Ureteral cancer: Complete ureteral obstruction on computed tomography urography is associated with periureteral fat invasion

Abstract

ObjectiveTo determine predictive factors for periureteral fat invasion of ureteral cancer on computed tomography (CT) urography. Materials and methodsConsecutive 77 patients with CT urography and surgically proven ureteral cancer were retrospectively evaluated. Three independent radiologists with different experiences in interpreting CT urography analyzed radiologic findings as follows: (a) periureteral infiltration (absent; equivocal; or present); (b) ureteral obstruction (complete obstruction or not); (c) lesion length; (d) lesion width; (e) hydronephrosis grade; and (f) renal atrophy (present or not). We analyzed univariate and multivariate logistic regression. We assessed sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of significant parameters. ResultsPeriureteral fat invasion was surgically identified in 33.8 % (26/77). For reader 1, periureteral infiltration (odds ratio [OR] = 3.330), complete ureteral obstruction (OR = 18.965), and renal atrophy (OR = 3.116) were significant in univariate analysis (p < 0.05), while only complete ureteral obstruction was significant in multivariate analysis (OR = 12.779; p = 0.018). For reader 2 (OR = 4.172; p = 0.019) and 3 (OR = 9.459; p = 0.035), only complete ureteral obstruction was significant in univariate analysis. Sensitivity, specificity, PPV, and NPV of complete ureteral obstruction were 96.2 % (25/26), 43.1 % (22/51), 46.3 % (25/54), and 95.7 % (22/23) for reader 1, 84.6 % (22/26), 43.1 % (22/51), 43.1 % (22/51), and 84.6 % (22/26) for reader 2, and 96.2 % (25/26), 27.5 % (14/51), 40.3 % (25/62), and 93.3 % (14/15) for reader 3, respectively. ConclusionComplete ureteral obstruction on CT urography is highly sensitive for extra-ureteral extension of ureteral cancer. Ureteral cancer without complete ureteral obstruction may be less likely to invade periureteral fat.