Manganese-Enhanced Magnetic Resonance Imaging (MEMRI) in Ischaemic, Inflammatory and Takotsubo Cardiomyopathy (MEMORY)

BACKGROUND Heart failure (HF) is a global pandemic affecting at least million people worldwide. Despite the significant advances in therapies and prevention, mortality and morbidity are still high and quality of life poor.

Cardiomyopathy refers to diseases affecting the myocardium which can lead to heart failure.

Ischaemic heart disease (IHD) is the leading cause of cardiovascular morbidity and mortality in the United Kingdom (UK) and an important cause of reversible cardiomyopathy.

Takotsubo cardiomyopathy (TTS) is a form of stress cardiomyopathy that is characterised by acute, transient left ventricular (LV) dysfunction. This typically affects the apex extending beyond a single epicardial vessel with subsequent angiography revealing non obstructive coronaries. It accounts for up to 2% of acute coronary syndrome (ACS) admissions and despite being first described 20 years ago its pathophysiology is poorly understood. Typically takotsubo resolves over time and is considered to have a benign prognosis however acutely it can cause serious complications such as severe heart failure, cardiac arrhythmias and cardiac arrest.

Myocarditis is an inflammatory cardiomyopathy with varying aetiology which comprises a wide clinical spectrum from sub-clinical disease to severe heart failure. Currently myocardial biopsy is considered the gold standard in diagnosis, although it is prone to sample error and clinically indicated in only a few scenarios. Although in some cases cardiac MRI can reveal a 'typical' pattern of myocarditis based on T2 oedema imaging, the sensitivity and specificity is unclear and not yet established giving a limited diagnostic yield. Myocarditis may resolve spontaneously, recur or become chronic. In a third of biopsy-proven cases, it can lead to heart failure, cardiac transplantation or death. As a result, it has significant contribution to the global burden of cardiovascular disease.

Better understanding of underlying mechanisms and assessment of reversible myocardial injury is vital in the diagnosis, risk stratification, treatment, and intervention with potential to change clinical practice.

Delayed-enhancement MRI with Gadolinium Cardiac imaging using delayed-enhancement magnetic resonance imaging with gadolinium (DEMRI) is the gold-standard method for non-invasive characterisation of myocardial scar and function, and has become an invaluable tool in the field.

Gadolinium (Gd) is an excellent blood-pool contrast agent. Despite some active uptake in hepatocytes, Gd acts as a passive marker of the extracellular space. Chelated to diethylenetriamine-pentacetic-acid (DTPA), Gd diffuses rapidly into the extra cellular space but its molecular weight prevents it penetrating intact cell membranes. Many pathological myocardial states have a greater extra cellular volume (ECV) and as a consequence, a greater osmotic gradient relative to healthy tissue. These diseased tissues also result in the delayed wash out of Gd-DTPA resulting in an increased signal seen in pathological areas on DEMRI. However, the non-specific distribution of Gd within the extracellular space means direct assessment of viability is not possible with DEMRI.

Manganese-enhanced MRI (MEMRI) Manganese (Mn) was the first clinical MRI contrast agent. Cardiac, hepatic and renal uptake of Mn causes marked shortening of MRI T1 relaxation times providing excellent tissue contrast imaging. Mn-based contrast media offer potential for a wide range of MRI platforms having been applied to the assessment of neuronal activity and function, detection and tracking of lymphocytes16 and pancreatic beta-cell activation as well as evaluation of myocardial viability in the setting of ischaemia.

The mechanisms by which Mn provides contrast imaging in the heart depend on calcium (Ca) handling in myocardial cells. During myocardial contraction, Ca2+ ions are taken up into myocytes predominantly through voltage-gated L-type Ca2+ channels where they then trigger further Ca2+ release from the sarcoplasmic reticulum (excitation-contraction coupling). In diastole, Ca2+ is actively transported into sarcoplasmic reticulum by Ca2+-ATPase (SERCA2a), in addition to passage into the extracellular space via the Na+/Ca2+ exchanger and uptake into mitochondria. Alterations in this Ca2+ handling, by protein dysfunction and defective regulatory mechanisms, impair the ability of the myocyte both to increase and decrease intracellular Ca2+ concentrations, impacting on systole and diastole respectively.

Calcium plays a central role in excitation-contraction coupling and defective Ca2+cycling is key in the pathophysiological and adaptive mechanisms of defective contractile function and impaired relaxation in heart failure. Experimental reports have demonstrated reduced sarcoplasmic reticulum Ca2+storage and release, with decreased SERCA2a activity and Na+- Ca2+ exchange up regulation. The altered expression and activity of voltage-gated L-type Ca2+ channels both in heart failure and hypertrophy observed in several studies is not yet fully understood, although clearly highlights the centrality of calcium handling to this issue.

Mn acts as a Ca2+ analogue allowing its uptake by voltage-gated Ca2+ channels therefore cells with active calcium handling avidly take up Mn ions, in contrast to infarcted tissue which lack the necessary viable mechanisms. The result is a marked shortening of the T1 relaxation times in tissue with functioning calcium handling mechanisms. This has particular relevance to myocardial tissue and mitochondria since it has the potential to identify viable myocardium and provide a measure of myocardial function. In contrast to DEMRI with Gd, which effectively functions as a marker for pathological myocardium, MEMRI highlights areas of viable myocardium, marking non-viable myocardium by its absence of uptake. With its paramagnetic properties also reducing the T1 relaxation times of water, Mn provides positive image contrast in the tissues where it accumulates giving excellent anatomical delineation.

T1 Mapping The majority of myocardial pathologies prolong T1 relaxation (infarction, myositis, hypertrophy, ischaemia, amyloidosis, sarcoidosis), although lipid deposition (arrhythmogenic right ventricular dysplasia), iron deposition (haemachromatosis) and lysosomal storage disorders (Fabry disease) shorten it. On account of its potent paramagnetic properties, Gd shortens T1 in both normal and abnormal myocardium. A T1 map is generated by combining multiple images obtained during diastole, but with different inversion times, to assess T1 relaxation time for each voxel within the image. Each voxel intensity value represents the T1 relaxation time for the voxel and is labelled with colour for ease of visual differentiation. Different imaging protocols and techniques have been developed to combine an even sampling of the T1 recovery curve whilst minimising artefact, maximising precision and avoiding long breath holds. The principal potential advantages of T1 mapping compared to late gadolinium enhancement relates to its ability to detect diffuse more subtle forms of myocardial disease and to provide quantification rather than a binary "black versus white" categorisation. Combined with manganese enhancement, and time-optimised T1 mapping protocols, this should allow specific tissue characterisation rather than the more generalised extracellular and intravascular assessment provided by late-gadolinium enhancement.

Whilst T1 mapping eliminates errors introduced by windowing and variable signal enhancement, post contrast T1 quantification can be affected by variable contrast kinetics and is heavily reliant on precise timing of acquisition. The derivation of the partition coefficient and use of plasma volume to calculate extracellular volume fraction (ECV) have been developed to correct for these factors. Previous studies have demonstrated the value of ECV in the assessment of myocardial fibrosis in aortic stenosis, shown excellent reproducibility at 3-Tesla (3T) , and have correlated these values with other important biomarkers.

Mangafodipir In its early development, Mn toxicity occurred in animal studies with administration of magnesium chloride (MnCl2) because it competed too strongly with Ca2+ entry into the myocardium. It caused decreased myocardial function, hypotension and cardiac arrest.1 Two approaches were developed to circumvent these risks whilst maintaining the desired magnetic and kinetic properties. These are (i) chelation to bind the Mn2+ ions preventing competition with Ca2+, and (ii) co-administration of Ca2+ ions, effectively competing with Mn and reducing its cardio-toxic potential.12 Mangafodipir utilises the first of these two approaches, in a similar way to which Gd contrast agents mitigate against toxicity.

The principal limitation of Gd-based DEMRI imaging lies in quantifying the heterogeneous peri-infarct zone, where it results in overestimation of the non-viable infarct region.33 To date, there is no established imaging strategy to identify the viable and potentially, salvageable cardiomyocytes within this key area of myocardium. The investigators hypothesize that Mn-enhanced MRI will provide a method of cellular imaging through specific, non-perfusion dependent distribution due to active uptake by viable myocardial cells. The potential clinical application of this novel contrast agent will be to assess manganese uptake and therefore calcium handling in the myocardium of patients with reversible causes of myocardial injury such as reversible ischaemia, myocarditis and takotsubo cardiomyopathy. This will guide therapy and enable prognostication in these patient groups with high unmet clinical need.

RATIONALE for STUDY Pre-clinical and clinical data Mangafodipir has been used in animal models where it was compared to a non-chelated Mn-based contrast medium, EVP1001-1. Whilst the latter agent uses co-administration with calcium to counteract toxicity, Mangafodipir (Teslascan) employs chelation with dipyridoxyl diphosphate (DPDP) to counteract potential toxic effects.

The study successfully assessed T1 shortening in the healthy rat myocardium using both Mn contrast agents and demonstrated reduced T1 shortening (Mn uptake) with the simultaneous administration of diltiazem by approximately one third.

Spath et al have shown that manganese-enhanced magnetic resonance (MEMRI) is a better measure of myocardial infarction size than late gadolinium enhancement. Data indicate two important findings: (i) over-estimation of the scar by DEMRI (Gd) and (ii) quantitative differences in the scar volume between MEMRI and DEMRI. Moreover, they have shown that MEMRI is superior in assessment of viability as it gives a direct quantification of viable myocardium.

Furthermore, MEMRI with T1 mapping has shown altered manganese uptake and reduced rate of extra-cellular re-distribution in patients with dilated cardiomyopathy. In addition to better quantification of myocardial infarct size and assessing viability, we hypothesise that MEMRI will also allow the study of myocardial recovery in reversible cardiomyopathy such as reversible ischaemia, myocarditis and takotsubo cardiomyopathy.

Safety data and clinical use Naturally occurring deficiency is rarely reported but toxic effects of Mn have historically been recognised with Mn accumulating in the striatum and globus pallidus, the predominant manifestations being headaches and emotional lability, with parkinsonian extrapyramidal symptoms and gait-disturbance (manganism) developing with increasing severity. These concerns have been overcome as described above with chelation and calcium-competition.

The manganese agent used in this study, Mangafodipir (now a generic product), has been widely used in humans for the investigation of hepatic and pancreatic lesions. The agent has now been patented for application to cardiovascular disease and shows great promise.

Extensive animal studies and use in humans have enabled Phase I, II and III trials of Mangafodipir. In Phase III clinical trials, 624 patients were evaluable for adverse events. Infusion associated discomfort was reported in 4% of patients (n=24) a feeling of warmth of mild (n=22) to moderate (n=2) intensity. Six serious adverse events occurred. Of these, two were considered possibly/likely to be drug related and occurred in one patient. These were reported as a rash of the left arm in the presence of pre-existing lymphedema followed by a septicaemia after a superficial cut of the left hand. Vital signs were monitored in 321 patients and clinically important changes in blood pressure were recorded in 2 of these, one with an increased systolic blood pressure of 70 mmHg the other with increased systolic/diastolic pressures of 50/40 mmHg respectively. Monitoring of blood chemistry indicated increases in bilirubin levels in one patient and high total serum iron in 5 patients.

Special populations: No specific interaction, renal impairment or high risk patient studies were submitted. Subjects less than 18 years old were excluded in the trials. In the phase III studies, 216 of the 624 patients were over 65. The rate of adverse events was lower in this sub-group (4% vs. 9.3%). Bilirubin was measured in 257 patients and was abnormal in 30 of them although severe obstructive hepatobiliary disease was an exclusion criterion. 138 patients were recorded as having cirrhosis. No increased adverse event rate was found in either sub-group. 86 out of the 624 patients had unspecified cardiovascular disease. No difference in adverse event rate was observed with this sub-group.

Most common adverse reactions reported were transient and of mild intensity, with headache, nausea and flushing being most commonly reported (1-10%). Uncommonly reported adverse reactions were skin reactions, rhinitis, pharyngitis, abdominal discomfort, palpitation, gastrointestinal disturbance, dizziness, paraesthesia, altered taste sensation, fever and discomfort at injection site (0.1-1%). Other adverse reactions were very uncommon or rare. The frequency of reactions was found to be increased with faster infusion rates (4-6mls/minute).

More recently, clinical studies have focused on the application of Mangafodipir to human cardiovascular imaging and disease, reinforcing this large body of safety data as applied to a cardiovascular population as well as demonstrating the potential of the agent to characterise myocardial function.

Control Data This study is designed to explore the application of this novel cardiac MRI technique in reversible cardiomyopathy. Participants from the patient cohort will be compared to a healthy volunteer group.