Small vessel disease and the resting functional architecture of the brain

Abstract

Small vessel disease (SVD) is linked to cognitive impairment and dementia, yet little is known regarding functional activation in patients with SVD. Resting fMRI recordings suggest reduced connectivity in prefrontal, parietal and cingulate nodes and reciprocally increased connectivity in cerebellum, alterations which predicted neuropsychological test performance. Together with diffusion tensor tensor imaging studies, these data support of a model of disrupted connectivity as a systems-level approach to the cognitive disturbances seen in SVD.

Small vessel disease (SVD) denotes a spectrum of lesions of vascular etiology, which localize to subcortical structures and are associated with clinical phenotypes that include gait disturbance, psychomotor slowing, cognitive impairment, dementia, depression, and a significantly increased risk of stroke. Recent research has underscored the degree to which SVD and neurodegenerative processes overlap and work synergistically in producing the pathology associated with dementia; by some estimates defects in vascular function contribute directly or indirectly to more than half of cases of dementia.^{1, 2} Atheroma and thromboembolic phenomena involving the cerebral microvasculature, as well as endothelial and blood–brain barrier changes are recognized as important mechanisms underlying SVD, yet many aspects of SVD pathobiology remain unknown.^{2, 3}

In vivo knowledge of SVD has developed largely as a result of progress in neuroimaging, which has allowed a detailed characterization and classification of anatomical and physiological changes associated with this condition.⁴ As an example, recent studies using diffusion tensor imaging show an association between SVD and microstructural changes within selected white matter tracts,^{5, 6} changes which have been linked to gait disturbances,⁷ decreased performance on cognitive testing,⁸ and may precede the development of white matter lesions detected using conventional sequences such as T2 or FLAIR.⁹ Despite such advances, a systems-level understanding of how SVD contributes to cognitive dysfunction is missing.

In this issue of the Journal, Schaefer et al.¹⁰ report on patients with early SVD who underwent resting state functional magnetic resonance imaging (fMRI). With the help of eigenvector centrality analysis, a measure used in graph theory and network analysis,¹¹ the authors found reduced connectivity of nodes located in prefrontal, parietal and cingulate cortices, and concurrent connectivity increases in cerebellar structures when comparing SVD patients with age matched controls. Associations were also identified between regional connectivity measures and morphological evidence of SVD, and also with neuropsychological test scores across a range of cognitive domains.

These results were obtained in a comparatively small patient sample, and longitudinal data were missing to understand how the reported functional–anatomical–neuropsychological associations might be viewed in terms of causal relationships and not simply correlations. Importantly, it is not possible to discern from these data the degree to which SVD-associated connectivity changes are direct consequences of the underlying disease, indirect compensatory or adaptive mechanisms, or a combination of both. The study raises fundamental questions of inference, i.e., what is the validity and reliability of fMRI in a population with SVD? Interpretation of the fMRI (BOLD) signal rests on the assumption of preserved neurovascular coupling, and it is possible that the observed group differences or correlations with neuropsychological tests are more reflective of alterations in vascular physiology rather than underlying differences in neural activation. Moreover, there is evidence of intrinsic microvascular oscillations in the aging and microangiopathic brain;¹² potential confounding of neurovascular coupling and connectivity measures by such oscillations have not been accounted for. Notwithstanding, the study contributes to mounting evidence that changes in connectivity within discrete cortical and subcortical hubs represent functional signatures of impaired cognition. These observations are consistent with studies of SVD made using diffusion tensor imaging, and support a paradigm of disconnectivity and hub reorganization as a unifying principle in a range of disabling neurological and psychiatric disorders including Alzheimer disease,¹³ multiple sclerosis,¹⁴ traumatic brain injury,¹⁵ and schizophrenia.¹⁶

The study by Schaefer et al.¹⁰ elicits important questions which might be evaluated in ongoing or future studies. How can we account for BOLD signal changes that are mediated by vascular pathology and are independent of neural activation? Assuming we can, where do changes in hub centrality and functional circuit connectivity stand in the temporal sequence linking vascular risk factors, morphological evidence of SVD, and subtle or overt clinical symptoms? Could functional connectivity changes serve as markers to predict and track disease progression and the potential impact of preventative strategies (e.g., control of hypertension) or therapeutic interventions (e.g., neuromodulatory interventions?). Can we disambiguate functional and anatomical markers of brain reorganization and plasticity from direct lesion-associated changes? Are the reported cerebellar increases in centrality noted in SVD patients an example of this reorganization, consistent with known networks linking cerebellum with hemispheric (primarily prefrontal) association areas?¹⁷ Answers to these and related questions will represent significant advances in the biology and treatment of the aging brain.

The authors declare no conflict of interest.