Objective: We compared the assessment of left ventricular function and mass by M-mode echocardiography (echo) with fast breath-hold cardiovascular magnetic resonance (CMR) in patients who received orthotopic heart transplantation. We also sought to establish the reproducibility of breath-hold CMR in this patient population.
Methods: We prospectively acquired 51 sets of echo and CMR data in 21 patients who had undergone orthotopic heart transplantation. We examined the intraobserver and interobserver reproducibility of breath-hold CMR in this group and compared it with published data. We compared the left ventricular ejection fraction (EF) and mass determined by echo with the CMR data.
Results: The average time between CMR and echo was 0 +/- 7 days (mean +/- SD), the time between each set of CMR-echo data acquisition was 5.1 +/- 4.1 months. Cardiovascular magnetic resonance showed good reproducibility in this population, with intraobserver percentage variability of 2.2% +/- 2.4% for EF and 3. 2% +/- 2.7% for mass, and interobserver percentage variability of 2. 4% +/- 1.9% for EF and 2.2% +/- 1.9% for mass. The Bland-Altman limits of agreement between echo and CMR were wide for both EF (-9. 6% to 15%) and mass, irrespective of the formula used (-61.3 to 198 g for the Bennett and Evans formula, -65.4 to 196.8 g for the American Society of Echocardiography (ASE) formula, -65.3 to 181 g for the Devereux formula, and -95.2 to 64.6 g for the Teichholz formula).
Conclusion: Fast-acquisition CMR is reproducible in recipients of transplanted hearts. We found poor agreement with the results of echo. The choice of technique will depend on local resources as well as the clinical importance of the result. Echo remains readily available and gives rapid assessment of volumes, EF, and mass. However, the good reproducibility of CMR may make it a more suitable technique for long-term follow-up of an individual or of a study population.