Multimodality Assessment of Myocardial Deformation: A Comparison Between Transthoracic Echocardiography and Cardiovascular Magnetic Resonance Imaging

Background:

Myocardial deformation measurements are now possible from routinely acquired Transthoracic Echocardiography (TTE) and Cardiovascular Magnetic Resonance (CMR) images. Accordingly, they are increasingly used in research and clinical practice. Impaired left atrial (LA) and left ventricle (LV) strain parameters were shown to be associated with adverse events in multiple cardiac conditions.

Purpose:

To investigate the inter-technique agreement and test-retest reproducibility of TTE and CMR in assessing LA and LV myocardial deformation. To compare the ability of CMR and TTE to detect change in myocardial deformation post-intervention. To establish the reproducibility of the novel LA function assessment using CMR.

Methods:

Prospective and retrospective studies, that included all prospectively recruited participants in ethically approved research studies, were included in this thesis. The inter-technique agreement was assessed from TTE and CMR scans performed on the same day on the same cohort. The test-retest reproducibility assessment was assessed from same-day CMR+/-TTE on two separate occasions within 2-weeks using the same protocol and scanner. Two separate cohorts were used to assess the change post-intervention: participants with type 2 diabetes (T2D) who were randomised to a 12-week intervention of: (i) routine care, (ii) aerobic exercise training, or (iii) low-energy meal replacement plan (MRP),and had same-day CMR and TTE at baseline and week-12. Another cohort of patients with severe aortic stenosis (AS) who were listed to undergo surgical aortic valve replacement (AVR) had CMR and TTE performed on the same day pre- and median of 6 months post-AVR. LA and LV deformation assessments were conducted blindly on both TTE and CMR images.

Results:

A greater proportion of TTE image analysis was limited by poor image quality (≥10%) compared to CMR (%5). LA strain (LAS) by CMR had moderate to good test-retest reproducibility, with moderate agreement with TTE. Only conduit LAS showed good test-retest reproducibility on TTE. Almost all LV strain (LVS) and strain rate (SR) parameters showed poor inter-modality agreement. The reproducibility of LVSR parameters was similar at both CMR and TTE, whilst LVS was more reproducible by CMR. Post-lifestyle intervention in people with T2D, only CMR showed a significant increase in LAS at both reservoir and booster pump, and mid-circumferential LVS in the MRP group. 6 months post-AVR, both STE and FT-CMR showed significant changes in LA and longitudinal-LV deformation with some discrepancy between modalities in terms of LAS. However, circumferential-LV deformation was significantly increased using FT-CMR only.

Conclusion:

Overall, these results favour the use of CMR for assessing LA and LV myocardial deformation in longitudinal studies, due to its good reproducibility and sensitivity to detect change post-intervention, leading to smaller numbers needed to detect change. Moreover, the superior image quality of CMR compared to TTE leads to more analysable images and a more complete dataset.