Abstract
BACKGROUND: Myocardial strain measurements are increasingly used in research and clinical practice. However, there are limited data on inter-modality agreement and reproducibility. We aimed to investigate the inter-technique agreement of transthoracic echocardiography (TTE) and cardiovascular magnetic resonance (CMR) imaging derived left atrial (LA) and left ventricular (LV) deformation parameters. METHODS: Subjects with or without cardiovascular disease were prospectively recruited and had TTE and CMR on the same day. Ten subjects with type 2 diabetes (T2D) had both scans repeated within two weeks for test-retest reproducibility assessment. Myocardial deformation analyses were undertaken including LA strain (LAS) corresponding to LA reservoir, conduit and booster pump phases, LV global longitudinal strain (GLS) and peak early/late diastolic strain rate (PE/PLDSR) and LV mid-circumferential strain (Mid-CS) and strain rates. RESULTS: 222 participants (T2D (n = 87); severe aortic stenosis (n = 78) and healthy volunteers (n = 57)) were included. There were no significant differences between TTE and CMR measured LAS parameters, with moderate agreement between imaging modalities (ICC = 0.55-0.69). LV parameters were significantly higher on CMR except for Mid-CS which was higher on TTE (-19.3 ± 3.19 vs. -23.0 ± 4.37; p < 0.001). Inter-technique agreement was poor for all LV deformation parameters, except PLDSR with modest agreement (ICC = 0.52-0.66). CMR test-retest reproducibility was good to excellent for LAS and LV strain rate parameters (ICC = 0.73-0.90). TTE test-retest reproducibility was good for conduit LAS and LV_PEDSR (ICC = 0.80). CONCLUSION: There is modest agreement between TTE and CMR for LAS and poor agreement for LV strain assessment, suggesting that these techniques cannot be used inter-changeably. In a small subset of participants CMR test-retest reproducibility was overall better than TTE.