Background: Cardiac magnetic resonance imaging (CMR) has become a diagnostic modality that allows for prognostic risk stratification in various cardiac diseases. CMR derived detection of myocardial necrosis by late gadolinium enhancement (LGE) and assessment of left ventricular functional parameters such as left-ventricular ejection fraction (LVEF) have been proven to be significantly associated with outcome and prognosis. Our study focusses on the validation of specific thresholds for these parameters in a multi-center daily all-comers cohort of stable coronary artery disease (CAD) patients.
Methods: Multi-center data from tertiary high-volume CMR centers were pooled. Patients referred for viability testing for known or suspected CAD were enrolled. Functional parameters of both ventricles and myocardial necrosis were assessed. The primary endpoint was defined as cardiac death and non-fatal myocardial infarction. A multi-model approach was used for the evaluation of the predictive power of several LVEF thresholds and LGE.
Results: The study cohort consisted of 2422 patients. Median age was 66 years; 25.9 % were female. Median follow-up was 2.86 years. During the follow-up period, 187 primary endpoints occurred. On multi-model testing, optimal thresholds for LVEF could be defined at ≤50 and ≤35 %. The addition of LGE as categorical variable further lead to a significant improvement of each risk prediction model, whilst quantification of LGE affection had no additional prognostic impact.
Conclusion: LVEF thresholds at ≤50 and ≤35 % in combination with the assessment of LGE presence allows for excellent discrimination between low, mid and high prognostic risk in stable CAD.
Keywords: Cardiac magnetic resonance imaging; Late gadolinium enhancement; Left-ventricular ejection fraction; Prognosis; Risk stratification; Stable coronary artery disease.