Context: Ureteroscopic biopsy is the gold standard for the histopathologic diagnosis of urothelial carcinoma of the upper urinary tract.
Objective: To assess the accuracy of endoscopically obtained biopsy samples in diagnosing, grading, and staging urothelial carcinoma and correlate diagnostic findings to biopsy sample size.
Design: We retrospectively reviewed endoscopic biopsies of the ureter, renal pelvis, and ureteropelvic junction from 2008 to 2011. Biopsy diagnoses that were discordant with follow-up pathology and/or ureteroscopic impression were re-reviewed and samples were immunohistochemically analyzed.
Results: Endoscopic biopsies (n = 118) yielded a sensitivity of 85.4% for the ureter (n = 79), 77.8% for the renal pelvis (n = 37), and 100% for the ureteropelvic junction (n = 2). A specificity of 100% for all locations and a diagnostic accuracy of 98.3% were identified. The median sample size was 0.3 cm for true positives, 0.3 cm for true negatives, and 0.2 cm for false negatives with no statistical significance. We found that 87.1% of tumors diagnosed on biopsy had concordant grade and 60.0% had concordant pT stage with follow-up surgical resections (n = 43) and biopsies (n = 24). Biopsy samples with concordant tumor grades (mean = 0.6 cm) compared with follow-up resection were larger than biopsy samples with discordant grades (mean = 0.3 cm) (P = .04).
Conclusions: Though highly specific, endoscopic biopsy does provide a significant false-negative rate owing to both sampling and diagnostic errors when assessing the upper urinary tract for urothelial carcinoma. Tumor grading is accurate, particularly with larger tissue samples, but tumor staging is unreliable.