Reproducibility of thoracic and abdominal aortic wall measurements with three-dimensional, variable flip angle (SPACE) MRI.

PURPOSE: To evaluate the reproducibility and repeatability of high-resolution, isotropic thoracic and abdominal aortic wall measurements, and determine the implications they have on the number of subjects necessary for future clinical trials.
MATERIALS AND METHODS: Using a T1-weighted three-dimensional MRI SPACE sequence, we evaluated the interobserver, intraobserver, and scan-rescan variability of isotropic thoracic and abdominal aortic wall measurements in 15 cardiovascular diseased patients and 6 normal volunteers. Main outcome analyses were intracorrelation coefficient (ICC), mean relative error (mRE), and sample size calculation at 80% power to be used to compare placebo group and treatment group means in future two-arm randomized clinical trials.
RESULTS: Excellent reliability, ICC > 0.8 (P < 0.001) and small mRE < 10% were demonstrated for the interobserver, intraobserver, and scan-rescan variability for all investigated measures: lumen area (LA), outer wall area (OWA), wall area (VWA), total wall volume (TWV), and percentage wall volume (%WV). Sample size calculation revealed slightly different sample size per treatment arm for thoracic and abdominal aorta segments (maximum number of subjects: 352 subjects for thoracic segment versus 421 subjects for abdominal segment for LA at 5% difference, and minimum of 3 thoracic versus 4 abdominal subjects needed for %WV evaluation at 25% difference).
CONCLUSION: Our study demonstrates the reproducibility and repeatability of SPACE aortic plaque measurements, and gives insight into the number of subjects needed for the design of therapeutic studies in aortic atherosclerosis.