References for this Review were identified through searches of PubMed with the search terms “network”, “network dysfunction”, “connectivity”, “resting state functional MRI”, “electroencephalography”, “magnetoencephalography”, “diffusion tensor imaging”, “tractography”, “dementia”, “neurodegenerative disorders”, “frontotemporal dementia”, “Alzheimer”, “mild cognitive impairment”, “Parkinson”, “Lewy bodies dementia”, “stroke”, and “tumour” from 1986 until June, 2011. Furthermore, we identified
ReviewFunctional network disruption in the degenerative dementias
Introduction
Historically, clinicians have identified patients with neurodegenerative dementias on the basis of their clinical symptoms. In recent years, advances in basic science have allowed researchers to recategorise these diseases on the basis of molecular phenotype—ie, the toxic, misfolded disease protein aggregates that are identified in the brain post mortem, such as amyloid β (Aβ) and hyperphosphorylated tau in Alzheimer's disease; tau, TAR DNA-binding protein of 43 kDa (TDP-43), or fused in sarcoma (FUS) in frontotemporal dementia; and α-synuclein in Parkinson's disease and dementia with Lewy bodies.1 These pathological changes are thought to be early events in a cascade that begins at the synaptic and neuronal levels and ultimately leads to the clinical syndrome. Within this temporal window, quantifiable biological, imaging, and physiological markers of pathology have been identified that can be thought of as in-vivo intermediate phenotypes. Such surrogate markers of pathology can improve understanding of disease pathophysiology—ie, indicate links between the molecular phenotype and clinical symptoms—and have the potential to allow earlier, more accurate diagnosis and monitoring of disease progression. In Alzheimer's disease, PET amyloid ligands enable in-vivo mapping of cerebral Aβ deposition,2 whereas structural MRI findings have been shown to relate to hyperphosphorylated-tau-mediated neurodegeneration.3 These biomarkers have recently been incorporated into the new diagnostic criteria for Alzheimer's disease.4, 5 In disorders such as Parkinson's disease, frontotemporal dementia, and dementia with Lewy bodies, structural biomarkers have been used to elucidate disease pathophysiology by showing patterns of atrophy associated with histopathology on the one hand,6, 7, 8 and clinical symptoms on the other (table 1).8, 9
Localisation-based approaches (such as in-vivo mapping of molecular changes and neurodegeneration) have helped build much of the present knowledge of disease pathophysiology. However, these approaches are less suited to investigation of neuronal or synaptic dysfunction, which is thought to underlie cognitive and functional deficits. Because brain functions rely on the integrity of dynamic communication between interconnected brain regions and circuits, a network perspective accounting for such interactions has the potential to provide new and meaningful intermediate phenotypes of pathology (table 1). Prevalent views on the relation between symptoms and pathological changes in Alzheimer's disease help illustrate this notion (figure 1). In typical Alzheimer's disease, the progression of symptoms occurs in a stereotyped order that relates to the topographic progression of hyperphosphorylated tau:10 episodic memory loss takes place first (hippocampus and medial temporal lobe, and posterior cingulate cortex), followed by semantic memory loss (lateral temporal cortex) and aphasic, apraxic, and visuospatial symptoms (frontal, temporal, and parietal neocortex), and finally by motor and visual deficits (sensorimotor and occipital cortex). Although atypical variants exist,11 this orderly progression might be indicative of an incremental spread throughout interconnected regions within large-scale networks, and, ultimately, a spread into adjacent or upstream regions.
The brain can be thought of as a complex neural network consisting of structurally and functionally interconnected regions at many scales (panel 1).12 At the macroscopic level, neural networks can be assessed non-invasively in health and disease with functional MRI (fMRI) and neurophysiological techniques (electroencephalography [EEG] and magnetoencephalography [MEG]).13, 14 The aim of our Review is to provide a comprehensive overview of findings on functional network disruption in the most prevalent neurodegenerative dementias. Although several reviews have addressed functional network disruption in Alzheimer's disease and in psychiatric disorders,15, 16, 17, 18, 19, 20 we summarise studies across many neurodegenerative dementias. By including frontotemporal dementia, dementia related to Parkinson's disease, and dementia with Lewy bodies, we highlight functional network similarities and differences in disorders that share common mechanisms (toxic protein aggregation and neuronal loss) but have distinct clinical phenotypes. Towards this aim, we review resting-state task-free functional imaging and neurophysiological studies. Because our primary goal is to review functional methods that are broadly applicable across neurodegenerative diseases, we have omitted task-activation studies, which require the design of disease-specific experiments (see Dickerson, 2007, for a review of task-activation studies in Alzheimer's disease21), as well as studies of grey-matter structural covariance.22, 23
Section snippets
Techniques to assess network integrity
fMRI, EEG, and MEG techniques enable researchers to assess large-scale neural networks at different spatial and temporal resolutions. Functional connectivity between brain regions can be measured at a spatial resolution of 2–3 mm with fMRI and about 5–30 mm with EEG or MEG. fMRI and neurophysiological techniques contrast most sharply in their temporal resolutions, which differ by three orders of magnitude (seconds vs milliseconds). Structural connectivity within networks can be measured at a
Functional networks and clinical impairment
Imaging and lesion studies have provided valuable information on the functional anatomy of the brain, and localisation principles are vital to the clinical neurologist. However, as we outlined in our introduction, localisation-based perspectives often do not explain the complex inter-relation between neurodegenerative pathological changes and clinical symptoms. Even focal lesions such as stroke (eg, strategic infarction), brain tumour, or traumatic brain injury can cause widespread disturbance
Conclusions and future directions
Brain connectivity studies allow questions to be addressed that have so far escaped a convincing answer. For example, what is the mechanism whereby, in Alzheimer's disease, the deposition of Aβ and hyperphosphorylated tau takes place in largely distinct but highly interconnected hub regions? Why does damage spread to the whole network? Similar questions apply to α-synuclein in dementia with Lewy bodies and tau, TDP-43, and FUS in frontotemporal dementia. Several working models for network-based
Search strategy and selection criteria
References (145)
- et al.
The diagnosis of dementia due to Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease
Alzheimers Dement
(2011) - et al.
The diagnosis of mild cognitive impairment due to Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease
Alzheimers Dement
(2011) - et al.
Voxel-based morphometry and its application to movement disorders
Parkinsonism Relat Disord
(2007) Use of magnetoencephalography (MEG) to study functional brain networks in neurodegenerative disorders
J Neurol Sci
(2010)- et al.
Assessing neuronal networks: understanding Alzheimer's disease
Prog Neurobiol
(2009) Advances in functional magnetic resonance imaging: technology and clinical applications
Neurotherapeutics
(2007)- et al.
Connectivity-behavior analysis reveals that functional connectivity between left BA39 and Broca's area varies with reading ability
Neuroimage
(2006) - et al.
EEG coherency: I—Statistics, reference electrode, volume conduction, Laplacians, cortical imaging, and interpretation at multiple scales
Electroencephalogr Clin Neurophysiol
(1997) - et al.
A diffusion tensor imaging tractography atlas for virtual in vivo dissections
Cortex
(2008) - et al.
The application of graph theoretical analysis to complex networks in the brain
Clin Neurophysiol
(2007)