Article Text

Download PDFPDF
11 CMR longitudinal strain analysis in aortic stenosis
  1. Nick B Spath1,
  2. Miquel Gomez2,
  3. Russell J Everett1,
  4. Scott I Semple3,
  5. Calvin WL Chin4,
  6. David E Newby1,5,
  7. Marc R Dweck1,5
  1. 1BHF/University Centre for Cardiovascular Science, University of Edinburgh, UK
  2. 2Hospital del Mar Medical Research Institute, Universitat Autònoma de Barcelona, Spain
  3. 3Clinical Research Imaging Centre, University of Edinburgh, UK
  4. 4Department of Cardiovascular Science, National Heart Centre, Singapore
  5. 5Department of Cardiology, Royal Infirmary of Edinburgh, UK


Introduction There is interest in measures of myocardial health in aortic stenosis (AS) to identify left ventricular (LV) decompensation and optimise timing of surgical intervention. CMR global longitudinal strain (GLS) is increasingly explored. We investigated the relationship of CMR GLS with established markers of AS severity, myocardial fibrosis and clinical outcomes.

Methods CMRs from AS patients (n=159) and healthy controls (n=42) were analysed using commercially available software. LV contours were drawn (long and short-axis cines). Automated analysis calculated 2D and 3D strain. Myocardial fibrosis was assessed using T1-mapping and late-gadolinium enhancement (LGE). Mortality was assessed at 1,466 days (median). Statistical analysis planning was published prior.

Results 2D-LVGLS correlated with AS severity (Vmax;r=0.24;p=0.0005), ejection fraction (EF, r=−0.33; p<0.0001) and was reduced in mid-wall (p=0.0002) and infarct LGE (p<0.0001). Extracellular volume fraction (ECV%) and indexed ECV were associated with 2D-LVGLS (r=0.16; p=0.02, r=0.42; p<0.0001). Compared to controls, 2D-LVGLS was unchanged in mild/moderate AS but lower in severe (p=0.02) and severe-symptomatic AS (p=0.002), although substantial overlap was seen across groups. 3D-LVGLS was unchanged between controls and AS, and no correlation with any parameter above was observed. Adjusted, iECV and EF were associated with reduced 2D-LVGLS (beta 0.08; p=0.001 and beta −0.13; p<0.001), similarly for 3D-LVGLS (beta 0.09; p=0.05 and beta −0.25; p<0.001). No mortality difference was demonstrated between 2D-LVGLS tertile. EF was the sole mortality predictor (HR 0.93; 95% CI: 0.88 to 0.98).

Conclusion CMR 2D-LVGLS measured may be a functional manifestation of myocardial fibrosis and LV decompensation in AS. However, substantial overlap between groups was observed and 2D-LVGLS did not predict clinical outcomes.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.