Split-bolus CT urogram: Is less more?

To determine the accuracy of split-bolus CT urogram (SB-CTU) without special maneuvers for ureteral distention in diagnosing upper urinary tract urothelial carcinoma (UCA).

A hospital database was searched from 1/1/10, to 9/1/15, for SB-CTU exams without special maneuvers for ureteral distention. Accuracy of SB-CTU for detecting upper and lower urinary tract UCA was computed by comparing the prospective radiology report with cystoscopy, ureteroscopy, and/or urologic clinical follow-up. Patients with less than 12 months of clinical follow-up were excluded.

339 studies were identified in 334 patients (60% male 40% female, avg. age 64). 119 studies were performed for microhematuria, 150 for gross hematuria, 13 for hematuria not otherwise specified, 57 for history of UCA, and one for a collecting system mass on a prior CT. There were five upper tract and 33 bladder tumors with overall prevalence of 1.5% and 9.7%, respectively. The prevalence varied significantly with patient age and clinical indication for SB-CTU. There were one false negative and four false positives for upper urinary tract UCA. Sensitivity, specificity, positive predictive value, and negative predictive value for detecting upper tract and bladder tumors were 80%, 99%, 44%, and 100%, respectively, and 55%, 98%, 78%, and 95%, respectively.

Based on this study, SB-CTU without special maneuvers for ureteral distention is highly sensitive for detecting upper tract UCA, although with a low positive predictive value, false positives do occur. The clinical utility of increasing accuracy in diagnosing this low-prevalence disease through other more complex CT urogram protocols should be weighed against the added cost and radiation dose associated with these protocols.