Intrarenal arterial Doppler sonography in the detection of renal vein thrombosis of the native kidney

OBJECTIVE: Previous studies of transplant kidneys and recent reports on native kidneys have suggested intrarenal arterial Doppler findings can be helpful in the noninvasive workup of renal vein thrombosis. We used arterial Doppler sonography to evaluate cases of possible acute renal vein thrombosis in native kidneys that had equivocal results on standard Doppler analysis of the renal vein. MATERIALS AND METHODS: Twenty native kidneys in 12 patients with clinical findings suggestive of acute renal vein thrombosis had Doppler studies of the main renal vein that failed to show normal flow. In all 20 kidneys, duplex Doppler study of arcuate/interlobar intrarenal arteries was done and the resistive index was determined. The Doppler findings were compared with subsequent findings on either renal venograms (n = 11) or MR images (n = 9), which served as the reference "gold" standards. RESULTS: The prevalence of renal vein thrombosis was 25% (5/20). Ten kidneys had very abnormal findings on arterial Doppler studies (absent or reversed end-diastolic flow), but only two of these were proved to have renal vein thrombosis. In six other kidneys, end-diastolic flow was identified but the resistive index was still elevated (> or = 0.70), and only one of these kidneys was proved to have renal vein thrombosis. Four kidneys had normal arterial Doppler studies, and 50% (two) of these were proved to have renal vein thrombosis. When absent or reversed end-diastolic flow was used as a sign of renal vein thrombosis, intrarenal arterial Doppler analysis had a sensitivity of 40% (2/5) and a specificity of 47% (7/15). CONCLUSION: Unlike the reported experience in transplanted kidneys, intrarenal arterial Doppler analysis is neither sensitive nor specific for renal vein thrombosis in native kidneys. An intrarenal arterial Doppler study with normal findings should not prevent further workup if Doppler findings in the renal vein are equivocal, nor should absent or reversed end-diastolic arterial signals be considered highly suggestive of renal vein thrombosis.