Screening patients for renal artery stenosis: value of three-dimensional time-of-flight MR angiography

The purpose of this study was to determine the value of three-dimensional time-of-flight MR angiography in screening patients for the presence of renal artery stenoses.

Fifty-three patients who were thought to have renovascular hypertension on clinical grounds were prospectively examined with three-dimensional time-of-flight MR angiography. IV digital subtraction angiograms and duplex sonograms were available for all patients. For 21 patients in whom IV angiograms were of high quality and showed the renal arteries to be normal, as confirmed with duplex sonograms, intraarterial digital subtraction angiograms were not obtained for MR correlation. In the other 32 patients, intraarterial angiograms of the abdominal aorta were obtained. MR angiograms were interpreted independently by two radiologists who were unaware of the findings on angiography and duplex sonography. Our preliminary clinical experience suggested that a signal loss in the renal arteries on maximum-intensity-projection MR angiograms indicated a potential stenosis and that the degree of stenosis could not be measured accurately with MR angiography. The search for stenoses was focused on the proximal and middle parts of the vessel, as far as 3 cm from the origin of the vessel. We used intraarterial angiography to measure and grade renal artery stenoses. On intraarterial angiograms, stenoses that involved more than 50% of the vessel's section were considered significant (n = 24) and stenoses that involved 50% or less of the section were considered insignificant (n = 7). These 31 stenoses were found in 26 of the 32 patients who had intraarterial digital subtraction angiography.

All 24 significant stenoses appeared as a cutoff of signal intensity on MR angiograms. Results were false-positive in 20 cases: in eight cases, arteries were of small diameter; in seven cases, the stenoses were insignificant; in two cases, the acquisition volume was incorrectly positioned; in two cases, there was a sharp angle in the proximal part of the renal artery; and in one case, no explanation was found. MR angiography had a sensitivity of 100% for detecting stenoses in the proximal 3 cm of the renal artery; stenoses more than 3 cm from the origin of the arteries could not be detected reliably.

Our results suggest that three-dimensional time-of-flight MR angiography is a simple method for detecting stenoses of the proximal portion of the main renal arteries. However, more work is needed before widespread clinical use of such a technique is feasible.