Login to Access Video or Poster Abstract: MP78-11
Sources of Funding: None

Introduction

CT Urography (CTU) is the initial imaging modality of choice for assessing upper urinary tract pathology, including upper tract urothelial carcinoma (UTUC). However, despite abnormal findings on CTU suggestive of UTUC, follow-up diagnostic ureteroscopy (DURS) is frequently negative. We aim to assess the positive predictive value (PPV) of CTU for UTUC using DURS as the reference standard.

Methods

The study group comprised 79 consecutive patients from 2 academic institutions. They had CTU findings of upper tract wall thickening, hydronephrosis, a filling defect and/or contrast enhancement diagnosed in consensus by 2 radiologists as suspicious for UTUC. DURS, with either wash cytology when endoscopically negative or biopsies when endoscopically positive, was used as the reference standard. The results of DURS were classified as UTUC, benign lesions (BL) and no pathological findings (NPF). Statistical analysis was conducted.

Results

Solitary CTU suspected findings were reported in 45 (57%) patients, including thickness in 8 (10%), hydronephrosis in 5 (6%), filling defect in 30 (38%) and enhancement in 2 (3%). Combinations of suspected findings were reported in 34 (43%) patients. DURS revealed 41 (52%) UTUC, 14 (18%) BL and 24 (30%) NPF. Table shows the calculated PPV. _x000D_ The combination of CTU findings had higher PPV in comparison to solitary findings for detection of UTUC (65% vs 42%, respectively; p<0.05, OR 2.5, 95% CI 1.007-6.28) as well as for overall endoscopic pathological findings (82% vs 60%, respectively; p<0.05, OR 3.1, 95% CI 1.07-9.02). Arbitrary stratification of solitary CTU findings as minor UTUC predictors (PPV<50%: thickening, hydronephrosis and enhancement) and major UTUC predictors (PPV≥50%: filling defect) resulted in a statistically significant better prediction for the major group (p<0.05; OR 7.2, 95% CI 1.39-38.15).

Conclusions

The best PPV of CTU diagnosis of UTUC is achieved based on a combination of findings, with filling defect appearing to be the most significant among them. In the absence of filling defect, other CTU findings, such as thickening, hydronephrosis and enhancement, are not predictive for UTUC. We suggest that the need for DURS in these cases should be re-considered in correlation with other data (e.g., cytology, biomarkers, history of heavy smoking, recurrent hematuria, etc.). DURS remains the diagnostic standard for deciding whether or not to proceed to nephroureterectomy.

Funding

None