Prognostic and pathophysiological features of uraemic cardiomyopathy using cardiovascular magnetic resonance imaging

Premature cardiovascular (CV) death is the commonest cause of death in patients with end stage renal disease (ESRD), which includes those receiving or close to requiring renal replacement therapy. In ESRD patients, CV deaths are most commonly caused by cardiac arrhythmia and sudden cardiac death compared to the general population where myocardial ischaemia and infarction predominate. Higher CV disease burden is due to accumulation of “conventional” risk factors (e.g. hypertension, diabetes mellitus, smoking) and “novel” risk factors (e.g. oxidative stress, proteinuria, anaemia, inflammation) in ESRD patients. In addition, risk factors specific to patients with renal disease have been identified including alteration in left ventricular (LV) structure, called uraemic cardiomyopathy. These structural abnormalities are common in patients with ESRD (between 60-80% of subjects upon initiation of dialysis) and include left ventricular hypertrophy (LVH), systolic dysfunction (LVSD) and dilatation. These changes in LV structure confer adverse CV outcome in ESRD patients and have proven difficult to reverse. Detection of these abnormalities is usually performed using echocardiography, however this technique is inaccurate in ESRD patients due to significant alterations in LV shape and geometric assumptions made during calculation of myocardial mass. Cardiovascular MRI (CMR) negates these assumptions and is the most accurate, reproducible and reliable method of assessing LV dimensions independent of intravascular volume, particularly in patients with altered myocardial architecture. Furthermore, maximal left atrial volume can be measured using CMR. The principle aims of the studies presented in this thesis were to elucidate prognostic and pathophysiological features of uraemic cardiomyopathy using CMR. In a large study (n=246) of haemodialysis patients, the determinants of each LV abnormality of uraemic cardiomyopathy were identified from past clinical history, haemodialysis and blood parameters and other LV measurements. For LV changes, major determinants were clinical features associated with advanced renal disease, namely expansion of intravascular/ extracellular fluid compartment, abnormal bone mineral biochemistry and hypertension. Furthermore, presence of one LV abnormality was one of the strongest predictors of presence of another, perhaps indicating differing stages of uraemic cardiomyopathy development. In a subsequent prognostic study including these patients (n=446), presence of LVSD and LV dilatation on CMR were significantly associated with poorer all cause and CV mortality. Presence of LVH, which is by far the most common structural change, was associated with poorer cardiovascular survival only. In addition, presence of two or three abnormalities (commonly LVH with another abnormality) had a significantly poorer prognosis and independently predicted CV and all cause mortality. This has implications for therapeutic strategies which should aim to slow or reverse cardiac changes of ESRD and prevent progression from one cardiac abnormality to 2 or more. In a further study (n=201) investigating additional prognostic features of ESRD patients with LVH, maximal left atrial volume (LAV) was measured using the bi-plane area length method at end LV systole. Elevated LAV and presence of LVSD were significantly associated with poorer all cause survival and were independent predictors of death. The most likely causes of elevated LAV in ESRD patients are LV diastolic dysfunction and expanded extracellular compartment and may provide a target for therapeutic intervention. The electrophysiological features of uraemic cardiomyopathy were assessed using microvolt T wave alternans (MTWA) which is a novel, non-invasive method of measuring small variations in surface electrocardiogram (ECG) T wave morphology and thus ventricular repolarisation. This technique has been used to stratify other cohorts at elevated risk of sudden cardiac death (such as ischaemic and non ischaemic cardiomyopathy, hypertensive LVH). A study presented in this thesis, compared MTWA results between ESRD (n=200) and hypertensive patients with LVH on echocardiography (n=30). Abnormal MTWA result was significantly more common in ESRD patients compared to hypertensive patients with LVH. Furthermore, abnormal MTWA result was significantly associated with myocardial abnormalities of uraemic cardiomyopathy and a history of macrovascular atheromatous disease in ESRD patients. Despite preservation of LV function on CMR, the frequency of abnormal MTWA result in ESRD patients was similar to previous studies in subjects with heart failure. 31Phosphorus magnetic resonance spectroscopy is a novel, non-invasive technique of estimating cardiac energetic status and high energy phosphate (HEP) metabolism in a myocardial area of interest and has previously been used to assess patients with global myocardial disease (dilated cardiomyopathy, hypertensive LVH). High energy phosphate metabolism was compared between patients with ESRD (n=53) and hypertensive LVH (n=30) and despite similar LV mass between both groups, PCr: ATP (an indicator of HEP metabolism) was significantly reduced in ESRD patients. These findings are most likely due to cardiac interstitial fibrosis and the alteration of tissue composition within the area of interest, and changes in metabolic function within cardiomyocytes of uraemic hearts. Finally, a small study (n=50) investigated the effect of successful renal transplantation on LV mass measured by CMR. On comparison of patients who remained on the renal transplant waiting list, there was no significant difference in LV mass in patients who received a renal transplant. It is likely that previous echocardiography studies that demonstrated significant regression of LVH, measured improvement in fluid control rather that actual reduction in myocardial mass. Future studies investigating benefit of therapeutic intervention may require identification of individuals at higher CV risk and the results of studies presented in this thesis aim to provide information for selecting such ESRD patients. With these results in mind, further prospective studies will be able to carefully select groups of ESRD patients with differing left ventricular, left atrial, electrophysiological and biochemical properties to demonstrate survival benefit with interventional agents. In this way, future therapies for ESRD patients can be tailored to improve cardiovascular survival