Multidimensional assessment of diastolic dysfunction using echocardiography and magnetic resonance imaging in acute coronary syndrome

Abstract

Cardiovascular disease (CVD) is the leading cause of death in the western world and is becoming more important in the developing world. Recent advances in diagnosis, revascularisation and pharmacotherapy have resulted in a reduction in mortality. Nevertheless, the burden of disease remains large resulting in high healthcare costs. This has necessitated the introduction of a reliable, non‐invasive and cost‐effective imaging modality which has the ability to accelerate the diagnosis of acute coronary syndrome (ACS), aid in the risk stratification of de novo coronary artery disease (CAD) and afford incremental information of prognostic value such as viability. Clinically, the realm of ACS assessment has always been dependent on conventional methods such as electrocardiography (ECG) and serological markers of infarction or otherwise known as cardiac enzymes (CE). Unfortunately only 15‐20% of patients presenting with chest pain are ever diagnosed with ACS and up to 10% of patients with eventual myocardial infarction (MI) are misdiagnosed and inappropriately sent home. This is due to the delay in the pathophysiological manifestation of ACS, from arterial occlusion to overt ECG changes and/or CE rise, the heterogeneity of its symptom manifestation which is dependent on the patients’ age, gender and any underlying comorbidities, such diabetes mellitus which may delay and/or attenuate the symptoms and the lack of a holistic investigative modality that is able to quickly confirm the diagnosis of ACS/MI and thus expedite intervention. Cardiovascular magnetic resonance (CMR) is unique in that it offers high spatial resolution enabling a detailed volume and functional assessment, excellent tissue characterisation in any tomographic plane and exceptional prognostic ability which is afforded by late gadolinium enhancement (LGE). Furthermore, radiation free examination also affords the CMR with the ability to incorporate extensive imaging protocols and repeated imaging necessary for both clinical and research imperatives. Studies have already shown that CMR techniques such as myocardial volume and function, perfusion imaging and LGE is able to provide a more accurate diagnosis of ACS compared with standard clinical assessment that includes ECG and CE. In this thesis, our aims were to address some of these concerns through utillisation of novel imaging techniques advancing the utility of CMR as an important imaging modality in the assessment of ACS. Specifically we planned to look into two major areas of interest, diastolic dysfunction assessment and myocardial characterisation through T1 mapping. Our first paper discussed the importance of accurate early diagnosis of patients presenting with chest pain and the need for an imaging modality that would serve as a ‘one‐stop‐shop’ towards diagnosing ACS. We discussed the move away from conventional methods of assessing ACS and the non‐invasive imaging modalities available currently. We discussed the pros and cons of several imaging modalities, notably echocardiography and computed tomography based imaging. We then discussed the many different imaging techniques availed by CMR. We surmised that its ability to accurately and reliably diagnose, risk stratify and prognosticate ACS affords us with a comprehensive investigative tool for assessing patients who present to the Emergency department with chest pain. Our second paper looked at the utility of assessing diastolic dysfunction as a means of identifying coronary artery occlusion. We discussed its presence signifying one of the earliest signs associated with coronary artery occlusion and the various imaging techniques available to assess it. Echocardiography, being the principal assessment tool for the identification of diastolic dysfunction has several disadvantages namely anatomical impediments to complete endocardial definition, inter‐observer variability and lack of tissue characterisation. CMR however offers a more comprehensive assessment of diastolic dysfunction with the introduction of phase‐contrast, velocity‐encoded imaging and myocardial tagging. We also discussed the possibility of introducing a novel myocardial characterisation technique, T1 mapping and its possible relationship with regional diastolic dysfunction as an indicator of underlying ischaemia or infarction. Our first step was to evaluate the use of phase‐contrast velocity‐encoded imaging in the assessment of transmitral flow and the heterogenous contour size and locations used for its assessment. To achieve this we compared the planimetry of the entire left ventricular annulus inflow, the mitral valve leaflet orifice and the mid‐mitral valve inflow area, all techniques utilised in many previous studies in assessing transmitral inflow. In confirming the most accurate and reproducible technique, we would then be in a position to be able to confidently use this technique in all subsequent assessments utilised in our study. We have found that CMR derived E wave, A wave and E/A ratio can vary substantially according to planimetric contour location used. We have also established that the greatest degree of agreement, correlation and reproducibility is observed with mitral valve leaflet orifice, which is detailed in our third paper. We then proceeded with the comprehensive evaluation of global diastolic dysfunction through phase‐contrast velocity‐encoded imaging and compared it to the current standard ‐ echocardiogram, which is the subject of our fourth paper. We compared the CMR assessment of E/A, S/D and E/E’ assessed through phase‐contrast velocity‐encoded assessment of transmitral flow, pulmonary vein flow and mitral annular velocities against Doppler echocardiogram. We found that CMR evaluation of diastolic function with mitral/pulmonary flow and mitral annular velocity is readily achievable, reproducible and demonstrates significantly positive correlation with echocardiogram. This was important to establish for two reasons. Firstly, previous assessments of diastolic dysfunction utilised different imaging modalities inclusive of echocardiogram, computed tomography and CMR and did not thoroughly assess all three aforementioned indices in each of their studies. Secondly, before we could proceed with CMR assessment of regional diastolic dysfunction, we needed to be sure that the CMR assessment of global left ventricular diastolic dysfunction is comparable to the reference standard, echocardiogram. In clinical practice, diastolic dysfunction precedes systolic dysfunction, electrocardiographic abnormalities and angina in the pathophysiological cascade following coronary artery occlusion. However, visual estimation of diastolic dysfunction through left ventricular motion abnormality and T2‐weighted oedema imaging is subjective and requires extensive clinical experience. Therefore, the development of a noninvasive and quantitative method for detecting coronary artery stenosis through T1 mapping and diastolic circumferential strain analysis of the affected segments would simplify the screening process and ensure early diagnosis, thereby guaranteeing speedy intervention and thus limit the complications post‐ MI. To this end, our final study looked at comparing the assessment of myocardial characterisation through pre‐contrast T1 relaxation post‐ACS, with the CMR measure of regional diastolic dysfunction via left ventricular, segmental diastolic circumferential strain analysis. To our knowledge, no such study has ever been done. Segmental scarring and oedema post MI is known to impair segmental myocardial relaxation and increase myocardial stiffness thereby causing regional diastolic dysfunction. Similarly, scarring and oedema prolongs T1 relaxation in humans. Intuitively, there should be a significantly positive correlation between measures of segmental diastolic function and T1 relaxation post MI. However, several studies have shown poor correlation between measures of global diastolic function and T1 relaxation of infarcted segments post MI. We have found similarly, however when we compared the diastolic circumferential strain analysis with T1 relaxation times of the same infarcted segments, we found a significantly positive correlation with a good degree of agreement and excellent reproducibility. Therefore in summary, we have been able to demonstrate that the multidimensional CMR assessment of diastolic dysfunction is practicable, easily achievable, readily reproducible and adds a further discriminating value to CMR assessment of ACS. This thesis identifies the strengths and weaknesses of the techniques utilised in CMR assessment of diastolic dysfunction and allows us to use them more appropriately in future. Future work stemming from this thesis is already in the pipeline and will look into the use of other measures of myocardial scarring, oedema and haemorrhage such as T2W and T2* imaging, each independently associated with infarct size and prognosis post MI, and their possible relationship to either global or regional diastolic function.