Diagnosis of ureteral obstruction in patients with compromised renal function: the role of noninvasive imaging modalities.

We compared the role of noncontrast computerized tomography (NCCT), magnetic resonance urography (MRU), and combined abdominal radiography (KUB) and ultrasonography (US) in the diagnosis of the cause of ureteral obstruction in patients with compromised renal function.

METHODS

The study included 149 patients, of whom 110 had bilateral obstruction and 39 had obstruction of a solitary kidney. Therefore, the total number of renal units was 259. All patients had renal impairment with serum creatinine greater than 2.5 mg/dl. Besides conventional KUB and US all patients underwent NCCT and MRU. The gold standard for diagnosis of the cause of obstruction included retrograde or antegrade ureterogram, ureteroscopy and/or open surgery. The sensitivity, specificity and overall accuracy of NCCT, MRU, and combined KUB and US in the diagnosis of ureteral obstruction were calculated in comparison with the gold standard.

RESULTS

The definitive cause of ureteral obstruction was calculous in 146 and noncalculous in 113 renal units, including ureteral stricture in 65, bladder or ureter in 43, extraurinary collection in 3 and retroperitoneal fibrosis in 2. The site of stone impaction was identified by NCCT in all 146 renal units (100% sensitivity), by MRU in 101 (69.2% sensitivity), and by combined KUB and US in 115 (78.7% sensitivity) with a difference of significant value in favor of NCCT (p <0.001). Ureteral strictures were identified by NCCT in 18 of the 65 cases (28%) and by MRU in 54 of 65 (83%). Bladder and ureteral tumors causing ureteral obstruction could be diagnosed in approximately half of the patients by NCCT (22 of 43) and in all except 1 by MRU (42 of 43). NCCT and MRU could identify all extraurinary causes of obstruction. Overall of the 113 kidneys with noncalculous obstruction the cause could be identified by MRU in 101 (89% sensitivity), by NCCT in 45 (40% sensitivity), and by combined KUB and US in only 20 (18% sensitivity) with a difference of significant value in favor of MRU (p <0.001).

CONCLUSIONS

In patients with renal impairment due to ureteral obstruction NCCT has superior diagnostic accuracy for detecting calculous causes of obstruction but MRU is superior for identifying noncalculous lesions.