[Evaluation of left ventricular volumes in patients with congenital heart disease and abnormal left ventricular geometry. Comparison of MRI and transthoracic 3-dimensional echocardiography].

PURPOSE: To assess the new method of 3-dimensional echocardiography in comparison to the "gold standard" MRI as to its ability to calculate left ventricular volumes in patients with congenital heart disease.

MATERIALS AND METHODS: Eighteen patients between the ages of 3.9 to 37.3 years (mean: 12.8 +/- 9.7) with a geometrically abnormal left ventricle were examined using a 1.5T scanner with a fast gradient-echo sequence (TR = 14 ms, TE = 2.6 - 2.9 ms, FOV = 300 - 400 mm, flip angle = 20 degrees, matrix = 128 : 256, slice thickness = 5 mm, retrospective gating) in multislice-multiphase technique. Transthoracic 3D-echocardiography was performed with a 3.5 MHz transducer and a Tomtec (Munich, Germany) system for 3D reconstruction.

RESULTS: Volume calculation was possible in all patients with 3D-echocardiography, but the muscle mass calculation only succeeded in 11 of 18 patients (61 %) due to inadequate visualization of the entire myocardium. Comparing MRI and 3D-echocardiography, the correlation was r = 0.97 for the end-systolic volumes, r = 0.98 for the end-diastolic volumes, r = 0.79 for the end-systolic muscle mass and r = 0.77 for the end-diastolic muscle mass. The agreement between both methods was considered good for the calculated end-diastolic volumes and sufficient for the calculated end-systolic volumes. The muscle mass calculations showed larger differences especially for the end-systolic mass. Mean intraobserver variability was 18.6 % for end-systolic and 8.3 % for end-diastolic volumes.

CONCLUSION: In patients with an abnormal left ventricular configuration due to congenital heart disease, the new method of 3D-echocardiography is sufficient for volume calculations in preselected patients. The high intraobserver variability is still a limitation of transthoracic 3D-echocardiography in comparison to MRI.