Advanced electrocardiography in myocardial electrical remodeling : insights from cardiovascular magnetic resonance imaging

Sammanfattning: The electrocardiogram (ECG) is a common diagnostic tool in cardiology thanks to its high accessibility and low cost. However, in several cardiovascular diagnoses, including left ventricular hypertrophy (LVH), current conventional ECG measures and criteria have a poor diagnostic performance. LVH is associated with hypertension and diabetes, but is often missed by the standard 12-lead ECG. LVH is typically diagnosed by non-invasive imaging methods. Cardiovascular magnetic resonance (CMR) is the gold standard for diagnosing LVH. Advanced-ECG (A-ECG) is a term used to describe a combination of advanced ECG analysis methods, and has been shown to be of diagnostic and prognostic utility. The aims of this thesis were to investigate the ability of A-ECG to diagnose LVH, using CMR as reference, as well as investigating the prognostic ability of A-ECG measures with regards to morbidity and mortality. We found that increased extracellular volume fraction by CMR reduces voltage measures of conventional ECG criteria for LVH, including the Sokolow-Lyon index and Cornell indices. This may explain the limited sensitivity of the ECG in detecting LVH. We further investigated different patterns of LVH based on the relation between increased mass and wall thickness, and found that the different patterns differ on their electrocardiographic manifestation by A- ECG. Furthermore, A-ECG had a higher diagnostic performance compared to conventional ECG LVH criteria. The ECG detects electrical changes, while LVH represents a structural change. Therefore, the electrical changes associated with LVH may be better referred to as left ventricular electrical remodeling (LVER). LVER, defined as the A-ECG measure spatial QRS-T angle exceeding the upper limit of normal, was found to have a higher accuracy in diagnosing LVH compared to conventional ECG LVH criteria. We also found that patients with LVER have a worse prognosis compared to patients without LVER. Lastly, we optimized a score based on ECG and CMR measures, respectively, to predict morbidity and mortality, and found that ECG and CMR are both strong and independent predictors of events. In conclusion, conventional ECG criteria lack sensitivity in detecting LVH, which may be explained by increased extracellular volume fraction or different structural patterns in LVH. A-ECG has a higher diagnostic accuracy than conventional ECG criteria for LVH and is prognostic beyond CMR measures. Lastly, we suggest that LVER should be used when electrical changes in LVH are addressed.