Exploring the role of white matter disconnection in the heterogeneity of Parkinson's disease

Idiopathic Parkinson’s disease is a neurodegenerative condition that presents with a myriad of motor and non-motor symptoms. The complex and heterogeneous character of Parkinson’s disease suggests that it is a multisystem disorder with widespread pathology affecting different anatomical structures. In addition to localized dysfunction, changes in connections between brain regions and networks may contribute to Parkinson’s symptomology. This thesis explores this hypothesis in two prevalent symptoms of Parkinson’s disease: freezing of gait and visual hallucinations. Freezing of gait is an involuntarily paroxysmal breakdown of an individual’s footstep pattern whilst walking. Freezing of gait can be triggered by environmental factors or additional cognitive load. Furthermore, freezing behaviour can also be observed in nongait tasks, such as during speech or handwriting. These observations suggest that freezing cannot solely be attributed to a motor dysfunction. This thesis provides support for the notion that freezing of gait results from ineffective communication between distinct brain regions. Specifically, this thesis showed that freezing of gait is associated with deficits in executive functioning, mood and sleep disturbances as well as with the non-tremor dominant motor phenotype, confirming its multi-system nature. To investigate the underlying mechanisms of this multi-system symptom, the structural network topology in patients with freezing of gait was assessed. Importantly, the brain displays a modular structure and changes in the diversity of inter-modular connections (between-module connectivity) and the connections of a node within its own module (within-module connectivity) can inform about the integration and segregation of specialized communities. This thesis showed that patients with freezing of gait showed reduced between-module connectivity in subcortical and frontoparietal nodes. Several of these nodes were found within the brain’s ‘rich club’, a group of highly interconnected nodes that form the structural backbone of the connectome and, therefore, play an important role in global information integration between different modules in the brain. Freezing of gait could thus arise due to ineffective communication between major cortical and subcortical integration centres. Furthermore, changes in within-module scores were observed across different sensory processing modalities. This may reflect a compensatory strategy and could indicate that compared to non-freezers, freezers rely more heavily on exogenous stimuli to guide their actions, perhaps due to a loss of automaticity. Visual hallucinations represent another common, yet poorly understood symptom of Parkinson’s disease. The pathophysiology of visual hallucinations is likely multi-faceted with changes across attentional and perceptual brain networks. First, we investigated attentional processing capability in patients with visual hallucinations using a computerized flanker task. With increasing task demands, Parkinson’s disease patients with visual hallucinations showed lower accuracy rates than patients without this symptom, yet the reaction times were similar between groups for correct trials. These findings suggest that hallucinators have specific impairments in sustained attention and conflict monitoring. Furthermore, the episodic nature suggests that visual hallucinations may arise during periods of ineffective integration of information. To investigate whether the severity of hallucinatory behaviour was associated with ability to successfully integrate information, changes in structural network topology were assessed. Severity of hallucinatory behaviour was associated with reduced connectivity across a structural bilateral sub-network. Nodes within this network showed higher between- and within-module scores compared to nodes outside this network. This may indicate that the system requires to reroute information across less efficient pathways, affecting the sensory integration process. Furthermore, severity of hallucinations was associated with changes in between-module scores across top-down visual processing centres and attentional networks. Thus, impaired integration across visual processing sites may result in the inefficient transfer of information that gives rise to visual hallucinations in Parkinson’s disease. The findings presented in this thesis suggest that freezing of gait and visual hallucinations may be the result of alterations in the brain’s information processing capability as indicated by reduced interactions between brain networks. By investigating the inherent complexity of the brain’s interacting subsystems and their connections, we can advance our understanding of neurodegenerative disorders and ultimately, aid the developments of new and improved treatment strategies