Is echocardiography or magnetic resonance imaging superior for precoarctation angioplasty evaluation?

We compared the dimensions of the aorta obtained by two-dimensional transthoracic echocardiography (echo) (median, 2.5 mo preangioplasty) and magnetic resonance imaging (MRI) (median, 4.2 mo preangioplasty) to those obtained by angiography (cath) in 13 patients (age, 7.7 +/- 1.6 yr; mean +/- SEM) who underwent evaluation for coarctation balloon angioplasty between April 1993-January 1996. Echo measurements were obtained from the suprasternal and subcostal sagittal planes, MRI measurements from axial and sagittal oblique views, and cath measurements from the straight lateral or oblique views. Measurements of the diameters of the aortic isthmus, coarctation, descending aorta at the diaphragm, and isthmus length were made by all three modalities. Presence of aorto-aortic collaterals was determined, and the coarctation length was delineated. Investigators were blinded to other measurement data prior to statistical analysis. Data analysis by repeated analysis of variance (ANOVA) and Student-Newman-Keuls testing revealed no statistically significant difference between systolic pressure gradient by clinical examination (32.2 +/- 5.9 mm Hg), peak instantaneous Doppler evaluation (37.5 +/- 2.9 mm Hg), or preangioplasty systolic pressure gradient (32.1 +/- 3.3 mm Hg). With the exception of measurements of the descending aorta (echo, 11.7 +/- 0.9 mm vs. MRI, 13.3 +/- 0.8 mm vs. cath, 14.0 +/- 1.3 mm; P = 0.04), there was no statistically significant difference in dimensions of the aortic isthmus (9.2 +/- 0.6 mm vs. 10.5 +/- 0.9 mm vs. 10.8 +/- 0.9 mm), coarctation site diameter (4.8 +/- 0.6 mm vs. 5.6 +/- 0.9 mm vs. 5.3 +/- 0.8 mm), or isthmus length (12.4 +/- 2.1 mm vs. 12.1 +/- 2.2 mm vs. 10.9 +/- 1.7 mm). The correlation coefficients derived from comparisons of MRI vs. cath to echo vs. cath were similar for all dimensions except for isthmus length (P < 0.01). MRI demonstrated aorto-aortic collaterals more frequently than echo, while echocardiography better demonstrated cardiac function and intracardiac anomalies. Both modalities detected the single patient with a long segment coarctation. We conclude that echocardiography and MRI provide similar definition of the coarcted thoracic aorta, and either test may be individually advantageous in specific clinical situations.