat de Parkinsonpatiënten die deelnamen aan onze studie geen hallucinaties hadden tijdens het scannen. Er werden geen verschillen gevonden tussen Parkinsonpatiënten zonder VH en gezonde vrijwilligers. Dit betekent dus dat er een specifiek verband bestaat tussen veranderingen in cerebrale activatie en VH en niet tussen deze veranderingen en de ZvP in het algemeen. Wij vonden geen argumenten voor compensatoire toename van activatie in Parkinsonpatiënten met VH en derhalve geen steun voor het idee dat Parkinsonpatiënten met VH meer afhankelijk zijn van toegenomen top-down verwerking tijdens visuele perceptie. Aangezien geen van de patiënten hallucineerde tijdens het scannen, zijn de geleidelijk zichtbaar wordende plaatjes een indirecte manier om functionele cerebrale veranderingen die geassocieerd zijn met VH te onderzoeken. Ondanks de relatief verminderde activatie van de ventrale/laterale temporale cortex en een deel van de prefrontale cortex, kan nog steeds worden verwacht dat deze gebieden geactiveerd worden tijdens VH in deze patiënten. Activatie van zowel frontale cortex als visuele associatiecortices werd gezien tijdens VH in een patiënt met de ZvP en in patiënten met schizofrenie (Kataoka et al., 2008; Silbersweig et al., 1995). Activatie van de ventrale visuele associatiecortex tijdens VH werd ook getoond in patiënten met het Charles Bonnet Syndroom (waarbij patiënten slechtziend of blind zijn en last hebben van VH) (Ffytche et al., 1998). Samenvattend kan gesteld worden dat ziekte-gerelateerde factoren zoals verstoorde visuele perceptie, weerspiegeld in corticale activatieafnames, een rol lijken te spelen in het ontstaan van VH in de ZvP. De oorzaak voor deze functionele stoornissen zou kunnen worden gezocht in de pathologie (depositie van Lewy lichaampjes), volumeveranderingen van de grijze stof (zie hieronder) of verminderde cholinerge innervatie (zie hieronder). In hoofdstuk 5 hebben wij Voxel Based Morphometry (VBM) gebruikt om te onderzoeken of de functionele verschillen tussen Parkinsonpatiënten met VH (van hoofdstuk 4) geassocieerd waren met structurele (d.w.z. van de grijze stof) veranderingen. Bovendien hebben wij onderzocht of er verschillen waren in grijze stof tussen Parkinsonpatiënten in het algemeen en gezonde vrijwilligers. In deze studie vonden wij geen verschillen tussen Parkinsonpatiënten met en zonder VH. Daarentegen toonden de Parkinsonpatiënten in het algemeen een afname van grijze stof in o.a. bilaterale prefrontale en pariëtale

138 129 cortices, vergeleken met de gezonde vrijwilligers. De meest uitgesproken afname van grijze stof volume werd gevonden in de linker pariëtale cortex in beide (niet-demente) patiëntengroepen. Dit resultaat was hemisfeer-specifiek en niet afhankelijk van de dominante kant van Parkinson-klachten. Deze resultaten laten zien dat de eerder getoonde functionele storingen geen verband houden met verlies van grijze stof. Dit kan betekenen dat functionele verschillen tussen Parkinsonpatiënten met en zonder VH voorafgaan aan anatomische verschillen. De grijze stofreductie in juist de linker pariëtaalkwab zou een secundair effect kunnen zijn van ziekte van de basale ganglia. De basale ganglia aandoening leidt tot een verstoord rekruteren van intern gegenereerde motor programma s en daardoor afname van volgehouden doelgerichte bewegingen. Deze functionele verklaring is consistent met bevindingen van toegenomen grijze stof in de linker pariëtale cortex na intensief trainen van doelgerichte bewegingen (bijv. jongleren). Farmacologische en niet-farmacologische interventies Het tweede deel van dit proefschrift (hoofdstukken 6, 7 en 8) was gericht op therapeutische interventies in patiënten met VH. In hoofdstuk 6 beschrijven wij de voorlopige resultaten van onze vervolgstudie naar een cholinerg effect op cerebrale activatie tijdens visuele object perceptie van hoofdstuk 4. Onze hypothese was dat de verminderde activaties in ventrale/laterale visuele associatiecortices in Parkinsonpatiënten met VH, vergeleken met zowel Parkinsonpatiënten zonder VH als gezonde vrijwilligers, mogelijk het gevolg zijn van een afgenomen cholinerge innervatie van deze gebieden. Dit idee is consistent met de gunstige effecten van cholinesterase-remmers, zoals rivastigmine, op VH in de ZvP en Parkinsondementie. De verwachting was daarom dat toediening van rivastigmine occipitale en temporale corticale activatie in Parkinsonpatiënten met VH zou kunnen normaliseren. Hiertoe onderzochten wij cerebrale activatiepatronen met fmri tijdens visuele perceptie in gezonde vrijwilligers en Parkinsonpatiënten met VH in twee sessies; een na toediening van placebo en een na toediening van rivastigmine (dubbel-blind, pseudogerandomiseerd design). Gezonde vrijwilligers toonden robuuste activatie van bilaterale fusiforme en linguale gyri na beide behandelingen, terwijl rivastigmine (vergeleken met placebo) resulteerde in activatietoename van de superior frontale gyrus, anterieure cingulate en insula. Parkinsonpatiënten met VH toonden een minder robuuste activatie dan gezonden, zonder significante verschillen tussen de twee behandelingen. Echter, nietsignificante toename van activatie werd gezien na toediening van rivastigmine (vergeleken met placebo) in bilaterale (para)hippocampus, posterior superior

139 130 SAMENVATTING frontale gyrus en striate en extrastriate visuele cortices. Dit laatste zou kunnen betekenen dat rivastigmine een normalisatie geeft van gestoorde visuele cortex activatie in Parkinsonpatiënten met VH zoals we zagen in hoofdstuk 4. Uitbreiding van deze voorlopige resultaten is noodzakelijk om dit effect te bevestigen. In dit proefschrift presenteren we de eerste resultaten. De studie loopt nog. In hoofdstuk 7 beschrijven wij een kleine studie naar het effect van apomorfine op visuele perceptie en aandacht in Parkinsonpatiënten met VH. In deze studie zagen wij dat de contrastgevoeligheid toenam na toediening van apomorfine, mogelijk door stimulatie van retinale dopaminerge receptoren, terwijl tegelijkertijd de reactietijden in de selectieve aandachtstaak toenamen. Deze data leidden tot de hypothese dat apomorfine VH zou kunnen verminderen in Parkinsonpatiënten met vooral visueel perceptieve problemen, terwijl het VH zou kunnen verergeren in Parkinsonpatiënten met vooral aandachtsproblemen. In hoofdstuk 8 beschrijven wij een blinde patiënte met visuele ervaringen die lijken op het Charles Bonnet Syndroom. Het ging om visuele sensaties van beweging en kleur die continu aanwezig waren. Wij lieten haar de aandacht richten op één van deze sensaties (kleur dan wel beweging) of op een controleconditie tijdens fmri. Op deze manier konden we hersengebieden identificeren die betrokken waren bij de twee types visuele sensaties. Activatie in hersengebied V5, betrokken bij visuele perceptie van beweging, werd vervolgens gebruikt voor neuronavigatie om de positie te bepalen voor repetitieve transcraniële magnetische stimulatie (rtms). Wij vonden geen duidelijk effect van rtms met inhibitoire frequentie na stimulatie van V5 noch van de primaire visuele cortex V1. Wel ondersteunt deze studie het idee dat hersengebieden betrokken bij kleur- en bewegingsperceptie geactiveerd werden op een top-down manier. Bovendien kan de beschreven procedure in de toekomst uitkomst bieden om doelgebieden voor rtms te identificeren. In hoofdstuk 9 werden de resultaten van onze klinische en imaging studies in een breder perspectief geplaatst. We schetsen een uitgebreid model met mechanismen van en interacties tussen verstoorde visuele perceptie en attentie, mogelijk lijdend tot VH in de ZvP. Deze functionele netwerkbenadering berust op drie basisprincipes van cerebrale organisatie; 1) bottom-up en topdown verwerking, 2) modulatie (vanuit hersenstam en basale voorhersenen naar cortex en thalamus) en 3) selectie (binnen cortex - basale ganglia thalamus netwerken). Interacties tussen stoornissen in de ZvP van verwerking, modulatie en selectie van visuele stimuli worden beschreven op een functioneel anatomisch niveau, leidend tot een model over VH in Parkinson dat kan worden gebruikt als startpunt voor toekomstig onderzoek naar de pathogenese en

140 131 behandeling van VH in de ZvP. Conclusie Visuele hallucinaties (VH) in de ziekte van Parkinson zijn geassocieerd met stoornissen van visuele perceptie en verminderde aandacht. Verminderde activatie van visuele associatiecortices voor herkenning van beelden weerspiegelt de toegenomen gevoeligheid voor het ontstaan van VH in de ziekte van Parkinson. Deze VH-geassocieerde functionele defecten in Parkinsonpatiënten gingen niet gepaard met volumeverlies van grijze stof, maar zouden wel vooraf kunnen gaan aan deze anatomische veranderingen.

141

142 Bibliography Aarsland, D., Andersen, K., Larsen, J. P., Lolk, A., and Kragh-Sorensen, P. (2003). Prevalence and characteristics of dementia in parkinson disease: an 8-year prospective study. Arch.Neurol., 60(3): Aarsland, D., Zaccai, J., and Brayne, C. (2005). A systematic review of prevalence studies of dementia in parkinson s disease. Mov Disord., 20(10): Abe, Y., Kachi, T., Kato, T., Arahata, Y., Yamada, T., Washimi, Y., Iwai, K., Ito, K., Yanagisawa, N., and Sobue, G. (2003). Occipital hypoperfusion in parkinson s disease without dementia: correlation to impaired cortical visual processing. J.Neurol.Neurosurg.Psychiatry, 74(4): Alegret, M., Valldeoriola, F., Marti, M., Pilleri, M., Junque, C., Rumia, J., and Tolosa, E. (2004). Comparative cognitive effects of bilateral subthalamic stimulation and subcutaneous continuous infusion of apomorphine in parkinson s disease. Mov Disord., 19(12): Aleman, A., Laroi, F., and Association, A. P. (2008). science of idiosyncratic perception. Washington DC. Hallucinations. The Archibald, N. K., Clarke, M. P., Mosimann, U. P., and Burn, D. J. (2009). The retina in parkinson s disease. Brain, 132(Pt 5): Arditi, A. (2005). Improving the design of the letter contrast sensitivity test. Invest Ophthalmol.Vis.Sci., 46(6): Ashby, F. G., Ennis, J. M., and Spiering, B. J. (2007). A neurobiological theory of automaticity in perceptual categorization. Psychol.Rev., 114(3): Baker, W. L., Silver, D., White, C. M., Kluger, J., Aberle, J., Patel, A. A., and Coleman, C. I. (2009). Dopamine agonists in the treatment of early parkinson s disease: a meta-analysis. Parkinsonism.Relat Disord., 15(4):

143 134 BIBLIOGRAPHY Ballanger, B., Strafella, A. P., van Eimeren, T., Zurowski, M., Rusjan, P. M., Houle, S., and Fox, S. H. (2010). Serotonin 2a receptors and visual hallucinations in parkinson disease. Arch.Neurol., 67(4): Ballard, C. G., Aarsland, D., McKeith, I., O Brien, J., Gray, A., Cormack, F., Burn, D., Cassidy, T., Starfeldt, R., Larsen, J. P., Brown, R., and Tovee, M. (2002). Fluctuations in attention: Pd dementia vs dlb with parkinsonism. Neurology, 59(11): Bar, M., Kassam, K. S., Ghuman, A. S., Boshyan, J., Schmid, A. M., Dale, A. M., Hamalainen, M. S., Marinkovic, K., Schacter, D. L., Rosen, B. R., and Halgren, E. (2006). Top-down facilitation of visual recognition. Proc.Natl.Acad.Sci.U.S.A, 103(2): Bar, M., Tootell, R. B., Schacter, D. L., Greve, D. N., Fischl, B., Mendola, J. D., Rosen, B. R., and Dale, A. M. (2001). Cortical mechanisms specific to explicit visual object recognition. Neuron, 29(2): Barnes, J. and Boubert, L. (2008). Executive functions are impaired in patients with parkinson s disease with visual hallucinations. J.Neurol.Neurosurg.Psychiatry, 79(2): Barnes, J., Boubert, L., Harris, J., Lee, A., and David, A. S. (2003). Reality monitoring and visual hallucinations in parkinson s disease. Neuropsychologia, 41(5): Barnes, J. and David, A. S. (2001). Visual hallucinations in parkinson s disease: a review and phenomenological survey. J.Neurol.Neurosurg.Psychiatry, 70(6): BECK, A. T., WARD, C. H., MENDELSON, M., MOCK, J., and ERBAUGH, J. (1961). An inventory for measuring depression. Arch.Gen.Psychiatry, 4: Beck, D. M. and Kastner, S. (2005). Stimulus context modulates competition in human extrastriate cortex. Nat.Neurosci., 8(8): Beck, D. M. and Kastner, S. (2009). Top-down and bottom-up mechanisms in biasing competition in the human brain. Vision Res., 49(10): Beyer, M. K., Janvin, C. C., Larsen, J. P., and Aarsland, D. (2007). A magnetic resonance imaging study of patients with parkinson s disease with mild cognitive impairment and dementia using voxel-based morphometry. J.Neurol.Neurosurg.Psychiatry, 78(3):

144 BIBLIOGRAPHY 135 Binkofski, F., Buccino, G., Stephan, K. M., Rizzolatti, G., Seitz, R. J., and Freund, H. J. (1999). A parieto-premotor network for object manipulation: evidence from neuroimaging. Exp.Brain Res., 128(1-2): Biousse, V., Skibell, B. C., Watts, R. L., Loupe, D. N., Drews-Botsch, C., and Newman, N. J. (2004). Ophthalmologic features of parkinson s disease. Neurology, 62(2): Blanke, O., Landis, T., and Seeck, M. (2000). Electrical cortical stimulation of the human prefrontal cortex evokes complex visual hallucinations. Epilepsy Behav., 1(5): Bodis-Wollner, I. (1990). The visual system in parkinson s disease. Res.Publ.Assoc.Res.Nerv.Ment.Dis., 67: Bodis-Wollner, I. (2009). Retinopathy in parkinson disease. J.Neural Transm., 116(11): Boecker, H., Ceballos-Baumann, A. O., Volk, D., Conrad, B., Forstl, H., and Haussermann, P. (2007). Metabolic alterations in patients with parkinson disease and visual hallucinations. Arch.Neurol., 64(7): Bohnen, N. I., Kaufer, D. I., Hendrickson, R., Ivanco, L. S., Lopresti, B. J., Constantine, G. M., Mathis, C., Davis, J. G., Moore, R. Y., and Dekosky, S. T. (2006). Cognitive correlates of cortical cholinergic denervation in parkinson s disease and parkinsonian dementia. J.Neurol., 253(2): Bos, E. H., Bouhuys, A. L., Geerts, E., van Os, T. W., Van der Spoel, I. D., Brouwer, W. H., and Ormel, J. (2005). Cognitive, physiological, and personality correlates of recurrence of depression. J.Affect.Disord., 87(2-3): Bosboom, J. L., Stoffers, D., and Wolters, E. C. (2004). Cognitive dysfunction and dementia in parkinson s disease. J.Neural Transm., 111(10-11): Braak, H., Ghebremedhin, E., Rub, U., Bratzke, H., and Del Tredici, K. (2004). Stages in the development of parkinson s disease-related pathology. Cell Tissue Res., 318(1): Braak, H., Rub, U., Jansen Steur, E. N., Del Tredici, K., and de Vos, R. A. (2005). Cognitive status correlates with neuropathologic stage in parkinson disease. Neurology, 64(8): Brainard, D. H. (1997). The psychophysics toolbox. Spat.Vis., 10(4):

145 136 BIBLIOGRAPHY Buning, H. (1997). Robust analysis of variance. Journal of Applied Statistics, 24(3): Burke, W. (2002). The neural basis of charles bonnet hallucinations: a hypothesis. J.Neurol.Neurosurg.Psychiatry, 73(5): Burn, D., Emre, M., McKeith, I., De Deyn, P. P., Aarsland, D., Hsu, C., and Lane, R. (2006). Effects of rivastigmine in patients with and without visual hallucinations in dementia associated with parkinson s disease. Mov Disord., 21(11): Burton, E. J., McKeith, I. G., Burn, D. J., Williams, E. D., and O Brien, J. T. (2004). Cerebral atrophy in parkinson s disease with and without dementia: a comparison with alzheimer s disease, dementia with lewy bodies and controls. Brain, 127(Pt 4): Buttner, T., Kuhn, W., Muller, T., Patzold, T., Heidbrink, K., and Przuntek, H. (1995). Distorted color discrimination in de novo parkinsonian patients. Neurology, 45(2): Buttner, T., Kuhn, W., Patzold, T., and Przuntek, H. (1994). L-dopa improves colour vision in parkinson s disease. J.Neural Transm.Park Dis.Dement.Sect., 7(1): Buttner, T., Muller, T., and Kuhn, W. (2000). Effects of apomorphine on visual functions in parkinson s disease. J.Neural Transm., 107(1): Catani, M. and Mesulam, M. (2008). The arcuate fasciculus and the disconnection theme in language and aphasia: history and current state. Cortex, 44(8): Collerton, D., Perry, E., and McKeith, I. (2005). Why people see things that are not there: A novel perception and attention deficit model for recurrent complex visual hallucinations. Behav.Brain Sci., 28(6): Colosimo, C., Hughes, A. J., Kilford, L., and Lees, A. J. (2003). Lewy body cortical involvement may not always predict dementia in parkinson s disease. J.Neurol.Neurosurg.Psychiatry, 74(7): Corbetta, M., Kincade, J. M., and Shulman, G. L. (2002). Neural systems for visual orienting and their relationships to spatial working memory. J.Cogn Neurosci., 14(3):

146 BIBLIOGRAPHY 137 Corbetta, M., Miezin, F. M., Dobmeyer, S., Shulman, G. L., and Petersen, S. E. (1991). Selective and divided attention during visual discriminations of shape, color, and speed: functional anatomy by positron emission tomography. J.Neurosci., 11(8): Coull, J. T., Frackowiak, R. S., and Frith, C. D. (1998). Monitoring for target objects: activation of right frontal and parietal cortices with increasing time on task. Neuropsychologia, 36(12): Crick, F. (1984). Function of the thalamic reticular complex: the searchlight hypothesis. Proc.Natl.Acad.Sci.U.S.A, 81(14): Cunnington, R., Egan, G. F., O Sullivan, J. D., Hughes, A. J., Bradshaw, J. L., and Colebatch, J. G. (2001). Motor imagery in parkinson s disease: a pet study. Mov Disord., 16(5): Dagher, A. and Nagano-Saito, A. (2007). Functional and anatomical magnetic resonance imaging in parkinson s disease. Mol.Imaging Biol., 9(4): Davidsdottir, S., Cronin-Golomb, A., and Lee, A. (2005). Visual and spatial symptoms in parkinson s disease. Vision Res., 45(10): de Jong, B. M. and Paans, A. M. (2007). Medial versus lateral prefrontal dissociation in movement selection and inhibitory control. Brain Res., 1132(1): de Jong, B. M., van der Graaf, F. H., and Paans, A. M. (2001). Brain activation related to the representations of external space and body scheme in visuomotor control. Neuroimage., 14(5): de Jong, B. M., van Zomeren, A. H., Willemsen, A. T., and Paans, A. M. (1996). Brain activity related to serial cognitive performance resembles circuitry of higher order motor control. Exp.Brain Res., 109(1): de Jong, B. M., Willemsen, A. T., and Paans, A. M. (1999). Brain activation related to the change between bimanual motor programs. Neuroimage., 9(3): Desimone, R. and Duncan, J. (1995). Neural mechanisms of selective visual attention. Annu.Rev.Neurosci., 18: Di Rosa, A. E., Epifanio, A., Antonini, A., Stocchi, F., Martino, G., Di Blasi, L., Tetto, A., Basile, G., Imbesi, D., La Spina, P., Di Raimondo, G., and

147 138 BIBLIOGRAPHY Morgante, L. (2003). Continuous apomorphine infusion and neuropsychiatric disorders: a controlled study in patients with advanced parkinson s disease. Neurol.Sci., 24(3): Diederich, N. J., Fenelon, G., Stebbins, G., and Goetz, C. G. (2009). Hallucinations in parkinson disease. Nat.Rev.Neurol., 5(6): Diederich, N. J., Goetz, C. G., Raman, R., Pappert, E. J., Leurgans, S., and Piery, V. (1998). Poor visual discrimination and visual hallucinations in parkinson s disease. Clin.Neuropharmacol., 21(5): Diederich, N. J., Goetz, C. G., and Stebbins, G. T. (2005). Repeated visual hallucinations in parkinson s disease as disturbed external/internal perceptions: focused review and a new integrative model. Mov Disord., 20(2): Diederich, N. J., Pieri, V., and Goetz, C. G. (2003). Coping strategies for visual hallucinations in parkinson s disease. Mov Disord., 18(7): Dotchin, C. L., Jusabani, A., and Walker, R. W. (2009). Non-motor symptoms in a prevalent population with parkinson s disease in tanzania. Parkinsonism.Relat Disord., 15(6): Downing, P. E., Chan, A. W., Peelen, M. V., Dodds, C. M., and Kanwisher, N. (2006). Domain specificity in visual cortex. Cereb.Cortex, 16(10): Draganski, B., Gaser, C., Busch, V., Schuierer, G., Bogdahn, U., and May, A. (2004). Neuroplasticity: changes in grey matter induced by training. Nature, 427(6972): Draganski, B. and May, A. (2008). Training-induced structural changes in the adult human brain. Behav.Brain Res., 192(1): Dubois, B. and Pillon, B. (1997). J.Neurol., 244(1):2 8. Cognitive deficits in parkinson s disease. Dubois, B., Slachevsky, A., Litvan, I., and Pillon, B. (2000). The fab: a frontal assessment battery at bedside. Neurology, 55(11): Eger, E., Henson, R. N., Driver, J., and Dolan, R. J. (2006). Mechanisms of top-down facilitation in perception of visual objects studied by fmri. Cereb.Cortex. Ellis, C., Lemmens, G., Parkes, J. D., Abbott, R. J., Pye, I. F., Leigh, P. N., and Chaudhuri, K. R. (1997). Use of apomorphine in parkinsonian patients with neuropsychiatric complications to oral treatment. Parkinsonism.Relat Disord., 3(2):

148 BIBLIOGRAPHY 139 Epstein, J., Stern, E., and Silbersweig, D. (1999a). Mesolimbic activity associated with psychosis in schizophrenia. symptom-specific pet studies. Ann.N.Y.Acad.Sci., 877: Epstein, R., Harris, A., Stanley, D., and Kanwisher, N. (1999b). The parahippocampal place area: recognition, navigation, or encoding? Neuron, 23(1): Esselink, R. A., de Bie, R. M., de Haan, R. J., Lenders, M. W., Nijssen, P. C., Staal, M. J., Smeding, H. M., Schuurman, P. R., Bosch, D. A., and Speelman, J. D. (2004). Unilateral pallidotomy versus bilateral subthalamic nucleus stimulation in pd: a randomized trial. Neurology, 62(2): Farnsworth, D. and Society, T. P. (1947). The Farnsworth Dichotomous Test for Color Blindness. New York. Faul, F., Erdfelder, E., Lang, A. G., and Buchner, A. (2007). G*power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav.Res.Methods, 39(2): Feldmann, A., Illes, Z., Kosztolanyi, P., Illes, E., Mike, A., Kover, F., Balas, I., Kovacs, N., and Nagy, F. (2008). Morphometric changes of gray matter in parkinson s disease with depression: a voxel-based morphometry study. Mov Disord., 23(1): Fenelon, G., Goetz, C. G., and Karenberg, A. (2006). Hallucinations in parkinson disease in the prelevodopa era. Neurology, 66(1): Fenelon, G., Mahieux, F., Huon, R., and Ziegler, M. (2000). Hallucinations in parkinson s disease: prevalence, phenomenology and risk factors. Brain, 123 ( Pt 4): Fenelon, G., Souals, T., de Langavant L., C., and C, B.-L. A. (2008). Sense of presence in parkinson s disease: A prospective phenomenological study. Ffytche, D. H., Howard, R. J., Brammer, M. J., David, A., Woodruff, P., and Williams, S. (1998). The anatomy of conscious vision: an fmri study of visual hallucinations. Nat.Neurosci., 1(8): Fink, G. R., Dolan, R. J., Halligan, P. W., Marshall, J. C., and Frith, C. D. (1997). Space-based and object-based visual attention: shared and specific neural domains. Brain, 120 ( Pt 11):

149 140 BIBLIOGRAPHY Flowers, K. A. and Robertson, C. (1995). Perceptual abnormalities in parkinsons-disease - top-down or bottom-up processes. Perception, 24(10): Folstein, M. F., Folstein, S. E., and McHugh, P. R. (1975). mini-mental state. a practical method for grading the cognitive state of patients for the clinician. J.Psychiatr.Res., 12(3): Fox, P. T. and Raichle, M. E. (1986). Focal physiological uncoupling of cerebral blood flow and oxidative metabolism during somatosensory stimulation in human subjects. Proc.Natl.Acad.Sci.U.S.A, 83(4): Francis, P. T. and Perry, E. K. (2007). Cholinergic and other neurotransmitter mechanisms in parkinson s disease, parkinson s disease dementia, and dementia with lewy bodies. Mov Disord., 22 Suppl 17:S351 S357. Frankel, J. P., Lees, A. J., Kempster, P. A., and Stern, G. M. (1990). Subcutaneous apomorphine in the treatment of parkinson s disease. J.Neurol.Neurosurg.Psychiatry, 53(2): Frederick, J. M., Rayborn, M. E., Laties, A. M., Lam, D. M., and Hollyfield, J. G. (1982). Dopaminergic neurons in the human retina. J.Comp Neurol., 210(1): Friston, K. J., Fletcher, P., Josephs, O., Holmes, A., Rugg, M. D., and Turner, R. (1998). Event-related fmri: characterizing differential responses. Neuroimage., 7(1): Friston, K. J., Holmes, A. P., Poline, J. B., Grasby, P. J., Williams, S. C., Frackowiak, R. S., and Turner, R. (1995). Analysis of fmri time-series revisited. Neuroimage., 2(1): Geerligs, L., Meppelink, A. M., Brouwer, W. H., and van Laar, T. (2009). The effects of apomorphine on visual perception in patients with parkinson disease and visual hallucinations: a pilot study. Clin.Neuropharmacol., 32(5): Goetz, C. G., Pappert, E. J., Blasucci, L. M., Stebbins, G. T., Ling, Z. D., Nora, M. V., and Carvey, P. M. (1998). Intravenous levodopa in hallucinating parkinson s disease patients: high-dose challenge does not precipitate hallucinations. Neurology, 50(2): Goetz, C. G. and Stebbins, G. T. (1993). Risk factors for nursing home placement in advanced parkinson s disease. Neurology, 43(11):

150 BIBLIOGRAPHY 141 Goetz, C. G., Wuu, J., Curgian, L. M., and Leurgans, S. (2005). Hallucinations and sleep disorders in pd: six-year prospective longitudinal study. Neurology, 64(1): Goodale, M. A., Westwood, D. A., and Milner, A. D. (2004). Two distinct modes of control for object-directed action. Prog.Brain Res., 144: Grill-Spector, K. (2003). The neural basis of object perception. Curr.Opin.Neurobiol., 13(2): Grill-Spector, K., Kushnir, T., Hendler, T., and Malach, R. (2000). The dynamics of object-selective activation correlate with recognition performance in humans. Nat.Neurosci., 3(8): Groenewegen, H. J. (2009). The basal ganglia. Grossi, D., Trojano, L., Pellecchia, M. T., Amboni, M., Fragassi, N. A., and Barone, P. (2005). Frontal dysfunction contributes to the genesis of hallucinations in non-demented parkinsonian patients. Int.J.Geriatr.Psychiatry, 20(7): Hajee, M. E., March, W. F., Lazzaro, D. R., Wolintz, A. H., Shrier, E. M., Glazman, S., and Bodis-Wollner, I. G. (2009). Inner retinal layer thinning in parkinson disease. Arch.Ophthalmol., 127(6): Harding, A. J., Broe, G. A., and Halliday, G. M. (2002). Visual hallucinations in lewy body disease relate to lewy bodies in the temporal lobe. Brain, 125(Pt 2): Harnois, C. and Di Paolo, T. (1990). Decreased dopamine in the retinas of patients with parkinson s disease. Invest Ophthalmol.Vis.Sci., 31(11): Harris, J. P., Calvert, J. E., and Phillipson, O. T. (1992). Processing of spatial contrast in peripheral vision in parkinson s disease. Brain, 115 ( Pt 5): Hattori, N., Shibasaki, H., Wheaton, L., Wu, T., Matsuhashi, M., and Hallett, M. (2009). Discrete parieto-frontal functional connectivity related to grasping. J.Neurophysiol., 101(3): Hazlett, E. A., Buchsbaum, M. S., Kemether, E., Bloom, R., Platholi, J., Brickman, A. M., Shihabuddin, L., Tang, C., and Byne, W. (2004). Abnormal glucose metabolism in the mediodorsal nucleus of the thalamus in schizophrenia. Am.J.Psychiatry, 161(2):

151 142 BIBLIOGRAPHY Hendry, S. H. and Yoshioka, T. (1994). A neurochemically distinct third channel in the macaque dorsal lateral geniculate nucleus. Science, 264(5158): Hilker, R., Thomas, A. V., Klein, J. C., Weisenbach, S., Kalbe, E., Burghaus, L., Jacobs, A. H., Herholz, K., and Heiss, W. D. (2005). Dementia in parkinson disease: functional imaging of cholinergic and dopaminergic pathways. Neurology, 65(11): Holroyd, S., Currie, L., and Wooten, G. F. (2001). Prospective study of hallucinations and delusions in parkinson s disease. J.Neurol.Neurosurg.Psychiatry, 70(6): Hopfinger, J. B. and West, V. M. (2006). Interactions between endogenous and exogenous attention on cortical visual processing. Neuroimage., 31(2): Horowitz, T. S., Choi, W. Y., Horvitz, J. C., Cote, L. J., and Mangels, J. A. (2006). Visual search deficits in parkinson s disease are attenuated by bottom-up target salience and top-down information. Neuropsychologia, 44(10): Huang, C., Mattis, P., Tang, C., Perrine, K., Carbon, M., and Eidelberg, D. (2007). Metabolic brain networks associated with cognitive function in parkinson s disease. Neuroimage., 34(2): Huettel, S. A., Song, A. W., and McCarthy, G. (2004). Functional magnetic resonance imaging. Ibarretxe-Bilbao, N., Ramirez-Ruiz, B., Junque, C., Marti, M. J., Valldeoriola, F., Bargallo, N., Juanes, S., and Tolosa, E. (2009). Differential progression of brain atrophy in parkinson disease with and without visual hallucinations. J.Neurol.Neurosurg.Psychiatry. Imamura, K., Wada-Isoe, K., Kitayama, M., and Nakashima, K. (2007). Executive dysfunction in non-demented parkinson s disease patients with hallucinations. Acta Neurol.Scand. James, T. W., Humphrey, G. K., Gati, J. S., Menon, R. S., and Goodale, M. A. (2000). The effects of visual object priming on brain activation before and after recognition. Current Biology, 10(17): Jancke, L., Koeneke, S., Hoppe, A., Rominger, C., and Hanggi, J. (2009). The architecture of the golfer s brain. PLoS.One., 4(3).

152 BIBLIOGRAPHY 143 Jellinger, K. A. (2009a). A critical evaluation of current staging of alpha-synuclein pathology in lewy body disorders. Biochim.Biophys.Acta, 1792(7): Jellinger, K. A. (2009b). Formation and development of lewy pathology: a critical update. J.Neurol., 256 Suppl 3: Johannsen, P., Jakobsen, J., Bruhn, P., Hansen, S. B., Gee, A., Stodkilde- Jorgensen, H., and Gjedde, A. (1997). Cortical sites of sustained and divided attention in normal elderly humans. Neuroimage., 6(3): Kandel, E. (2000). Principles of Neural Science. McGraw-Hill. Kanwisher, N., McDermott, J., and Chun, M. M. (1997). The fusiform face area: a module in human extrastriate cortex specialized for face perception. J.Neurosci., 17(11): Kastner, S., De Weerd, P., Desimone, R., and Ungerleider, L. G. (1998). Mechanisms of directed attention in the human extrastriate cortex as revealed by functional mri. Science, 282(5386): Kastner, S., O Connor, D. H., Fukui, M. M., Fehd, H. M., Herwig, U., and Pinsk, M. A. (2004). Functional imaging of the human lateral geniculate nucleus and pulvinar. J.Neurophysiol., 91(1): Kataoka, H., Furiya, Y., Morikawa, M., Ueno, S., and Inoue, M. (2008). Increased temporal blood flow associated with visual hallucinations in parkinson s disease with dementia. Mov Disord., 23(3): Kempster, P. A., Williams, D. R., Selikhova, M., Holton, J., Revesz, T., and Lees, A. J. (2007). Patterns of levodopa response in parkinson s disease: a clinico-pathological study. Brain, 130(Pt 8): Klein, R. C., de Jong, B. M., de Vries, J. J., and Leenders, K. L. (2005). Direct comparison between regional cerebral metabolism in progressive supranuclear palsy and parkinson s disease. Mov Disord., 20(8): Kleinschmidt, A., Buchel, C., Hutton, C., Friston, K. J., and Frackowiak, R. S. J. (2002). The neural structures expressing perceptual hysteresis in visual letter recognition. Neuron, 34(4): Kobayashi, Y. and Isa, T. (2002). Sensory-motor gating and cognitive control by the brainstem cholinergic system. Neural Netw., 15(4-6):

153 144 BIBLIOGRAPHY Koerts, J., Borg, M. A., Meppelink, A. M., Leenders, K. L., van Beilen, M., and van Laar, T. (2010). Attentional and perceptual impairments in parkinson s disease with visual hallucinations. Parkinsonism.Relat Disord. Ku, J., Kim, J. J., Jung, Y. C., Park, I. H., Lee, H., Han, K., Yoon, K. J., Kim, I. Y., and Kim, S. I. (2008). Brain mechanisms involved in processing unreal perceptions. Neuroimage., 43(4): Kveraga, K., Boshyan, J., and Bar, M. (2007). Magnocellular projections as the trigger of top-down facilitation in recognition. J.Neurosci., 27(48): La Vega-Talbot, M., Duchowny, M., and Jayakar, P. (2006). Orbitofrontal seizures presenting with ictal visual hallucinations and interictal psychosis. Pediatr.Neurol., 35(1): Laatu, S., Revonsuo, A., Pihko, L., Portin, R., and Rinne, J. O. (2004). Visual object recognition deficits in early parkinson s disease. Parkinsonism.Relat Disord., 10(4): LaBerge, D. and Buchsbaum, M. S. (1990). Positron emission tomographic measurements of pulvinar activity during an attention task. J.Neurosci., 10(2): Lang, A. E. and Lozano, A. M. (1998). Parkinson s disease. first of two parts. N.Engl.J.Med., 339(15): Leh, S. E., Chakravarty, M. M., and Ptito, A. (2008). The connectivity of the human pulvinar: a diffusion tensor imaging tractography study. Int.J.Biomed.Imaging, Lieb, K., Brucker, S., Bach, M., Els, T., Lucking, C. H., and Greenlee, M. W. (1999). Impairment in preattentive visual processing in patients with parkinson s disease. Brain, 122 ( Pt 2): Litvinenko, I. V., Odinak, M. M., Mogil naya, V. I., and Perstnev, S. V. (2010). Use of memantine (akatinol) for the correction of cognitive impairments in parkinson s disease complicated by dementia. Neurosci.Behav.Physiol, 40(2): Liu, T., Pestilli, F., and Carrasco, M. (2005). Transient attention enhances perceptual performance and fmri response in human visual cortex. Neuron, 45(3):

154 BIBLIOGRAPHY 145 Lo, R. Y., Tanner, C. M., Albers, K. B., Leimpeter, A. D., Fross, R. D., Bernstein, A. L., McGuire, V., Quesenberry, C. P., Nelson, L. M., and Van Den Eeden, S. K. (2009). Clinical features in early parkinson disease and survival. Arch.Neurol., 66(11): Ma, Y., Tang, C., Spetsieris, P. G., Dhawan, V., and Eidelberg, D. (2007). Abnormal metabolic network activity in parkinson s disease: test-retest reproducibility. J.Cereb.Blood Flow Metab, 27(3): Malach, R., Reppas, J. B., Benson, R. R., Kwong, K. K., Jiang, H., Kennedy, W. A., Ledden, P. J., Brady, T. J., Rosen, B. R., and Tootell, R. B. (1995). Object-related activity revealed by functional magnetic resonance imaging in human occipital cortex. Proc.Natl.Acad.Sci.U.S.A, 92(18): Manford, M. and Andermann, F. (1998). Complex visual hallucinations. clinical and neurobiological insights. Brain, 121 ( Pt 10): Manganelli, F., Vitale, C., Santangelo, G., Pisciotta, C., Iodice, R., Cozzolino, A., Dubbioso, R., Picillo, M., Barone, P., and Santoro, L. (2009). Functional involvement of central cholinergic circuits and visual hallucinations in parkinson s disease. Brain, 132(Pt 9): Marinus, J., Visser, M., Stiggelbout, A. M., Rabey, J. M., Martinez-Martin, P., Bonuccelli, U., Kraus, P. H., and van Hilten, J. J. (2004). A short scale for the assessment of motor impairments and disabilities in parkinson s disease: the spes/scopa. J.Neurol.Neurosurg.Psychiatry, 75(3): Marinus, J., Visser, M., Verwey, N. A., Verhey, F. R., Middelkoop, H. A., Stiggelbout, A. M., and van Hilten, J. J. (2003). Assessment of cognition in parkinson s disease. Neurology, 61(9): Matsui, H., Nishinaka, K., Oda, M., Hara, N., Komatsu, K., Kubori, T., and Udaka, F. (2006a). Hypoperfusion of the visual pathway in parkinsonian patients with visual hallucinations. Mov Disord., 21(12): Matsui, H., Udaka, F., Tamura, A., Oda, M., Kubori, T., Nishinaka, K., and Kameyama, M. (2006b). Impaired visual acuity as a risk factor for visual hallucinations in parkinson s disease. J.Geriatr.Psychiatry Neurol., 19(1): Mayo, J. P. (2009). Intrathalamic mechanisms of visual attention. J.Neurophysiol., 101(3):

155 146 BIBLIOGRAPHY McAlonan, K., Cavanaugh, J., and Wurtz, R. H. (2008). Guarding the gateway to cortex with attention in visual thalamus. Nature, 456(7220): Meppelink, A. M., de Jong, B. M., Renken, R., Leenders, K. L., Cornelissen, F. W., and van Laar, T. (2009). Impaired visual processing preceding image recognition in parkinson s disease patients with visual hallucinations. Brain, 132(Pt 11): Meppelink, A. M., de Jong, B. M., van der Hoeven, J. H., and van Laar, T. (2010). Lasting visual hallucinations in visual deprivation; fmri correlates and the influence of rtms. J.Neurol.Neurosurg.Psychiatry. Meppelink, A. M., Koerts, J., Borg, M., Leenders, K. L., and van Laar, T. (2008). Visual object recognition and attention in parkinson s disease patients with visual hallucinations. Mov Disord. Merabet, L. B., Kobayashi, M., Barton, J., and Pascual-Leone, A. (2003). Suppression of complex visual hallucinatory experiences by occipital transcranial magnetic stimulation: a case report. Neurocase., 9(5): Merims, D., Shabtai, H., Korczyn, A. D., Peretz, C., Weizman, N., and Giladi, N. (2004). Antiparkinsonian medication is not a risk factor for the development of hallucinations in parkinson s disease. J.Neural Transm., 111(10-11): Mesulam, M. M. (2004). The cholinergic innervation of the human cerebral cortex. Prog.Brain Res., 145: Mesulam, M. M. and Geula, C. (1994). Chemoarchitectonics of axonal and perikaryal acetylcholinesterase along information processing systems of the human cerebral cortex. Brain Res.Bull., 33(2): Middleton, F. A. and Strick, P. L. (1996). The temporal lobe is a target of output from the basal ganglia. Proc.Natl.Acad.Sci.U.S.A, 93(16): Mink, J. W. (1996). The basal ganglia: focused selection and inhibition of competing motor programs. Prog.Neurobiol., 50(4): Morgante, L., Basile, G., Epifanio, A., Spina, E., Antonini, A., Stocchi, F., Di Rosa, E., Martino, G., Marconi, R., La Spina, P., Nicita-Mauro, V., and Di Rosa, A. E. (2004). Continuous apomorphine infusion (cai) and neuropsychiatric disorders in patients with advanced parkinson s disease: a follow-up of two years. Arch.Gerontol.Geriatr.Suppl, (9):

156 BIBLIOGRAPHY 147 Mosimann, U. P., Rowan, E. N., Partington, C. E., Collerton, D., Littlewood, E., O Brien, J. T., Burn, D. J., and McKeith, I. G. (2006). Characteristics of visual hallucinations in parkinson disease dementia and dementia with lewy bodies. American Journal of Geriatric Psychiatry, 14(2): Muller, T., Benz, S., and Przuntek, H. (2002). Apomorphine delays simple reaction time in parkinsonian patients. Parkinsonism.Relat Disord., 8(5): Muslimovic, D., Post, B., Speelman, J. D., and Schmand, B. (2005). Cognitive profile of patients with newly diagnosed parkinson disease. Neurology, 65(8): Nagano-Saito, A., Washimi, Y., Arahata, Y., Iwai, K., Kawatsu, S., Ito, K., Nakamura, A., Abe, Y., Yamada, T., Kato, T., and Kachi, T. (2004). Visual hallucination in parkinson s disease with fdg pet. Mov Disord., 19(7): Nagano-Saito, A., Washimi, Y., Arahata, Y., Kachi, T., Lerch, J. P., Evans, A. C., Dagher, A., and Ito, K. (2005). Cerebral atrophy and its relation to cognitive impairment in parkinson disease. Neurology, 64(2): Newell, F. N., Wallraven, C., and Huber, S. (2004). The role of characteristic motion in object categorization. J.Vis., 4(2): Oishi, N., Udaka, F., Kameyama, M., Sawamoto, N., Hashikawa, K., and Fukuyama, H. (2005). Regional cerebral blood flow in parkinson disease with nonpsychotic visual hallucinations. Neurology, 65(11): Okada, K., Suyama, N., Oguro, H., Yamaguchi, S., and Kobayashi, S. (1999). Medication-induced hallucination and cerebral blood flow in parkinson s disease. J.Neurol., 246(5): Parkkinen, L., Pirttila, T., and Alafuzoff, I. (2008). Applicability of current staging/categorization of alpha-synuclein pathology and their clinical relevance. Acta Neuropathol., 115(4): Pauling, L. and Coryell, C. D. (1936). The magnetic properties and structure of hemoglobin, oxyhemoglobin and carbonmonoxyhemoglobin. Proc.Natl.Acad.Sci.U.S.A, 22(4): Pelli, D. G. (1997). The videotoolbox software for visual psychophysics: transforming numbers into movies. Spat.Vis., 10(4):

157 148 BIBLIOGRAPHY Pereira, J. B., Junque, C., Marti, M. J., Ramirez-Ruiz, B., Bargallo, N., and Tolosa, E. (2009). Neuroanatomical substrate of visuospatial and visuoperceptual impairment in parkinson s disease. Mov Disord., 24(8): Perry, E. K., Curtis, M., Dick, D. J., Candy, J. M., Atack, J. R., Bloxham, C. A., Blessed, G., Fairbairn, A., Tomlinson, B. E., and Perry, R. H. (1985). Cholinergic correlates of cognitive impairment in parkinson s disease: comparisons with alzheimer s disease. J.Neurol.Neurosurg.Psychiatry, 48(5): Perry, E. K. and Perry, R. H. (1995). Acetylcholine and hallucinations: diseaserelated compared to drug-induced alterations in human consciousness. Brain Cogn, 28(3): Pieri, V., Diederich, N. J., Raman, R., and Goetz, C. G. (2000). Decreased color discrimination and contrast sensitivity in parkinson s disease. J.Neurol.Sci., 172(1):7 11. Pitzalis, S., Sereno, M. I., Committeri, G., Fattori, P., Galati, G., Patria, F., and Galletti, C. (2010). Human v6: the medial motion area. Cereb.Cortex, 20(2): Playford, E. D., Jenkins, I. H., Passingham, R. E., Nutt, J., Frackowiak, R. S., and Brooks, D. J. (1992). Impaired mesial frontal and putamen activation in parkinson s disease: a positron emission tomography study. Ann.Neurol., 32(2): Quene, H. and van den Bergh, H. (2004). On multi-level modeling of data from repeated measures designs: A tutorial. Speech Communication, 43: Ramirez-Ruiz, B., Junque, C., Marti, M. J., Valldeoriola, F., and Tolosa, E. (2006). Neuropsychological deficits in parkinson s disease patients with visual hallucinations. Mov Disord., 21(9): Ramirez-Ruiz, B., Junque, C., Marti, M. J., Valldeoriola, F., and Tolosa, E. (2007a). Cognitive changes in parkinson s disease patients with visual hallucinations. Dement.Geriatr.Cogn Disord., 23(5): Ramirez-Ruiz, B., Marti, M. J., Tolosa, E., Falcon, C., Bargallo, N., Valldeoriola, F., and Junque, C. (2008). Brain response to complex visual stimuli in parkinson s patients with hallucinations: A functional magnetic resonance imaging study. Mov Disord.

158 BIBLIOGRAPHY 149 Ramirez-Ruiz, B., Marti, M. J., Tolosa, E., Gimenez, M., Bargallo, N., Valldeoriola, F., and Junque, C. (2007b). Cerebral atrophy in parkinson s disease patients with visual hallucinations. Eur.J.Neurol., 14(7): Redgrave, P., Prescott, T. J., and Gurney, K. (1999). The basal ganglia: a vertebrate solution to the selection problem? Neuroscience, 89(4): Reinders, A. A. T. S., Glascher, J., de Jong, J. R., Willemsen, A. T. M., den Boer, J. A., and Buchel, C. (2006). Detecting fearful and neutral faces: Bold latency differences in amygdala-hippocampal junction. Neuroimage, 33(2): Robinson, D. L. and Petersen, S. E. (1992). The pulvinar and visual salience. Trends Neurosci., 15(4): Roselli, F., Pisciotta, N. M., Perneczky, R., Pennelli, M., Aniello, M. S., De Caro, M. F., Ferrannini, E., Tartaglione, B., Defazio, G., Rubini, G., and Livrea, P. (2009). Severity of neuropsychiatric symptoms and dopamine transporter levels in dementia with lewy bodies: a 123i-fp-cit spect study. Mov Disord., 24(14): Rye, D. B. (1997). Contributions of the pedunculopontine region to normal and altered rem sleep. Sleep, 20(9): Saalmann, Y. B. and Kastner, S. (2009). Gain control in the visual thalamus during perception and cognition. Curr.Opin.Neurobiol., 19(4): Sanchez-Ramos, J. R., Ortoll, R., and Paulson, G. W. (1996). Visual hallucinations associated with parkinson disease. Arch.Neurol., 53(12): Santangelo, G., Trojano, L., Vitale, C., Ianniciello, M., Amboni, M., Grossi, D., and Barone, P. (2007). A neuropsychological longitudinal study in parkinson s patients with and without hallucinations. Mov Disord., 22(16): Sarter, M., Givens, B., and Bruno, J. P. (2001). The cognitive neuroscience of sustained attention: where top-down meets bottom-up. Brain Res.Brain Res.Rev., 35(2): Schapira, A. H., Mann, V. M., Cooper, J. M., Krige, D., Jenner, P. J., and Marsden, C. D. (1992). Mitochondrial function in parkinson s disease. the royal kings and queens parkinson s disease research group. Ann.Neurol., 32 Suppl:S116 S124.

159 150 BIBLIOGRAPHY Schmeichel, A. M., Buchhalter, L. C., Low, P. A., Parisi, J. E., Boeve, B. W., Sandroni, P., and Benarroch, E. E. (2008). Mesopontine cholinergic neuron involvement in lewy body dementia and multiple system atrophy. Neurology, 70(5): Schneider, E., Fischer, P. A., Jacobi, P., Becker, H., and Hacker, H. (1979). The significance of cerebral atrophy for the symptomatology of parkinson s disease. J.Neurol.Sci., 42(2): Schotte, A., Janssen, P. F., Megens, A. A., and Leysen, J. E. (1993). Occupancy of central neurotransmitter receptors by risperidone, clozapine and haloperidol, measured ex vivo by quantitative autoradiography. Brain Res., 631(2): Schufried, G. and Mîdling (1992). Test Management Program: Reaction Test -S7. Seger, C. A. and Cincotta, C. M. (2005). The roles of the caudate nucleus in human classification learning. J.Neurosci., 25(11): Selby, G. (1968). Cerebral atrophy in parkinsonism. J.Neurol.Sci., 6(3): Serences, J. T. and Yantis, S. (2006). Selective visual attention and perceptual coherence. Trends Cogn Sci., 10(1): Shimada, H., Hirano, S., Shinotoh, H., Aotsuka, A., Sato, K., Tanaka, N., Ota, T., Asahina, M., Fukushi, K., Kuwabara, S., Hattori, T., Suhara, T., and Irie, T. (2009). Mapping of brain acetylcholinesterase alterations in lewy body disease by pet. Neurology. Silbersweig, D. A., Stern, E., Frith, C., Cahill, C., Holmes, A., Grootoonk, S., Seaward, J., McKenna, P., Chua, S. E., Schnorr, L., and. (1995). A functional neuroanatomy of hallucinations in schizophrenia. Nature, 378(6553): Silva, M. F., Faria, P., Regateiro, F. S., Forjaz, V., Januario, C., Freire, A., and Castelo-Branco, M. (2005). Independent patterns of damage within magno-, parvo- and koniocellular pathways in parkinson s disease. Brain, 128(Pt 10): Simons, J. S., Henson, R. N., Gilbert, S. J., and Fletcher, P. C. (2008). Separable forms of reality monitoring supported by anterior prefrontal cortex. J.Cogn Neurosci., 20(3):

160 BIBLIOGRAPHY 151 Sincich, L. C., Park, K. F., Wohlgemuth, M. J., and Horton, J. C. (2004). Bypassing v1: a direct geniculate input to area mt. Nat.Neurosci., 7(10): Smith, A. T., Cotton, P. L., Bruno, A., and Moutsiana, C. (2009). Dissociating vision and visual attention in the human pulvinar. J.Neurophysiol., 101(2): Snow, J. C., Allen, H. A., Rafal, R. D., and Humphreys, G. W. (2009). Impaired attentional selection following lesions to human pulvinar: evidence for homology between human and monkey. Proc.Natl.Acad.Sci.U.S.A, 106(10): Sossi, V., Fuente-Fernandez, R., Schulzer, M., Troiano, A. R., Ruth, T. J., and Stoessl, A. J. (2007). Dopamine transporter relation to dopamine turnover in parkinson s disease: a positron emission tomography study. Ann.Neurol., 62(5): Stebbins, G. T., Goetz, C. G., Carrillo, M. C., Bangen, K. J., Turner, D. A., Glover, G. H., and Gabrieli, J. D. (2004). Altered cortical visual processing in pd with hallucinations: an fmri study. Neurology, 63(8): Steriade, M. (2004). Acetylcholine systems and rhythmic activities during the waking sleep cycle. Prog.Brain Res., 145: Summerfield, C., Junque, C., Tolosa, E., Salgado-Pineda, P., Gomez-Anson, B., Marti, M. J., Pastor, P., Ramirez-Ruiz, B., and Mercader, J. (2005). Structural brain changes in parkinson disease with dementia: a voxel-based morphometry study. Arch.Neurol., 62(2): Summerfield, C. and Koechlin, E. (2008). A neural representation of prior information during perceptual inference. Neuron, 59(2): Tagliati, M., Bodis-Wollner, I., and Yahr, M. D. (1996). The pattern electroretinogram in parkinson s disease reveals lack of retinal spatial tuning. Electroencephalogr.Clin.Neurophysiol., 100(1):1 11. Tan, E. K. and Skipper, L. M. (2007). disease. Hum.Mutat., 28(7): Pathogenic mutations in parkinson Teunisse, R. J., Cruysberg, J. R., Hoefnagels, W. H., Verbeek, A. L., and Zitman, F. G. (1996). Visual hallucinations in psychologically normal people: Charles bonnet s syndrome. Lancet, 347(9004):

161 152 BIBLIOGRAPHY Thannickal, T. C., Lai, Y. Y., and Siegel, J. M. (2007). Hypocretin (orexin) cell loss in parkinson s disease. Brain, 130(Pt 6): Thannickal, T. C., Moore, R. Y., Nienhuis, R., Ramanathan, L., Gulyani, S., Aldrich, M., Cornford, M., and Siegel, J. M. (2000). Reduced number of hypocretin neurons in human narcolepsy. Neuron, 27(3): Thomas, C., Moya, L., Avidan, G., Humphreys, K., Jung, K. J., Peterson, M. A., and Behrmann, M. (2008). Reduction in white matter connectivity, revealed by diffusion tensor imaging, may account for age-related changes in face perception. J.Cogn Neurosci., 20(2): Tir, M., Delmaire, C., le Thuc, V., Duhamel, A., Destee, A., Pruvo, J. P., and Defebvre, L. (2009). Motor-related circuit dysfunction in msa-p: Usefulness of combined whole-brain imaging analysis. Mov Disord., 24(6): Treue, S. (2003). Climbing the cortical ladder from sensation to perception. Trends Cogn Sci., 7(11): Ungerleider, L. G. and Haxby, J. V. (1994). what and where in the human brain. Curr.Opin.Neurobiol., 4(2): Verleden, S., Vingerhoets, G., and Santens, P. (2007). Heterogeneity of cognitive dysfunction in parkinson s disease: a cohort study. Eur.Neurol., 58(1): von Campenhausen, S., Bornschein, B., Wick, R., Botzel, K., Sampaio, C., Poewe, W., Oertel, W., Siebert, U., Berger, K., and Dodel, R. (2005). Prevalence and incidence of parkinson s disease in europe. Eur.Neuropsychopharmacol., 15(4): Walker, M. P., Ayre, G. A., Cummings, J. L., Wesnes, K., McKeith, I. G., O Brien, J. T., and Ballard, C. G. (2000). Quantifying fluctuation in dementia with lewy bodies, alzheimer s disease, and vascular dementia. Neurology, 54(8): Warrington (1991). The visual object and space perception battery. Thames Valley Test Company, Bury St. Edmonds UK. Wesnes, K. A., McKeith, I., Edgar, C., Emre, M., and Lane, R. (2005). Benefits of rivastigmine on attention in dementia associated with parkinson disease. Neurology, 65(10):

162 BIBLIOGRAPHY 153 Westerink, B. H. (2002). Can antipsychotic drugs be classified by their effects on a particular group of dopamine neurons in the brain? Eur.J.Pharmacol., 455(1):1 18. Wheaton, L. A. and Hallett, M. (2007). Ideomotor apraxia: a review. J.Neurol.Sci., 260(1-2):1 10. Williams, D. R. and Lees, A. J. (2005). Visual hallucinations in the diagnosis of idiopathic parkinson s disease: a retrospective autopsy study. Lancet Neurol., 4(10): Wolters, E., van Laar, T., and Berendse, H. (2007). Parkinsonism and related disorders. Zeki, S. (2001). Localization and globalization in conscious vision. Annu.Rev.Neurosci., 24: Zweig, R. M., Jankel, W. R., Hedreen, J. C., Mayeux, R., and Price, D. L. (1989). The pedunculopontine nucleus in parkinson s disease. Ann.Neurol., 26(1):41 46.

163

164 Dankbrein 155

165 156 DANKBREIN Figure 1: Dankbrein. In deze figuur wil ik graag een aantal mensen bedanken, die een directe of indirecte bijdrage hebben geleverd aan de totstandkoming van dit proefschrift. Zie legenda hiernaast voor uitleg over de functies van de verschillende hersengebieden waarin ik mensen heb geplaatst.

166 157