Background: The Paris System (TPS) introduced diagnostic criteria for urine cytology to improve reproducibility among pathologists. Thus far, most cytology studies have investigated application of TPS on lower urinary tract specimens. Also, it is unclear which cytologic features are most predictive of malignancy, particularly in the upper urinary tract. We evaluate concordance rates of preoperative upper urinary tract cytology specimens before and after application of TPS criteria with surgical resections and assess cytologic features associated with malignancy.
Design: 54 resections with high- and low-grade urothelial carcinoma (HGUC, LGUC) from 2000-2016 with available preoperative cytology (n = 61) were identified. Cytology was re-reviewed to evaluate cytologic features and provide diagnoses before and after TPS implementation.
Results: The most common cytologic features associated with HGUC were N:C ratios ≥0.7 (88%), hyperchromasia (83%), coarse chromatin distribution (67%), and nuclear pleomorphism in cell clusters (65%). Application of TPS criteria resulted in 10 diagnostic downgrades and 6 diagnostic upgrades. After TPS criteria were applied, the sensitivity of a positive diagnosis decreased from 29% to 19%. The morphologic feature most consistently associated with a downgrade from positive to suspicious was a lack of marked nuclear contour irregularities in atypical urothelial cells.
Conclusion: Using strict TPS criteria in upper urinary tract cytology specimens may decrease the frequency of positive diagnoses with a concurrent increase in suspicious diagnoses. These findings may indicate that different morphologic features, particularly markedly irregular nuclear contours, may have different predictive values for HGUC in upper urinary tract cytology specimens compared with those from the lower urinary tract.
Keywords: Paris system; upper urinary tract; urine; urine cytology.
© 2018 Wiley Periodicals, Inc.