Peripheral arterial occlusive disease: 3.0-T versus 1.5-T MR angiography compared with digital subtraction angiography.

Abstract

To prospectively evaluate the diagnostic accuracy of 3-T versus 1.5-T contrast material-enhanced (CE) magnetic resonance (MR) angiography with high spatial resolution in patients who have peripheral arterial occlusive disease, with conventional digital subtraction angiography (DSA) serving as the reference standard. Institutional review board approval and written informed consent were obtained. DSA and standardized single-injection, three-station, moving-table CE MR angiography, with similar acquisition protocols and contrast agent doses at 3 T and 1.5 T, were consecutively performed in 19 patients (13 men and six women; mean age ± standard deviation, 67 years ± 9). Stenosis was scored visually in 500 arterial segments (97.5% of all available) in consensus by two radiologists in a blinded manner (the radiologists were unaware of the field strength and prior DSA and MR angiographic results and used randomized analysis order). Contrast-to-noise ratio was determined in the vascular tree of both legs. Statistical significance in stenosis scoring was evaluated by using generalized estimating equations. Contrast-to-noise differences were evaluated with paired t tests. Agreement between MR angiography and DSA was evaluated by using Fleiss-Cohen κ statistics. Both 3-T and 1.5-T CE MR angiography showed similar excellent agreement with DSA regarding stenosis classification (κ = 0.96 and 0.93, respectively). All sensitivity and specificity values exceeded 90%. Mean contrast-to-noise ratio was 3.0-4.2 times higher at 3 T than at 1.5 T. Standardized single-injection, three-station, moving-table 3-T CE MR angiography is reliable for classification of stenosis in patients suspected of having peripheral arterial occlusive disease, and diagnostic performance was similar to that seen with 1.5-T MR angiography. There was a significantly increased contrast-to-noise ratio for identical contrast agent dose at 3-T MR angiography.