It has been reported and also has been our preliminary experience that many false ostial stenoses are attributable to a loss of signal intensity at the origin of the renal arteries when three-dimensional (3D) phase-contrast MR angiography is used. Our objective was to add a 3D time-of-flight MR angiography sequence to the 3D phase-contrast MR angiography sequence to better analyze the origin of the main renal arteries. We assessed the value of the combination of these two MR angiography sequences for the depiction of renal artery stenosis. Forty-six patients suspected of having renal artery stenosis on the basis of clinical history, physical examination, and laboratory data were prospectively enrolled. Intraarterial digital subtraction angiography findings were available for all patients. Using intraarterial digital subtraction angiography, we considered stenosis to be significant when the vessel was narrowed more than 50%. During MR angiography, half of the data were reconstructed by interpolation to avoid long acquisition times. Total acquisition times were less than 15 min. MR angiography findings were interpreted independently by two radiologists who were unaware of the findings of intraarterial digital subtraction angiography. With 3D phase-contrast MR angiography, any cutoff in signal intensity or any narrowing of the vessel diameter of more than 50% from the renal ostium to the renal hilum was considered to represent significant stenosis. With 3D time-of-flight MR angiography, our image analysis was focused on the origin of the arteries. Any cutoff in signal intensity in the first centimeter of the renal artery was considered to represent significant stenosis. Intraarterial digital subtraction angiography showed 105 renal arteries, including 15 supernumerary renal arteries. Eleven stenoses were localized to the main hilar renal arteries. Using time-of-flight MR angiography, we found that polar supernumerary renal arteries of small caliber and intrarenal branches of renal arteries were not adequately displayed. Using phase-contrast MR angiography to evaluate only whether the main hilar renal arteries were stenotic, we calculated the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy to be 100%, 65%, 28%, 100%, and 69%, respectively. Using a combination of the two imaging sequences, we found that the specificity, positive predictive value, and accuracy were increased to 90%, 58%, and 92%, respectively. For detecting stenoses of the main renal arteries but not for visualizing small accessory renal arteries or distal branches, our results support the use of a combination of the two MR angiography sequences. For now, this combination of sequences should be viewed primarily as a technique for screening patients.
Read full abstract