Free-Breathing Accelerated Gadolinium-Enhanced MR Angiography in the Diagnosis of Renovascular Disease

The purpose of this study was to assess the feasibility and accuracy of accelerated free-breathing and breath-hold gadolinium-enhanced MR angiography of the main renal arteries compared with digital subtraction angiography.Renal MR angiograms and catheter angiograms of 47 patients (19 men, 28 women; mean age, 68.1 +/- 15.1 years; range, 28-86 years) were reviewed. Thirty-one of the patients underwent free-breathing and 16 underwent breath-hold MR angiography with the same accelerated multiphase imaging protocol. Images were analyzed for examination quality, percentage narrowing of the main renal artery, and visibility of the branches. Diagnostic values of MR angiography were calculated with catheter angiography as the standard of reference.Sixty-five arteries, 24 of which (37%) had49% narrowing, were evaluated in the free-breathing group, and 37 arteries, six of which (16%) had49% narrowing, were evaluated in the breath-hold group. Comparison with digital subtraction angiography showed 100% (24/24) sensitivity and 95% (39/41) specificity for49% renal artery stenosis and 88% sensitivity (15/17) and 100% (48/48) specificity for74% renal artery stenosis in the free-breathing group. In the breath-hold group, sensitivity was 100% (6/6) and specificity 97% (30/31) for49% renal artery stenosis, and sensitivity was 100% (5/5) and specificity 100% (32/32) for74% renal artery stenosis. None of the examinations was nondiagnostic for the main renal arteries, but a smaller number of visible arterial tree subdivisions were found in the free-breathing group (average, 3.64 per patient) than in the breath-hold group (average, 5.87 per patient) (p = 0.035).Like breath-hold examinations, accelerated free-breathing MR angiographic examinations are feasible and accurate in evaluation of the main renal arteries.