Improved diagnostic precision of urine cytology by implementation of The Paris System and the use of cell blocks.

Abstract

The Paris System for Reporting Urinary Cytology (TPS) published in 2016 provides a standardized approach to evaluating urine cytology. In the authors' practice, a TPS-like approach was adopted in 2012 using similarly defined cytologic criteria and correlating cystoscopic findings, and they also began incorporating the use of cell block (CB) material. The objective of the current study was to assess whether this TPS-like approach with the use of CB, as well as direct implementation of TPS, improved the diagnostic value of urine cytology. In total, 188 consecutive urine cytology specimens from 2010 through 2016 that had concurrent or subsequent histologic specimens available were retrospectively analyzed for diagnostic correlation. Urine cytology performance was compared between the periods 2010 to 2012 (pre-TPS-like), 2013 to 2016 (TPS-like), and after TPS reclassification, including blind review by a cytopathologist. The contribution of available CB material to final diagnoses also was assessed. Both the TPS-like approach and TPS reclassification resulted in significantly lowering the rate of atypical urothelial cells (AUC) diagnosis from 48%, to 21%, to 9% (from pre-TPS, to TPS-like, to after TPS reclassification, respectively; P < 0.01) while increasing the positive predictive value of an AUC diagnosis for high-grade urothelial carcinoma from 21%, to 43%, to 83%, respectively. The use of CBs contributed to a definitive final diagnosis in 24 of 36 cases (66.7%). In support of the new TPS recommendations, the application of stringent "TPS-like" and TPS criteria improves the value of an AUC diagnosis for clinicians and patients by lowering the AUC rate and increasing the positive predictive value of AUC for high-grade urothelial carcinoma. CBs can be used to help resolve problematic cases for more definitive diagnostic categorization.