Validation of Feature Tracking Cardiac MRI for Myocardial Strain: The Dallas Heart Study

BACKGROUND: Prior studies using speckle tracking echocardiography or cardiac magnetic resonance imaging (CMR) with myocardial tissue tagging show several factors to be associated with higher (worse) left ventricular (LV) strain (e.g. African-American race and male gender). The study aims to assess the validity of a new technique, Feature-tracking (FT) CMR, by evaluating the associations between strain measured by FT CMR and demographics and cardiovascular risk factors previously shown to be associated with strain by more established techniques. METHODS: Left ventricular longitudinal (GLS) and circumferential (GCS) strain values were measured in participants from the Dallas Heart Study (DHS), a multi-ethnic, population-based probability sample of adults in Dallas County. Linear regression models were constructed with GCS and GLS as the outcome variables in separate models adjusted for cardiovascular risk factors and left ventricular MRI characteristics (stroke volume (SV), LV mass, end diastolic volume (EDV), and LV ejection fraction(LVEF)). RESULTS: An interim analysis was conducted on 1,134 MRI studies. Greater LV mass was associated with increased (worse) GCS and GLS (ß=0.21, p<0.0001; ß=0.14, p<0.0001, respectively), while higher SV (ß=-0.15, p<0.0001; ß=-0.24, p<0.0001, respectively) and higher LVEF (ß=-0.-0.50, p<0.0001; ß=-0.22, p<0.0001, respectively) were found to be associated with decreased (better) strain values of both types. Higher EDV was associated with increased (worse) GCS strain values (ß=0.18, p<0.0001) but decreased (better) GLS strain values (ß= -0.07, p=0.0279). Male gender was associated with increased (worse) GCS and GLS strain values (ß=0.23, p<0.0001; ß=0.28, p<0.0001, respectively), and African-American race was also similarly associated with increased strain values (ß=0.12, p=0.0004; ß=0.11, p=0.0011, respectively). CONCLUSION: FT CMR is a reliable method of measuring GCS and GLS. Known factors associated with higher (worse) strain in prior studies were reflected in the FT CMR data, as conversely were factors that associated with decreased (better) strain, such as higher LVEF and SV.