Intraarterial gadolinium-enhanced magnetic resonance angiography of the renal arteries in humans: feasibility, contrast agent reduction, and accuracy for detection of stenoses.

Objective: Feasibility of intraarterial MR angiography of the renal arteries and comparison of the accuracy of intraarterial MR angiography with selective intraarterial digital subtraction angiography (DSA) for detection of stenoses.

Materials And Methods: Ten consecutive patients (mean, 68 years) with suspected renal artery stenosis underwent a digital subtraction angiography and an intraarterial gadolinium-enhanced MR angiography, performed on a 1.5-T system. For intraarterial MR angiography 60 ml diluted contrast agent (10 ml gadodiamide in 50 ml 0.9% saline solution) was injected through a conventional angiography catheter placed in the suprarenal abdominal aorta using a flow rate of 3.5 ml/s. A three-dimensional (3D) gradient-echo sequence was performed. Differences in the quantitative measurement of stenoses of lesions between DSA and intraarterial MR angiography were evaluated by three observers. Overall impression of the intraarterial MR angiography was documented on a four-point scale (1 = excellent to 4 = poor). Interobserver variability was calculated.

Results: Intraarterial MR angiography of the renal arteries was feasible in all patients (100%) with a mean overall impression of all images of 1.8 (SD: 0.71). One of 9 accessory renal arteries was not visualized with intraarterial MR angiography. The overall sensitivity/specificity for detection of significant stenoses (>or=50% stenosis) were 83%/87%. Interobserver variability of intraarterial MR angiography ranged between fair and substantial (0.359-0.622).

Conclusion: Intraarterial MR angiography of the renal arteries in humans is feasible and has an acceptable sensitivity in detecting stenoses using injections of diluted contrast agent at concentrations as low as 17%.