Search

Pregnant women represent an under-investigated population in clinical research and the mechanisms of long-term cardiovascular complications in women with obstetric complications remain to be elucidated. This is the basis which stemmed our interest in conducting our research in pregnant women with gestational diabetes mellitus (GDM) and preeclampsia(pE). A third of parous women experience adverse pregnancy outcomes; an increased risk of heart failure, stroke and ischemic heart disease. GDM and pE are the leading obstetric complications of pregnancy. Both are associated with a substantial increase in the risk of long-term cardiovascular disease. The aims of my longitudinal cohort study were to assess maternal cardiac alterations during the third trimester of pregnancy and their recovery twelve-months post-partum in women with GDM (n=30) or pE (n=22) compared to women with healthy pregnancies (HP)(n=38). Phosphorus-magnetic resonance spectroscopy (31 P-MRS) and cardiovascular magnetic resonance imaging (CMR) were used to define myocardial phosphocreatine to ATP ratio (PCr/ATP), tissue characteristics, left ventricular (LV) volumes, mass, ejection fraction (EF), global longitudinal shortening (GLS), and diastolic function (mitral in-flow E/A ratio). Participants were monitored to birth, baby birth weights were recorded, placental tissue histology was assessed. Investigations were repeated 12-months postpartum to monitor cardiac changes. With obesity as a major and common risk factor for both GDM and pE, ten overweight and ten normal-weight nulliparous women were recruited as non-pregnant control groups. These studies have demonstrated that despite distinct etiologies, women with GDM and pE exhibit a similar myocardial phenotype during and after pregnancy with persistent subtle abnormalities in myocardial PCr/ATP ratio, LV mass and GLS compared to women with HP during pregnancy and twelve months' postpartum assessments. Women with GDM and pE show similar myocardial alterations to nulliparous women with obesity, suggesting that the maternal myocardial changes are predominantly driven by obesity-associated metabolic dysfunction. In summary, the work in this thesis demonstrates the role of CMR and MRS in elucidating the cardiac structural, functional and energetic abnormalities in pregnancies complicated with gestational diabetes and preeclampsia. Most importantly, this work highlights that the persistent myocardial changes seen in women with GDM and pE at 12 months are mainly driven by obesity-associated metabolic dysfunction. With obesity becoming a huge worldwide health burden, it is very crucial to create societal awareness to support patients for a healthier lifestyle. This can be implemented through evidence based medicine and population based policies. Further research can be conducted to identify patients at risk and implement lifestyle interventions early to avoid long term complications to both the mother and fetus.

Type 2 diabetes mellitus (T2D) is associated with an increased risk of heart failure and cardiovascular mortality even in the absence of coronary artery disease. Although the reasons for this are not clear, candidate mechanisms are impaired cardiac high energy phosphate metabolism and coronary microvascular dysfunction. Cardiac magnetic resonance imaging (CMR) and magnetic resonance spectroscopy (MRS) are powerful tools for the noninvasive assessment of the functional, structural and the metabolic status of the heart. The overarching aims of this thesis were to combine advanced CMR and MRS techniques to study disease mechanisms in asymptomatic patients with uncomplicated T2D who had no prior diagnosis of cardiovascular disease. In chapter 3, prospective longitudinal cardiac structural and functional changes were studied in T2D patients over a six-year follow-up period. In this study T2D patients who have not experienced any cardiovascular events and remained asymptomatic over the follow-up period showed significant reductions in cardiac size and biventricular systolic function over time. The work represented in chapter 4 showed that while T2D is associated with greater visceral adiposity in overweight patients, accumulation of visceral fat was evident even in T2D patients with a normal body weight at levels similar to overweight non-diabetic controls. Secondly, in T2D patients, all measures of adiposity strongly correlated with one another, and visceral adipose tissue, body mass index and waist circumference were each related to abnormalities in cardiac systolic and diastolic strain, and insulin resistance. Thirdly, myocardial stress perfusion was only reduced in overweight T2D participants, with no reduction in global stress myocardial blood flow or myocardial perfusion reserve index in the normal body weight T2D patients or in overweight healthy volunteers compared to normal body weight healthy volunteers. The work represented in Chapter 5 was set out to establish the links between myocardial perfusion, systolic and diastolic function and energetics in the healthy and the diabetic human heart, at rest and during pharmacological stress combining 31phosphorus-MRS (31P-MRS) and CMR techniques. It is well established that the diabetic heart is characterised by reductions in myocardial phosphocreatine to ATP ratio which is a sensitive indicator of myocardial energetic status that can be noninvasively measured in vivo using 31P-MRS. This reduction is among the earlier subclinical cardiac changes in T2D patients. This work showed that in response to dobutamine stress, patients with T2D as well as healthy volunteers and age-matched veteran athletes show decrements in myocardial energetics, similar increments in global longitudinal shortening and left ventricular ejection fraction (LVEF), but with a blunted increment in stress perfusion in T2D patients. This work also showed that rest and stress myocardial blood flow are is associated with rest and stress LVEF, while rest and stress energetics are associated with rest and stress diastolic parameters respectively, suggesting that diastolic function is a more energetically sensitive process than global systolic function. Chapter 6 was a randomised, phase-2, single centre, open-label, cross-over design mechanistic drug trial of glucagon-like peptide-1 (GLP-1) receptor agonist liraglutide compared to peroxisome proliferator-activated receptor (PPAR)-gamma agonist pioglitazone. T2D is characterized by dysregulated insulin secretion and resistance to insulin action. Both the insulin secretion and insulin resistance are amenable to pharmacological intervention. GLP-1 receptor activation promotes insulin secretion and causes weight loss. Pioglitazone is a peroxisome proliferator activated receptor gamma agonist which targets peripheral insulin sensitivity. This work was set out to compare the efficacies of two distinct glycaemic control strategies of targeting beta-cell dysfunction (liraglutide) or insulin resistance (pioglitazone) in improving subclinical cardiac energetic, structural, functional and perfusion alterations in T2D patients with no known prior cardiovascular disease. This randomised cross-over study showed that four months treatment with incretin mimetic liraglutide results in significant improvements in myocardial perfusion and energetics, while the insulin sensitiser pioglitazone shows no effect in modulation of these parameters. Both treatments led to improvements in insulin sensitivity. Pioglitazone results in significant increases in LV mass and an isolated improvement in rest diastolic function. In conclusion, the work in this thesis demonstrates the power of CMR and MRS in elucidating the changes in patients with T2D who are overweight/obese, and of studying biological effects of novel treatment agents on the heart.

Introduction (Part 1): The nerves of the brachial plexus control movement and feeling in the upper limb. The most common form of traumatic brachial plexus injury (BPI) is root avulsion. Morphological magnetic resonance imaging (MRI) is used clinically to diagnose root avulsion, but its accuracy remains unclear. Diffusion MRI (dMRI) techniques characterise tissue microstructure and generate proxy measures of nerve 'health' which are sensitive to myelination, axon diameter, fibre density and organisation. Part 2: Chapter 1 describes a meta-analysis of 11 studies, showing that conventional (morphological) MRI has modest diagnostic accuracy for diagnosing root avulsion in adults with BPI. This represents the rationale for developing dMRI. Part 2: Chapter 2 shows a clinically viable dMRI sequence which is sensitive to established root avulsion in adults, highlighting uncertainties which warrant investigation before the technique is applied to acutely injured patients. Part 2: Chapter 3 is concerned with modelling the geometry of the brachial plexus in fixed cadavers, to inform the step angle in dMRI processing. Part 2: Chapter 4 is a meta-analysis of 9 studies which defines the normal fractional anisotropy and mean diffusivity values in healthy adult the brachial plexus, and how experimental factors influence dMRI parameter estimates. Part 2: Chapter 5 explores the effect of fractional anisotropy thresholding on deterministic tractography in the brachial plexus, identifying areas of uncertainty in the intrathecal and intraforminal areas. Part 2: Chapter 6 shows that two most common pre-processing pipelines worldwide generate important differences in dMRI parameters and tractograms. Part 2: Chapter 7 deploys high b-value multishell dMRI to show that up to 44% of the brachial plexus has multiple fibre orientations.

Diffusion tensor cardiac magnetic resonance (DT-CMR) is an emerging technique capable of characterising myocardial architecture in-vivo, something previously only possible in post-mortem studies. By measuring the velocity of diffusion of water molecules as it is impeded by cell membranes in the myocardium, DT-CMR can measure the mean diffusivity (MD) and fractional anisotropy (FA) - markers of magnitude and anisotropy of diffusion of water molecules. Based on the principle that water diffusion occurs preferentially along the long axis of cardiomyocytes, DT-CMR can infer upon the orientations of myocardial sheetlets by measuring the absolute secondary eigenvector angle (E2A), and the orientations of cardiomyocytes. This offers the potential to elucidate many of the pathophysiological mechanisms of cardiovascular diseases that are currently not well understood, as well as providing diagnostic and prognostic biomarkers which can complement the techniques that are currently available. The overall aim for this thesis was to validate the technical and clinical applications for DT-CMR in cardiovascular diseases. Specifically, this thesis includes work on sequence development, post-processing refinements and clinical studies of patients following ST-elevation myocardial infarction (STEMI) and patients with hypertrophic cardiomyopathy (HCM). Recent advancements in percutaneous coronary intervention have improved survival rates following STEMI, but despite optimal management, up to a third of STEMI survivors develop adverse left ventricular (LV) remodelling in the long-term. This is characterised by progressive cavity dilatation, impaired contractility and is associated with reduced survival. Adverse remodelling is difficult to predict and this maybe in part due to the incomplete understanding of the pathophysiological processes, particularly earlier in the disease pathway. By studying the acute and dynamic changes in microstructure post-injury, DT-CMR may help explain the ensuing maladaptive, mechanical changes in the myocardium. The aims of chapters 3-4 were to examine how DT-CMR parameters are affected following STEMI, comparing infarcted myocardium with adjacent and remote areas. The aim of chapter 5 was to establish the predictive relevance of acute DT-CMR for adverse-remodelling at 12 months post-STEMI. Through our results, we demonstrate that patients are at significantly greater risk if their acutely infarcted myocardium exhibits reduced FA (signifying cardiomyocyte disarray) and reduced E2A (signifying the underlying myocardial sheetlets remain fixed in a hypo-angulated state during systole). These results demonstrate the potential clinical utility of DT-CMR for prognostic risk-stratification following STEMI. These findings encouraged me to explore the behaviour of DT-CMR parameters in patients with HCM. This is another condition where the prognosis is diverse and difficult to predict, again partly owing to incomplete understanding about the underlying pathophysiological mechanisms particularly early on in the disease pathway. The aim of chapter 6 was to compare DT-CMR parameters between healthy volunteers, athletes and HCM patients. Our results demonstrated than HCM patients have globally higher MD (signifying diffusion is more unrestricted), reduced FA (signifying cardiomyocyte disarray) and higher E2A (signifying the myocardial sheetlets adopt a hyper-angular orientation during systole) than athletes and healthy volunteers. In chapter 8, the aim was to establish if these findings are evident beyond areas of macroscopic abnormalities. Our results demonstrate that in comparison with healthy volunteers, HCM patients - even in myocardial segment with normal wall thickness, normal perfusion and no scarring - have reduced FA, increased E2A and increased MD. These results shed light on the sequence of pathology and demonstrate the clinical utility of DT-CMR for early detection of phenotypic expression in HCM. In conclusion, DT-CMR is safe, feasible and can provide quantitative information about the histological state of the myocardium, which can complement clinical and imaging risk factors and improve clinical decision making for managing a range of cardiovascular diseases.

Background: Mitral regurgitation (MR) is a heterogeneous disease requiring accurate investigations to guide optimal management. Cardiovascular magnetic resonance (CMR) provides reference standard biventricular assessment and highly reproducible MR quantification. Exercise-CMR (Ex-CMR) combines CMR with physiological stress; further development may allow comprehensive MR assessment. Therefore CMR is ideal to assist clinical decision making and assess research outcomes. Aims: The thesis aims were to: 1) Develop and validate an Ex-CMR protocol assessing biventricular volumes and great vessel flow in healthy volunteers, 2) Evaluate the validated Ex-CMR protocol in primary MR patients, 3) Compare cardiac reverse remodelling and residual MR post mitral valve repair (MVr) vs replacement (MVR) in primary MR patients 4) Assess cardiac reverse remodelling after percutaneous mitral valve intervention for primary MR. Methods: 1) Free-breathing, respiratory navigated Compressed-SENSE short-axis cines and aortic/pulmonary phase contrast magnetic resonance sequences were validated against clinical sequences at rest and used during Ex-CMR in 12 healthy volunteers, 2) 10 primary MR patients underwent the validated Ex-CMR protocol, 3) Of 83 moderate-severe primary MR patients, 72 (30 MVr, 22 MVR, 20 controls) completed CMR imaging at baseline and 6 months after mitral surgery or observation (control group). 4) Of 11 primary MR patients, 10 completed CMR imaging at baseline and 6-months after percutaneous intervention. Findings: 1) Biventricular volumes and great vessel flow assessment during continuous supine Ex-CMR is feasible in healthy volunteers using the Compressed-SENSE Ex-CMR protocol, demonstrating good/excellent intra/inter-observer reproducibility, 2) The validated Ex-CMR protocol is feasible in asymptomatic primary MR patients demonstrating effective forward left ventricular ejection fraction is augmented by decreases in MR, 3) MVR results in comparable cardiac reverse remodelling to MVr with lower residual quantitated MR and better right ventricular ejection fraction (compared with controls) 4) In primary MR, percutaneous valve intervention results in MR reduction and positive left-ventricular reverse remodelling.

Abdominal aortic aneurysm (AAA) disease is characterised by an asymptomatic, permanent, focal dilatation of the abdominal aorta progressing towards rupture, which confers significant mortality. Patient management and surgical decisions currently rely on aortic diameter measurements via abdominal ultrasound screening. However, AAA rupture can occur at small diameters or may never occur at large diameters. Therefore, there is a need to develop molecular imaging-based biomarkers independent of aneurysm diameter that may help stratify patients with early-stage AAA to reduced surveillance. AAA uptake of [18F]fluorodeoxyglucose on positron emission tomography (PET) has been demonstrated previously; however, its glucose-dependent uptake may overlook other key mechanisms. The cell proliferation marker [18F]fluorothymidine ([18F]FLT) is primarily used in tumour imaging. The aim of the overall study for this thesis was to explore the feasibility of [18F]FLT PET / computed tomography (CT) to visualise and quantify AAA in the angiotensin II (AngII)-infused mouse model. The experiments presented in this thesis revealed increased uptake of [18F]FLT in the 14-day AngII AAA model than in saline controls, followed by a decrease in this uptake at 28 days. Moreover, in line with the in vivo PET/CT findings, Western blotting of aortic tissue revealed increased levels of thymidine kinase-1 (the substrate of [18F]FLT) and nucleoside transporters in the 14-day AngII AAA model than in saline controls, followed by decreased expression levels at 28 days. A pilot experiment further demonstrated that [18F]FLT PET/CT could be used to detect an early therapeutic response to oral imatinib treatment in the AngII AAA model. Therefore, [18F]FLT PET/CT may be a feasible modality to detect and quantify cell proliferation in the AngII AAA murine model. The findings of this thesis are encouraging for the application of [18F]FLT PET/CT in patients with small AAA.

Introduction: Right ventricular (RV) pacing is recommended treatment for bradycardia and can normalise life expectancy and restore quality of life. Clinical trials have demonstrated that chronic RV apical pacing may be associated with an adverse prognosis and development of heart failure. Multi-parametric cardiovascular magnetic resonance (CMR) has several clinical and research applications and can quantitatively assess cardiac anatomy, function, intracardiac flow and myocardial tissue characteristics. Aims: To assess 1.) the feasibility of 4D flow and validity of transvalvular flow quantification in pacemaker patients. 2.) haemodynamic response of intrinsic atrioventricular (AV) conduction and forced RV pacing in those with and without focal myocardial fibrosis. 3.) the hypothesis that long term RV pacing in presence of focal myocardial fibrosis leads to adverse cardiac remodelling. Methods: Between November 2017 and August 2019, patients with MRI conditional devices were prospectively recruited from a single centre. Patients underwent multi-parametric CMR at 1.5 Tesla. Patients in Chapters 3 and 4 underwent CMR in two separate pacing modes during a single visit. Patients in Chapter 5 underwent CMR before and 6 months after pacemaker implantation. Results: 1. Flow across aortic, mitral and tricuspid valves was consistent and reproducible in both pacing modes (p > 0.05 for all). 2. A greater decline in left ventricular (LV) ejection fraction (p=0.02) was observed during forced RV pacing in patients with myocardial fibrosis compared to patients without fibrosis. 3. Patients with AV block and myocardial fibrosis undergoing pacemaker implantation have greater electromechanical dyssynchrony and a consequent increase in LV end systolic volume (p=0.008) at 6 months compared to those without fibrosis. Conclusions: Right ventricular pacing in patients with myocardial fibrosis, compared to those without, leads to greater deterioration in cardiac function both immediately and after 6 months.

Primary systemic sclerosis primary heart involvement (pSSc-HI) is one of the leading causes of death in SSc. Sensitive testing suggests that the majority of patients have subclinical involvement; however, there is little understanding of its pathogenesis, dynamics and importantly, the optimal approach to clinical detection and management. Identifying the at-risk patient for more intensive monitoring and management is thus essential. By applying sensitive tools, including implantable loop recorder (ILR) and cardiovascular magnetic resonance (CMR) the thesis proposes to provide a comprehensive assessment of pSSc-HI and to identify clinical and cardiac serum markers that could aid early identification of patient at risk of pSSc-HI. The thesis also evaluates the predictive value of subclinical CMR abnormalities for the development of cardiovascular (CV) outcomes and assesses for any interval change of CMR detected pSSc-HI and whether disease-modifying anti-rheumatic (DMARD) or vasodilator treatment can alter the course of pSSc-HI. The thesis demonstrates significant ILR findings especially in a poor prognosis group, supporting therefore the use of ILR in SSc patients. CMR-extracellular volume (ECV), marker of diffuse fibrosis and cardiac biomarkers high-sensitivity troponin I (hs-TnI) and N-terminal pro-brain natriuretic peptide (NT-proBNP) associate with ILR abnormalities, suggesting their potential value in detecting pSSc-HI arrythmias. Three distinctive CMR findings are identified: diffuse and focal fibrosis and reduced myocardial perfusion. Skin score, digital ulcers and cardiac biomarkers are associated with CMR abnormalities, whilst CMR-ECV and NT-proBNP predict the development of future CV outcomes. Finally, no significant interval change in CMR parameters is observed, albeit a higher likelihood of CMR deterioration is noted in a poor prognosis group, whilst DMARD and vasodilator treatment do not seem to alter the course of CMR pSSc-HI. The results of the thesis are thus providing a first step towards risk stratification of pSSc-HI, that will support future larger studies.

Immune-mediated inflammatory diseases (IMID) are a group distinguished by specific pathways of immune-dysregulation that lead to inflammation, organ damage and dysfunction, of which Rheumatoid Arthritis (RA) and Giant Cell Arteritis (GCA) are two examples. An increased risk of cardiovascular disease (CVD) is observed in patients with IMID. Inflammation plays an important role in atherosclerosis. The inflammatory process has confounding effects on lipid and glucose metabolism, blood pressure and haemostatic factors. RA is a chronic inflammatory arthritis and one of the most common systemic autoimmune diseases affecting approximately 1% of the UK population. In RA, the risk of myocardial infarction (MI) is independent of, and incremental to, traditional CVD risk factors. CVD, as a direct result of the IMID process and independent of atherosclerosis can be an additional insult and seen in GCA. GCA is associated with an increased mortality mainly due to CVD, including aortic syndromes. Aortic involvement in GCA may manifest clinically as a syndrome of systemic inflammation without specific symptoms, or even be asymptomatic until aortic dissection or rupture supervenes with life-threatening consequences. The early identification of disease can benefit patients clinically by initiation of earlier disease modifying therapy, potentially reducing morbidity and mortality. Determining disease phenotype is important to develop effective screening strategies. This thesis aims to identify subclinical cardiovascular (CV) change using cardiovascular magnetic resonance (CMR) in the IMIDs of RA and GCA. It is also tested whether therapeutic interventions can modulate surrogate markers of CV outcomes in RA. Using CMR, this thesis demonstrates the presence of CV abnormalities in early RA (ERA) and established GCA. Changes in vascular function, myocardial tissue composition and ventricular geometry/performance are present in ERA, inferring an increased risk of CVD at the earliest stages of the RA disease continuum. There are thoracic aortic structural changes in established, treated, GCA. Dilatation of the thoracic aorta is common in GCA, with polymyalgic symptoms a possible risk factor. Patients with GCA also had increased aortic arterial stiffness than controls. CMR is able to detect subclinical CV change in asymptomatic ERA and established GCA patients without known CVD. Surrogate markers of CV events can be modulated in ERA with early aggressive RA therapy and CMR can identify silent aortic dilatation, associated with adverse outcomes, in GCA. CMR can aid the early and appropriate diagnosis of disease and complications, and potentially identify those at the highest risk of adverse outcomes.

Introduction: Coronary artery disease (CAD) remains the number one cause of mortality worldwide; improving diagnosis and treatment is a priority. Multi- parametric cardiovascular magnetic resonance (CMR) offers quantitative assessment of the cardiovascular system with a variety of techniques allowing assessment of anatomy, function, myocardial composition and perfusion during a single scan. Aims: To assess 1.) diagnostic accuracy of visual and quantitative perfusion CMR to single-photon emission computed tomography (MPS-SPECT) in patients with left main stem CAD. 2.) the hypothesis that patients with ischaemic (ICM) and non-ischaemic cardiomyopathy (NICM) have different torsion and strain parameters 3.) development and validation of a contemporary multivariable risk model of CAD from a large population undergoing X-ray angiography. 4.) a rapid 3D mDIXON pulse sequence for image quality and quantitation of MI. 5.) T1 rho prepared (T1ρ) dark blood sequence and compare to blood nulled PSIR (BN) and standard myocardium nulled PSIR (MN) for detection and quantification of scar. Methods: Patients were recruited between 2008 and 2017. Patients in chapters 3,4,6,7 underwent multi-parametric CMR including late gadolinium enhancement (LGE) imaging at 1.5 or 3.0T. Patients in chapter 5 underwent angiography. Results: 1.) CMR demonstrated significantly higher area under the curve for detection of LMS or equivalent disease over MPS-SPECT(P=0.0001). 2.) Despite no difference in LV dimensions, EF and strain between ICM and NICM, NICM patients had significantly lower LV twist(P=0.023) and torsion(P=0.017) compared to ICM. 3.) The developed model discriminated well and was well-calibrated. Diamond and Forrester and Duke scores substantially over-predicted CAD risk, whilst CAD Consortium risk models slightly under-estimated risk. 4.) Image quality was comparable between 3D and 2D LGE(P=0.162). Time for 3D image acquisition was only 5% of the time required for a standard 2D acquisition. 5.) CNRscar-blood was significantly increased for BN and T1ρ compared to MN LGE. BN LGE demonstrated significantly higher reader confidence scores.

Background: Mitral regurgitation (MR) is the second most frequent valve disease in the developed world and accurate assessment of MR severity and its complications are important. With its excellent accuracy and reproducibility, cardiovascular magnetic resonance (CMR) imaging is an ideal tool to quantitatively assess MR severity and its cardiac remodelling in various clinical settings. Aims: The aims of the thesis were to 1) Assess the impact of MR severity on cardiac reverse remodeling and patients' outcome in the TAVI population 2) Assess the impact of mitral valve (MV) repair versus MV replacement on cardiac reverse remodelling 3) Evaluate the feasibility and reproducibility of a navigated image acquisition method for biventricular physiological assessment during continuous physical exercise 4) Evaluate the feasibility of exercise CMR (exCMR) in patients with significant MR Methods: 1) 85 patients undergoing TAVI with CMR pre- and 6m post-TAVI were evaluated 2) Of 65 patients with significant MR, 37 patients (9 MV repair, 10 MV replacement and 18 medical management) with paired CMR scans at baseline and 6-months were evaluated 3) 10 healthy volunteers underwent exCMR on two separate occasions using a free-breathing, multi-shot, navigated, balanced steady-state free precession cine pulse sequence 4) 12 patients with significant degenerative MR underwent navigated exCMR. Findings: 1) Significant MR is common in patients undergoing TAVI and improves in the majority post-procedure. Improvement in MR was not associated with more favourable LV reverse remodelling and baseline MR severity was not associated with mortality 2) MV surgery leads to positive atrial and left ventricular reverse remodelling. In this small series, MV replacement with chordal preservation has similar cardiac reverse remodelling benefits to MV repair 3) The navigated exCMR protocol allows simultaneous biventricular physiological assessment during continuous exercise. Intra- and inter-observer reproducibility were excellent 4) The navigated exCMR protocol is feasible in clinical patients with significant MR.

Background. In patients with coronary artery disease (CAD), multi-parametric cardiovascular magnetic resonance (CMR) provides detailed information about the myocardial tissue composition and its function. This thesis aims to investigate how myocardial tissue composition characteristics influence its function. Additionally, this thesis aims to develop methods for time-resolved, three-dimensional flow imaging in-vitro and in health. Methods. We have prospectively recruited and conducted multi-parametric CMR in 76 patients with ST-elevation myocardial infarction (STEMI), 44 patients with stable chest pain and 35 healthy volunteers. Results. Study 1 and 2 evaluated if CMR derived function parameters are associated with the presence of microvascular obstruction (MVO) or intramyocardial haemorrhage (IMH) in STEMI patients. Study 1 demonstrated the diagnostic accuracy of mitral annular plane systolic excursion (MAPSE) to detect MVO/IMH. Study 2 demonstrated that peak global longitudinal strain (GLS) is most strongly associated with MVO/IMH when compared to other myocardial deformation parameters. Study 3 demonstrated that during first-pass perfusion stress CMR, at peak hyperemic stress, GLS is reduced in the presence of a perfusion defect in patients being investigated for CAD. Study 4 demonstrated that acute ECV-maps can reliably quantify area at risk and final infarct size (at 3-month follow) in reperfused STEMI patients. Study 5 investigated 3 main acceleration methods for 4-dimensional flow (4D Flow) in phantoms and healthy volunteers. In phantoms, all 4D Flow methods underestimated peak velocity but their mean errors were within reasonable limits for clinical applications. In volunteers, 4D echo planer imaging (EPI) method had the shortest acquisition time, best agreement with 2-dimensional phase contrast, best internal consistency between mitral and aortic valve flows. Conclusions. CMR derived longitudinal functional parameters of the left ventricle are associated with pathologies predominantly in the sub-endocardium and could be used as a surrogate imaging markers. ECV-maps offer an alternative to IS quantification. 4D EPI is the most robust acceleration method for whole-heart flow assessment.

Introduction: Cardiovascular disease (CVD) remains the number one cause of mortality in the world for both men and women, thus improving its diagnosis and treatment is a priority. Rheumatoid Arthritis (RA) is a common auto-immune disease associated with high rates of CVD. Cardiovascular magnetic resonance (CMR) offers a multi-parametric, quantitative approach to the assessment of the heart and cardiovascular system with a host of techniques allowing assessment of anatomy, ventricular function, myocardial composition, myocardial perfusion, vascular performance and myocardial metabolism during a single scan. Its quantitative nature and lack of ionising radiation lend themselves ideally to the longitudinal study of subclinical CVD. Aims: To assess 1) whether blood longitudinal relaxation (T1) can be used to estimate blood haematocrit value to allow calculation of extracellular myocardial volume fraction (ECV), 2) whether CMR feature tracking (CMR-FT) is a feasible means of assessing aortic stiffness, 3) whether global longitudinal strain (GLS) is reduced in patients with prior MI with preserved left ventricular ejection fraction versus healthy controls, 4) whether aortic stiffness is present at all time points in the disease course of RA and 5) whether subclinical CV abnormalities in newly diagnosed RA improve with treatment and if the anti-tumour necrosis α Inhibitor Etanercept offers additional benefit over standard treatment. Methods: Patients were recruited between and February 2011 and February 2017. All patients underwent a comprehensive, multi-parametric CMR study including cine and late Gadolinium enhancement imaging at either 1.5 or 3.0T. Results: 1) estimation of blood haematocrit from blood T1 value provides accurate estimation of ‘synthetic’ ECV, 2) CMR-FT assessment of aortic stiffness is feasible and provides reproducible values for descending and ascending aortic strain values, 3) GLS is reduced in prior MI patients versus healthy controls (-17.3 ± 3.7% versus -19.3 ± 1.9% respectively, p=0.012). A GLS cut-off value of 18% correctly identifies prior MI with a sensitivity of 60% and specificity of 72.5%, 4) Aortic stiffness is evident in at risk RA individuals, newly diagnosed RA as well as established RA and 5) Aortic distensibility and left ventricular mass improve significantly in newly diagnosed RA patients following treatment. Etanercept appears to offer additional benefit over standard treatment evidenced by numerical improvements in aortic distensibility.

Introduction First-pass myocardial perfusion cardiovascular magnetic resonance imaging (CMR) has become an established method for the non-invasive diagnosis of ischaemic heart disease (IHD). As complete datasets are acquired every heart beat, first-pass perfusion CMR is challenging, and compromises have to be made between competing demands for spatial resolution, cardiac coverage, and temporal resolution. Accelerating data acquisition with k-t undersampling techniques is a strategy that could overcome some of the remaining limitations of first-pass perfusion CMR, and influence the management of patients with IHD. The use of a k-t Principle Component Analysis (PCA) acceleration method will be investigated for its technical feasibility and clinical merit for acquisition with improved spatial resolution, greater myocardial coverage, and higher heart rates. Supplemented by three additional chapters as appendices, this thesis presents the use of spatiotemporal undersampling for advanced perfusion CMR. The research has the following aims. Aims and Methods 1. To establish the clinical feasibility of 3D perfusion CMR technique in relation to fractional flow reserve (FFR) and the Duke Jeopardy Score. Accuracy in diagnosis and comparison of ischaemic burden with invasive indices of myocardial ischaemia, will be determined. 2. To compare the estimation of ischaemic burden using 3D perfusion CMR against the estimation using the clinical standard single-photon emission computed tomography (SPECT). 3. Optimization and design of a 3D perfusion CMR sequence based on k-t acceleration and PCA reconstruction using a turbo field-echo (TFE) pulse sequence, and development of balanced steady-state free precession (bSSFP) perfusion imaging at 3 Tesla (3T). 4. To use k-t acceleration schemes to establish the feasibility of first-pass stress perfusion CMR in a rodent model, and to validate this against the microspheres method. 5. Multicentre evaluation of 3D perfusion CMR imaging for the detection of IHD defined by fractional flow reserve (FFR)(appendix). 6. To use k-t acceleration schemes for high-resolution quantitative first-pass perfusion imaging, and to determine reproducibility (appendix). 7. To compare advanced perfusion CMR imaging techniques: high-spatial resolution versus 3D whole-heart coverage (appendix). Results This research provides novel experimental evidence and technical advancements on the clinical utility of k-t PCA acceleration methods, and demonstrates the following: 1. That 3D whole-heart myocardial perfusion CMR imaging at 3T accurately detects functionally significant CAD; 2. That 3D whole-heart myocardial perfusion CMR imaging can determine ischaemic burden with accuracy comparable to current imaging techniques that rely on ionising radiation; 3. That the development of a 3D bSSFP myocardial perfusion CMR sequence is feasible using radio frequency (RF) shimming with dual-source parallel RF transmission at 3T. 4. That first-pass myocardial stress perfusion CMR imaging is feasible in a murine model using a 3T clinical scanner. The use of k-t spatio-temporal undersampling also yielded the following findings: 5. 3D whole-heart myocardial perfusion CMR imaging was highly efficient in the detection of functionally significant CAD in a multicentre study (appendix). 6. Quantitative high-resolution myocardial perfusion CMR showed inter-study reproducibility with no significant diurnal variation (appendix). Conclusion These findings have important implications, lending support to the clinical use of 3D whole-heart perfusion imaging for the detection of coronary disease. The use of k-t acceleration schemes is feasible for both clinical and pre-clinical models. When combined with CMR assessment of function and viability, the technique holds promise as a completely non-invasive and radiation-free diagnostic and risk-stratification tool for patients with known or suspected CAD.

Background: Aortic valve stenosis is the most common degenerative valve disease affecting the Western elderly population. Medical therapy is ineffective at treating the mechanical obstruction of blood flow. Surgical Aortic Valve Replacement (SAVR) is the current recommended treatment for symptomatic severe AS but often high risk patients are declined for this. Transcatheter Aortic Valve Implantation (TAVI) allows the percutaneous delivery of the prosthetic valve but this novel approach is associated with complications. Aims: This thesis aims to focus on the effects of TAVI and contemporary SAVR on patients’ quality of life, neurocognitive function and the left ventricular reverse remodelling. Methods: High risk patients with symptomatic aortic stenosis were studied at baseline, 30 days, 6 month and 12 months after intervention. Cerebral MRI with diffusion weighted imaging for micro-embolism was conducted before and after intervention and again at 6 months. Cardiac MR was conducted at baseline and 6 months. Health related quality of life and a comprehensive battery of neurocognitive functional assessments were also conducted across 3 and 4 time points respectively. Results: The incidence (54(77%) vs. 17(43%), p=0.001) and number (3.4±4.9 vs. 1.2±1.8, p=0.001) of new micro-infarcts was greater after TAVI compared to SAVR. Physical component scores (PCS) in TAVI increased after 30 days (32.1±6.6 vs. 38.9±7.0, p<0.0001) and 6m (40.4±9.3, p<0.0001); the improvement occurred later in SAVR (baseline: 34.9±10.6, 30d: 35.9±10.2, 6m: 42.8±11.2, p<0.001). At 12 months, the majority of neurocognitive function tests did not show a significant change in the proportion of patients categorised as having impaired NCF compared to baseline in the TAVI or SAVR groups. After 6 months, there were significant improvements in indexed end-diastolic volumes (TAVI: 100±25mls vs. 87±26mls, p<0.001; SAVR 91±28mls vs. 82±17mls, p<0.05) Extracellular volumes were similar for both groups at baseline (range 22.8 to 24.6%). There was no significant change in ECV after 6 months (TAVI, 24.0±9% vs. 29.3±11%, SAVR, 23.8±7 vs. 23.5±9, p= 0.76). Conclusions: TAVI patients experience higher numbers of cerebral micro-infarcts than contemporary SAVR patients, but this appears to have no effect on HRQoL; TAVI patients experienced earlier improvements in quality of life than SAVR patients. There was also no evidence of neurocognitive functional decline after TAVI using a battery of very sensitive neurocognitive function tests. Both TAVI and SAVR improved cardiac imaging parameters with evidence of reverse LV remodelling but no change in diffuse myocardial fibrosis after 6 months.

Introduction: Aortic stenosis (AS) is the commonest valvular lesion in the developed world and is associated with adverse cardiac remodelling. With its excellent accuracy and reproducibility, cardiovascular magnetic resonance (CMR) imaging is an ideal tool to assess cardiac remodelling and reverse remodelling following surgical aortic valve replacement (SAVR) and transcatheter aortic valve implantation (TAVI). The aims of this thesis were: 1) to evaluate gender differences in AS and following aortic valve replacement, 2) to evaluate the incidence of post-procedural myocardial infarction following SAVR and TAVI, 3) to describe the immediate effect of TAVI on reverse remodelling and 4) to assess the impact of TAVI-induced left bundle branch block (LBBB) . Methods: Between January 2009 and April 2015, patients with severe AS undergoing either TAVI or SAVR were prospectively recruited. Patients underwent comprehensive 1.5T CMR evaluation pre-procedure, prior to hospital discharge and 6m post-procedure. Results: 1) Women with severe AS have a lower indexed left ventricular (LV) mass than men (65.3± 18.4 vs. 81.5±21.3g/m2, p < 0.001). 6m following valve replacement, LV mass regression is similar between genders (men 21.7±10.1 vs. women 18.4±11.0%, p=0.121). 2) Myocardial infarction (MI) is more frequent following SAVR than TAVI (n=10 (26%) vs. n=3 (5%), p=0.004). 3) Over 10% of LV mass regression occurs prior to hospital discharge following TAVI and is more pronounced in the absence of myocardial fibrosis (p=0.005). 4) TAVI-induced LBBB is associated with a reduced LVEF 6m following TAVI compared with those with a narrow QRS (-2.1±6.9 vs. +4.6±7.8%, p=0.002). Conclusions: TAVI and SAVR are associated with favourable cardiac reverse remodelling which does not differ according to gender and begins prior to hospital discharge. SAVR is associated with a higher incidence of post-procedural MI than TAVI. TAVI-induced LBBB should be avoided where possible due to its unfavourable effects on cardiac reverse remodelling.

Immune-mediated inflammatory disease (IMID) represents a group of diseases characterised by dysregulation of immune processes with a shared common inflammatory pathway, leading to end-organ damage, of which Rheumatoid Arthritis (RA) and Systemic Sclerosis (SSc) are two important examples. An accelerated risk of cardiovascular disease (CVD) with associated increased mortality is observed in those with IMID; the risk in RA being similar to those with diabetes mellitus. In addition, IMID can directly affect the myocardium independently of atherosclerosis, termed primary myocardial disease, causing further excess mortality; best described in SSc. Determining the disease phenotype most at risk of either macrovascular or primary myocardial disease and having a greater understanding of the underlying pathophysiology is vital to develop effective screening strategies to prevent and manage its complications. Surrogate markers of CVD, including soluble cardiovascular (CV) biomarkers and cardiovascular magnetic resonance (CMR) imaging, can inform of subclinical CVD or risk of progression to clinical CVD; with currently limited data in IMID. Using soluble CV biomarkers and CMR, this thesis demonstrates the presence of subclinical CVD in patients with RA and SSc free of clinical CVD. In RA, these abnormalities associate with traditional CV risk factors; emphasising the importance of their aggressive management. Using CMR, this thesis describes a reduction in left ventricular mass in RA; suggesting pathology other than atherosclerosis. This work investigates specific treatment strategies in the reduction of CV risk, reporting improvement in insulin resistance with TNF inhibition in a randomised controlled trial of early RA. In SSc, the utility of CMR in the assessment of primary myocardial disease is demonstrated, describing cardiac fibrosis in those free of known cardiac disease, associating with a poor prognostic phenotype. Finally, this thesis reports the novel use of an implantable loop recorder in SSc, detecting arrhythmias in patients free of known cardiac disease.

Background: Surgical aortic valve replacement (SAVR) remains first-line treatment for symptomatic severe aortic stenosis, whereas transcatheter aortic valve implantation (TAVI) is indicated in patients who are inoperable or considered too high-risk for surgery. Current focus is centred on differences in the impact of valve replacement upon cardiovascular function to guide patient selection and the development of novel prosthetic valves to improve outcomes. Cardiovascular Magnetic Resonance (CMR) imaging is the investigative modality of choice for such a purpose. Objectives: To compare the impact of SAVR and TAVI upon aortic stiffness, right ventricular function and myocardial strain, and to compare two vendor designs in the quantity of post-TAVI aortic regurgitation and reverse remodelling. Methods: A prospective study of patients with severe aortic stenosis under surveillance and subsequently requiring SAVR or TAVI, recruited between September 2009 and December 2015. A 1.5 Tesla CMR study was performed pre and 6 months post SAVR, and pre, immediately and 6 months post implantation of Medtronic CoreValve and Boston Lotus TAVI. Aortic distensibility (AD), pulse wave velocity (PWV), right ventricular (RV) volumes, myocardial strain and aortic regurgitation (AR) were quantified. Results: At 6 months, SAVR was associated with a significant worsening in PWV (6.38±4.47 vs. 11.01±5.75ms-1, p=0.001) and ascending AD (1.95±1.15 vs. 1.57±0.68x10-3mmHg-1, p=0.044), whereas no change was seen following TAVI. A significant reduction in RV ejection fraction (58±8 vs. 53±8%, p=0.005) was seen flowing SAVR, with no change following TAVI. A significant and comparable decline in LV torsion and twist was observed. Baseline circumferential strain was significantly associated with all-cause mortality (hazard ratio, 1.03; 1.01–1.05; p=0.009). Significantly less AR was seen immediately following Lotus than CoreValve TAVI (4.3±3.4 vs.11.7±8.4%, p=0.001) with equivalent degrees of reverse remodelling observed at 6 months. Conclusion: Compared with TAVI, SAVR is more detrimental upon aortic stiffness and right ventricular function at 6 months. CMR derived circumferential strain is associated with survival following SAVR and TAVI.

Background: Myocardial perfusion imaging with cardiovascular magnetic resonance (CMR) is bound by spatio-temporal constraints. Standard perfusion CMR techniques permit the acquisition of 3-4 myocardial slices with a spatial resolution of 2-3mm. However, acceleration techniques can be applied to achieve higher spatial resolution (< 2mm) or 3-dimensional (3D) acquisitions with increased spatial coverage (12-16 slices). Acceleration can also be used to simultaneously acquire perfusion data at different time-points in the cardiac cycle. Accordingly, this thesis includes studies that modify the standard approach to perfusion CMR in order to investigate the impact of spatial resolution, spatial coverage and cardiac phase of acquisition. Methods and Results: Study 1 and 2 compared high-resolution and standard-resolution perfusion CMR in patients with suspected coronary artery disease (CAD). Study 1 found high-resolution acquisition had greater diagnostic accuracy compared to standard-resolution for detecting CAD (area under curve [AUC]: 0.93 vs. 0.83; p < 0.001); and study 2 found it also had greater diagnostic accuracy for specifically identifying 3-vessel CAD (AUC: 0.90 vs. 0.69; p < 0.0001). Study 3 compared high-resolution and 3D perfusion CMR in patients with CAD and found limited agreement between myocardial ischaemic burden estimates (95% limits of agreement: -8.68%, 9.82%). Study 4 and 5 compared systolic and diastolic acquisitions using standard perfusion CMR (limited to 1 slice) and 3D perfusion CMR respectively. Both studies found higher estimates of myocardial blood flow (MBF) in diastole compared to systole at stress (p < 0.05). Study 6 utilised accelerated perfusion CMR to compare MBF estimates at 5 different time-points in the cardiac cycle in healthy volunteers. Estimates of stress MBF peaked at end-diastole and fell steadily to end-systole (p < 0.0001). Conclusion: By altering the spatial resolution, spatial coverage and cardiac phase of acquisition of perfusion CMR, we have gained valuable insights into the relative impact of these parameters on both qualitative and quantitative assessment of ischaemia.

Background: Use of contrast enhanced cardiac magnetic resonance imaging (MRI) for identification of focal pathology (perfusion deficit and scar) is widespread. Quantitative analysis of dynamic contrast enhanced (DCE) MRI data may allow objective assessment of focal and diffuse disease. However it is a complex process and not widely adopted outside the research domain. For accurate quantification temporal variation in relative contrast agent concentration in the myocardium and feeding blood supply must be measured. While MRI signal intensity can be used as a probe of contrast agent concentration its response is non-linear. Aims: In this thesis non-linearity correction methods for quantitative myocardial DCE-MRI are compared, the feasibility of a novel bookend T1 based correction is tested and the method is used in clinical studies to assess myocardial characteristics in health and ischaemic disease. Methods: Signal non-linearity correction methods were compared using simulation, phantom experiments and a volunteer study. Methods compared were independent sampling strategies (dual-bolus and dual-sequence), previously proposed model based correction (native T1 or proton density weighted image based) and bookend T1 based correction which is proposed as a method to account for imperfect magnetisation preparation. The feasibility of the bookend T1 method was tested and characteristics of heathy and diseased myocardium were assessed in clinical studies of ischaemia and infarction. Conclusions: Native T1 based correction has been found to be highly sensitive to imperfect magnetisation preparation, and is thus recommended against. Model based correction using proton density weighted images or bookend T1 data have been found to be more accurate and precise than dual-sampling methods. The clinical studies have demonstrated the feasibility of the bookend T1 based method and have yielded insights into myocardial characteristics in a range of conditions.

Background: The early identification of disease can benefit patients clinically and provide a powerful research tool. This thesis aims to identify subclinical cardiac change using cardiovascular magnetic resonance (CMR) in both disease and health and evaluate its diagnostic and prognostic uses. Methods: We have prospectively recruited and conducted multi-parametric CMR in 50 patients with hypertrophic cardiomyopathy (HCM), 40 endurance athletes and 100 asymptomatic patients with type 2 diabetes mellitus. Results: Study 1 and 2 evaluated CMR in the early diagnosis of HCM. Study 1 demonstrated the diagnostic accuracy of extracellular volume (ECV) mapping is superior to volumetric methods of differentiating HCM from athletic left ventricular (LV) hypertrophy. Study 2 demonstrated that ECV expansion could be detected prior to overt hypertrophy or impairment of contractile function in patients with HCM. Study 3 demonstrated that LV torsion is lower in endurance athletes than controls and is predominantly influenced by lactate threshold and intensity of training. Study 4 and 5 investigated the role of CMR in identifying patients with type 2 diabetes at risk of heart failure and silent myocardial infarction. Study 4 found that the increased risk of heart failure in patients with type 2 diabetes was mediated by ECV expansion and diffuse fibrosis. There was a high rate of silent myocardial infarction (17 %) which was unrelated to heart failure risk. In study 5 we developed a simple screening tool, using measures that can be carried out in a cardiology clinic, for the detection of silent myocardial infarction in type 2 diabetes. Conclusions: CMR is able to detect subclinical change in both tissue characteristics and function of the heart. This can aid the early and appropriate diagnosis of disease and identify those at the highest risk of adverse outcomes.

Objectives: To apply and refine conventional and developing cardiac magnetic resonance (CMR) imaging tissue characterisation techniques, with the aim of applying them to better understand cardiac remodelling in health, disease and as an outcome surrogate in clinical trials. Background: Cardiac structure, function and perfusion are routinely assessed with a range of imaging modalities in both research and clinical practice. Cardiac magnetic resonance (CMR) imaging is emerging as the gold standard tool for many of these assessments. The use of gadolinium containing contrast agents in CMR protocols allows the detection of myocardial scar and focal fibrosis and provides important prognostic information. The developing field of T1 mapping allows measurement of the extracellular volume, a surrogate for fibrosis that offers further insights into diffuse myocardial change not previously possible. Methods: CMR tissue characterisation techniques were applied in sequential studies of: ischaemic cardiomyopathy, health, athletic cardiac adaptation and in a randomised controlled study examining the effects of spironolactone in heart failure with preserved ejection fraction. Results & Conclusions: Late gadolinium enhancement (LGE) imaging is of limited application in predicting functional recovery of dysfunctional segments in ischaemic cardiomyopathy (Chapter 3). Modified Look-Locker Inversion recovery (MOLLI) T1 and extracellular mapping techniques can be applied reproducibly in health and following either bolus or split dose gadolinium administration (Chapters 4 & 5). T1 mapping provides important insights into athletic cardiac remodelling that may allow its application in distinguishing between athletic and myopathic change (Chapter 6). Individuals with heart failure with preserved ejection fraction (HF-PEF) have diffuse myocardial fibrosis as measured with T1 mapping. Early results of a randomised controlled study suggest that the beneficial effects of spironolactone in HF-PEF may in part be due to regression of diffuse myocardial fibrosis (Chapter 7).

Objectives:- To evaluate myocardial tissue characterisation by cardiovascular magnetic resonance (CMR) to predict functional recovery in reperfused acute myocardial infarction (AMI). Background:- Prognosis following AMI is closely related to recovery of myocardial contractile function. Accurate early prediction of functional recovery may allow for additional therapies in high risk patients, and avoid over-treatment of lower risk patients. Clinical prognostication commonly relies on echocardiographic evaluation of function, which may be misleading acutely. CMR offers a number of ways to refine prediction of functional recovery by characterising myocardial tissue, but these have not been extensively evaluated. Methods:- Patients following reperfusion for first-presentation ST-elevation AMI were scanned by CMR acutely, subacutely and in convalescence. Tissue pathologies visible on acute CMR were evaluated for their ability to predict recovery of contractile function. Oedema in the peri-infarct zone, microvascular obstruction (MO) and intramyocardial haemorrhage (IMH) in the infarct zone, and extracellular volume (ECV) of the infarct zone were evaluated. In addition, susceptibility-weighted MR imaging (SW-MRI) was evaluated against the reference standards of T2-weighted and T2* imaging to detect the known prognostic marker of IMH. Results:- Acutely oedematous myocardium demonstrated recovery of function over time (p<0.05), whereas non-oedematous peri-infarct myocardium did not (p>0.05). Recovery of function closely mirrored resolution of oedema. Infarct contractile recovery was attenuated in infarcts that demonstrated MO acutely as compared to those without (p<0.01), and attenuated further in those that demonstrated IMH acutely (p<0.01). SW-MRI had sensitivity 93% and specificity 86% as compared to T2-weighted imaging, with excellent inter-observer reliability and shorter breath-hold times (4 seconds vs. 16 seconds). Infarct ECV had higher accuracy to predict improved wall motion than late gadolinium enhancement imaging (c-statistic 0.80 vs. 0.70, p=0.04). Conclusion:- Tissue characterisation by CMR offers a variety of ways to predict functional outcome following AMI, using both established and novel imaging techniques.

Background: Severe symptomatic aortic stenosis (AS) heralds a poor prognostic outlook and significant co-morbidity, with valve replacement the only definitive cure. Transcatheter aortic valve implantation (TAVI) has developed as an alternative to the standard treatment of surgical aortic valve replacement (SAVR) in high-risk or inoperable AS patients. The clinical and cost effectiveness of TAVI compared to SAVR requires further investigation. Methods: A prospective study of sixty seven TAVI and twenty seven SAVR patients, recruited from September 2009 to September 2011. Baseline assessments included a cerebral and cardiovascular magnetic resonance scan (1.5 Tesla MRI system) and the completion of two health surveys (EQ 5D and SF 12). Follow-up MRI was performed at 5±2 days (cerebral MRI) and 6 months (cardiovascular MRI) post AVR. Health status was assessed at 30 days, 6 months and one year. A cost-effectiveness analysis was performed using a 10 year Markov model with deterministic and probabilistic sensitivity analyses. Results: TAVI and SAVR resulted in similar levels of ventricular reverse remodelling. TAVI had a greater reduction in valvular impedance (21±8mmHg vs. 35±13mmHg, p=0.017) and myocardial fibrosis (10.9±6 % vs. 8.5±5%, p=0.03). Cerebral emboli occurred in 77% of TAVI patients. Age (r=0.37, p=0.042), severity of atheroma (r=0.91, p<0.001) and catheterisation time (r=0.45, p=0.02) were predictors of cerebral infarcts. HRQOL significantly improved over 12 months (PCS, p=0.02; EQ-5D, p=0.02; VAS, p=0.01 and SF6D p=0.03). Male gender (SF6D, p=0.01) and increased operator experience (PCS, EQ5D and VAS, p<0.05) predicted an improvement in HRQOL. Despite greater procedural costs, TAVI was cost-effective compared to SAVR over the 10 year model horizon (costs £52,593 vs. £53,943 and QALYs 2.81 vs. 2.75) indicating that TAVI dominated SAVR. Conclusions: TAVI has comparable cardiac and health benefits to SAVR, but greater cerebral complications. TAVI is likely to represent a clinical and cost effective alternative to SAVR.