A Comparative Study of the Paris System and Common Reporting System for Urine Cytology With Cyto-Histology Correlation: A Study of 829 Urine Cytology Specimens.

Abstract

The Paris System (TPS) diligently detects high-grade urothelial carcinoma (HGUC) and creates a uniform, standardized, reproducible reporting system for urine cytology. However, many centres might still use a common reporting system (CRS). The study aims to compare TPS and CRS for urine cytology with histology correlation. It was a cross-sectional study done from July 2016 to December 2022. The study included 829 urine cytology samples (96% voided urine) from 478 patients. Histology correlation was available for 138 (16.6%) samples of 115 patients. The frequency of NHGUC, AUC, HGUC and SHGUC was 40.6%, 17.4%, 12.2% and 5.5%, respectively, in TPS. In contrast, in CRS, the frequency of NM, AUS, SM and PM was 69.2%, 13.3%, 4.5% and 13.0%, respectively. TPS and CRS had 64% agreement overall with the kappa test (κ-value 0.479, moderate strength). The agreement between TPS and CRS was 39.8% for NHGUC, 10.97% for AUC and 10.85% for HGUC. After combining a few TPS categories, the agreement increased to 87.7% (κ-value 0.7640, good strength). Histological concordance for AUC, HGUC and NHGUC was 75%, 31.8% and 31.3% in TPS, and it was 50% and 33.3% for AUS and PM, respectively, in CRS. The sensitivity and specificity of TPS and CRS against histology were 37.5% vs. 26.0%, p = 0.0005 and 76.5% vs. 85.3%, p = 0.0083, respectively. TPS and CRS have moderate strength of agreement for urine cytology. TPS was more sensitive than CRS. It may be easy for institutes to transition to a newer TPS system if they still use a CRS.