Renal duplex sonography: main renal artery versus hilar analysis

The purpose of this study was to compare the accuracy of main renal artery Doppler scanning interrogation and hilar analysis to diagnose hemodynamically significant renal artery disease.

From January 1998 to August 1999, 41 patients had renal duplex sonography with both main renal artery interrogation and hilar analysis followed by angiography. They form the basis of this review. The sample consisted of 24 men and 17 women, with a mean age of 68.9 +/- 10.2 years, who provided 80 kidneys for comparative analysis. Significant renal artery disease revealed through angiography was defined as >/= 60% diameter-reducing stenosis or occlusion. Peak systolic velocity (PSV) (in meters per second) and the presence of poststenotic turbulence (PST) were determined from main renal artery interrogation. Acceleration time (AT) (in milliseconds) was measured by means of hilar analysis. Significant renal artery stenosis was defined by a PSV of 2.0 m/s or more and a PST or an AT more than 100 ms. Sensitivity analyses of both PSV and AT were examined, and 95% CIs were computed. Receiver operating characteristic curves were used to estimate optimal values for PSV and AT.

Angiography revealed hemodynamically significant fibromuscular dysplasia in 5 kidneys (4 patients), atherosclerotic stenosis >/= 60% in 48 kidneys (30 patients), and renal artery occlusion in 4 kidneys (4 patients). Kidneys with significant renal artery stenosis had a higher PSV (2.54 +/- 0.11 vs 1.28 +/- 0.08, P <.001) and AT (82.43 +/- 7.2 vs 30.0 +/- 2.8, P <.001) compared with those without stenosis. Compared with angiography, a PSV of 2.0 m/s or more and PST demonstrated a sensitivity of 91%, specificity of 96%, and overall accuracy of 92% for detection of significant renal artery stenosis. Two of five studies with false-negative results reflected diseased polar vessels. By contrast, AT of more than 100 ms had a sensitivity of 32%, specificity of 100%, and overall accuracy of 54%. Receiver operating characteristic curve analysis revealed a PSV of more than 1.8 m/s and an AT of 58 ms or greater as optimal values. With an AT of 58 ms or more, the sensitivity was 58%, and specificity was 96%, with an overall accuracy of 70%. There were no apparent associations between PSV or AT and type or location of renal artery lesion, serum creatinine level, or end-diastolic ratio.

Main renal artery interrogation is an accurate screening test to detect significant stenosis or occlusion of the main renal artery. Hilar analysis alone does not provide sufficient sensitivity to be used as a sole screening study. Neither method detects the presence of renovascular disease associated with polar vessels.