|dc.description.abstract||Sporadic human cerebral small vessel disease (SVD) is a significant problem in our aging population. It is the most common cause of haemorrhagic stroke, causes one quarter of all ischaemic strokes, causes vascular dementia and synergistically worsens other dementias. It also causes a range of other psychological and physical problems and is increasingly common with increased age. SVD is well characterised on neuroimaging, with lesions throughout the brain, and particularly in the white matter. The cause(s) and pathophysiological mechanisms underlying the development of SVD, however, remain poorly understood with poor correlations between the abnormalities seen at the cellular level, on neuroimaging, and clinically.
There are many theories as to the cause of SVD. However, observational studies and experimental evidence point towards an abnormality in the small vasculature of the brain initiating a cascade of events leading to a variety of vascular and brain parenchymal lesions. What these abnormalities are, and how exactly they result in the pathology seen, is unknown. Different structural components of the vessel wall and parenchymal brain cells appear to be involved, as well as functional abnormalities such as abnormal vascular reactivity. Risk factors also play a role, hypertension being the most significant, but how these interact with the normal vasculature is not fully understood.
To provide an overview of our current understanding of SVD in human tissue I first completed a systematic review of the literature comparing the appearances of SVD on post-mortem imaging and histology. This revealed the inconsistency in methods and reporting in these studies and the lack of histopathology agreement on SVD terminology and definitions.
I then studied the histological appearances of the lesions identified by post-mortem imaging to provide a reliable precise histological-imaging correlation. I developed a new protocol for ex vivo 7 Tesla magnetic resonance imaging (7T MRI) scanning of human brain tissue on post-mortem material and developed a grading system to assess SVD burden on MRI and histology with histological definitions, to try to encourage standardised, comprehensive and transparent reporting so that results in small studies can be more easily compared. I studied human post-mortem brain tissue to better investigate the disease in the appropriate context.
In our cases from individuals with haemorrhagic SVD, normal aging and young controls, the most severe SVD pathology on ex vivo imaging and histology was, as expected, in the haemorrhagic SVD group. The normal aging group also had significant levels of pathology, perhaps representing the increasing burden of disease present but not necessarily detected clinically with increased age. It is possible the underlying pathophysiology in this group might develop by different mechanisms compared to the haemorrhagic group. Directly comparing the imaging and histological lesions confirmed the histological appearances of some lesions on imaging such as enlarged perivascular spaces, lacunes, microinfarcts and microbleeds. However, making direct comparisons is complex. Some lesions, such as small vessel fibrinoid necrosis, presumed to be below the resolution for detection on 7T MRI, were identified on both histology and imaging. Some features seen on histology in association with recognised SVD lesions, such as perivascular inflammation in an area of white matter rarefaction, were present in a variety of different histological contexts with no apparent correlation on imaging. And some lesions, such as white matter rarefaction around enlarged perivascular spaces, were present often on both imaging and histology, but their significance and contribution to SVD is unknown.
To try to further understand the mechanisms underlying SVD and the lesions seen on imaging I undertook biochemical studies of protein expression in the deep white matter of the haemorrhagic SVD group, young controls and an Alzheimer’s disease group, who also have white matter pathology on neuroimaging. Increased fibrinogen levels suggested vascular leakage in both disease groups. However, haemorrhagic SVD had more severe white matter hypoxic changes and increased vasoconstrictor levels while in Alzheimer’s disease there was increased amyloid 42 and levels of a pericyte marker, possibly reflecting different pathophysiological mechanisms causing the similar appearing radiological changes.
When assessing radiologically defined white matter hyperintensities (WMH) I found hypoxic-induced changes throughout brains with WMH, including in normal appearing areas of white matter. This suggests these brains have abnormalities in areas that appear radiologically normal, as found in in vivo imaging studies.
To conclude, this work has confirmed the importance of reaching a consensus in histopathological reporting, terminology and definitions which is a basic requirement before we can better understand the pathophysiology of SVD. This has led to the formation of definitions and a practical grading system that could be used as a basis upon which to build a future agreement. The complexity of the histological lesions underlying radiological SVD changes was apparent, and the frequency with which some other potentially important histological changes were identified suggests these have not, to date, been fully appreciated. Investigating the underlying mechanisms of white matter hyperintensities showed vascular leakage was a shared abnormality in two different diseases with white matter changes on imaging, suggesting it may be a common factor upon which variable pathways converge. Future work is needed to further understand the importance of these less well characterised histological features. Investigating the role of vascular leakage and exploring drugs that maintain or improve vascular integrity could be a potential route for helping to treat SVD. Studies into underlying transcriptomic abnormalities around vascular leakage in human tissue may be informative.||en