Search

Large vessel vasculitis (LVV), including giant cell arteritis (GCA) and Takayasu arteritis (TAK), is the most common form of primary vasculitis and is characterised by chronic inflammation of medium and large arteries. Accurately determining disease activity in LVV is challenging, particularly once treatment has started. This often leads to a mismatch between disease activity and treatment intensity leaving patients at risk of the consequences of over-treatment, including adverse effects of toxic therapies, or under-treatment, leading to unchecked vascular inflammation and arterial damage. The aim of this thesis was to develop novel methods of assessing disease activity in order to better guide treatment for patients with LVV. The results presented here demonstrate that hybrid positron emission tomography with magnetic resonance imaging (PET/MR) is able to accurately distinguish active from inactive LVV and can track disease activity longitudinally. MR metrics provide complementary information to that provided by PET metrics for determination of disease activity status. A novel, PET/MR-specific disease activity assessment score - the VAMP score - appears more effective at distinguishing active from inactive LVV than established PET-only scoring systems. The imaging data provided by PET/MR have the potential to positively influence 'real-world' clinical decision making. Considering that PET/MR utilises a significantly lower radiation dose compared with PET/CT, these results support the use of PET/MR for longitudinal disease monitoring in LVV. This thesis has also demonstrated the efficacy of a series of new and emerging serological biomarkers of disease activity in LVV. Plasma concentrations of leucine-rich alpha-2- glycoprotein-1 (LRG1), angiopoietin-2 (Ang-2), soluble fms-like tyrosine kinase-1 (sFlt-1), osteopontin, and calprotectin are elevated in LVV compared with health. Additionally, LRG1, Ang-2 and osteopontin can distinguish active LVV from inactive LVV, with all biomarkers outperforming C-reactive protein (CRP) for this indication. Correlations were observed between several of the evaluated biomarkers and previously defined measures of disease activity. Finally, I have demonstrated the ability of microcirculatory changes to inform LVV disease activity based on examination of retinal microstructures using optical coherence tomography (OCT). Patients with LVV exhibit chorioretinal thinning which is reversible with treatment. Correlations were also observed between OCT metrics and both PET/MR-based and serological measures of disease activity. These results suggest that microcirculatory dysfunction plays a role in the development and maintenance of inflammation in LVV and may be amenable to therapeutic targeting. Additionally, consideration should be given to the eye as a window through which to define and track disease activity in LVV. In summary, this thesis has identified several ways in which monitoring of disease activity in LVV might be improved. These findings have the potential to enhance LVV patient care by facilitating more accurate matching of treatment intensity with disease activity. Larger, multi- centre studies are now required to confirm and expand on these findings.

Leucine-rich alpha-2 glycoprotein 1 (LRG1) has been implicated in a variety of inflammatory and vasoproliferative diseases. This leucine-rich repeat family protein interacts with active TGF-beta1 and modulates TGFbeta signalling to promote pathological neovascularisation through activation of the pro-angiogenic SMAD 1/5/8 pathway. Emerging data continues to reveal TGFbeta-associated functions of LRG1 in disease, such as regulation of fibrosis. Despite this knowledge, little is known about the structure or molecular mechanism of LRG1. In this study, biophysical techniques were employed to characterise the LRG1 protein. In addition, a series of LRG1 variants were designed and expressed to test specific amino acid activity in TGFbeta pathway protein interaction studies, and in models of in vitro and ex vivo angiogenesis. Findings from small angle X-ray scattering, multi-angle light scattering and analytical ultracentrifugation experiments demonstrated that LRG1 exists as a mostly spherical monomer in solution. These data combined with glycosylation state findings showed that LRG1 expressed in HEK cells is glycosylated at 4 N-linked glycan sites. This project also revealed a previously unknown interaction between LRG1 and the BMP-2 ligand of the TGFbeta superfamily. Variant LRG1 studies located a functional amino acid (S297) in the LRG1 concave face, that when mutated altered the LRG1-TGF-beta1 interaction, decreased LRG1-activation of SMAD 1/5/8 signalling, and significantly inhibited LRG1s pro-angiogenic function in the metatarsal sprouting assay (p < 0.001). The characterisation of LRG1 structure and function in this study contributes to understanding of LRG1 mechanism in the TGFß pathway, and this information can aid future developments of anti-LRG1 therapeutics.

The role of leucine rich alpha-2-glycoprotein-1 (LRG1) as a biomarker has been demonstrated in various cancers, acute myeloid leukaemia, ulcerative colitis and others, where its functions have been studied in aberrant neovascularisation in the eye as well as being a key player in epithelial-mesenchymal transition (EMT). Its involvement in angiogenesis and inflammation have recently been established, making it a potential candidate as a therapeutic target for many diseases. We hypothesised based on the known pathophysiology of preeclampsia (PE) involving dysfunctional angiogenesis and inflammation, that LRG1 levels may be altered in women who develop PE during pregnancy, with particular interest in women with diabetes. We found that LRG1 plasma concentration is increased in women who developed PE, and this occurred specifically in diabetic women. In vitro studies revealed that the expression of LRG1 from trophoblast cells is reduced in hypoxia and this reduction is exacerbated in hyperglycaemic conditions. This is coupled with a reduction of PPARβ/δ, although PPARβ/δ does not seem to act downstream of LRG1 in trophoblasts. The mechanism of LRG1 reduction was found not to be mediated by miR-17. In vivo mouse models showed that LRG1 was reduced in diabetic Akita (Ins2Akita/Ins2+) mouse placenta compared to non-diabetic controls. This was accompanied with a net reduction in mesenchymal markers indicating a less invasive nature in the placenta. Due to the relationship of PE and the development of future cardiovascular disease. We studied the connection between LRG1 and heart failure. Left ventricular septal tissue from patients who died from end stage heart failure revealed a downregulation of LRG1 expression in patients with ischaemic cardiomyopathy, dilated cardiomyopathy and hypertrophic cardiomyopathy compared to patients who died from non-cardiac issues. No significant difference was found between control and patients with left ventricular hypertrophy. We then studied the functional association of LRG1 with the migration of human cardiac fibroblast cells and found that recombinant LRG1 increased the migration of cells whereas LRG1 knockdown reduced cell migration. This process may be mediated by PPARβ/δ. Our studies led us to investigate LRG1 (and sST2 and IL6) as a biomarker in patients with heart failure to differentiate between heart failure with preserved ejection fraction (HFpEF), heart failure with reduced ejection fraction (HFrEF) or non-heart failure patients. Use of two cohorts of patients from Sub-Saharan Africa and two of European Caucasian showed that LRG1, sST2 and IL-6 did not predict the development of heart failure in general and would be a poor marker for diagnosis.