MR imaging (including MR angiography) of abdominal aortic aneurysms: Comparison with conventional angiography

OBJECTIVE. The purpose of this report is to compare the results of MR imaging, including MR angiography, with the results of conventional angiography in the imaging of abdominal aortic aneurysms (AAA). SUBJECTS AND METHODS. Twenty seven nonconsecutive patients with known AAA underwent both MR imaging (including MR angiography) and conventional angiography preoperatively. MR imaging was done before angiography according to the following protocol in all patients: sagittal and coronal T1-weighted images of the abdomen, followed by axial MR angiograms obtained with a multiple overlapping thinslab acquisition (MOTSA) three-dimensional time-of-flight (3D TOF) volume centered on the renal arteries. Last, coronal MR angiograms were obtained with a dynamic contrast-enhanced 3D TOF volume centered on the aorta. Conventional angiography was subsequently performed in all patients. Data from each study were collected prospectively and analyzed in a blinded manner. For the MR studies, all sequences were used to determine the extent of the AAA and the presence of iliac aneurysms; stenotic disease of the celiac, superior mesenteric, renal, and iliac arteries and the number and location of renal arteries were determined from the MR angiograms. For evaluation of the extent of the aneurysm, results from both techniques were compared with surgical results; otherwise MR angiography was compared with conventional angiography. Confidence intervals for sensitivity and specificity were calculated at the 95% level. The McNemar test was used for comparison of proportions. RESULTS. MR angiography had a sensitivity of 100% (4/4) and a specificity of 91% (21/23) for the four stenoses of the celiac artery that were greater than 50%. For detection of renal artery stenoses of greater than 50%, present in nine arteries, the sensitivity was 89% (8/9) and specificity was 98% (49/50). MR angiography showed 100% (54/54) of all main renal arteries and 78% (7/9) of all accessory renal arteries seen on conventional angiograms. The proximal extent of the aneurysm was correctly predicted on the basis of MR findings in 100% (26/26) of patients with surgical correlation, compared with 92% (24/26) when conventional angiograms were used. This difference was not statistically significant (p > .1). MR imaging had a sensitivity of 100% (19/19) for detection of isolated aneurysms of the common iliac artery, and MR angiography had a 100% sensitivity (5/5) and specificity (103/103) for detection of stenoses greater than 50% in the common and external iliac arteries. One circumaortic left renal vein was visualized on MR images. CONCLUSION. These results suggest that MR angiography may be a useful noninvasive method of determining the presence of coexistent celiac, superior mesenteric, renal, and iliac atherosclerotic disease in patients with AAA.