The Paris System for Urine Cytology: Prospective Assessment of Reproducibility, Diagnostic Accuracy, and Malignancy...

Urine cytology is widely used for the detection and surveillance of urothelial carcinoma but has historically suffered from variability in interpretation. The Paris System for Reporting Urinary Cytology (TPS) was developed to standardize reporting and improve clinical relevance by focusing on high-grade urothelial carcinoma (HGUC).

This prospective observational study included 170 urine samples from patients clinically suspected of urothelial neoplasia. Cytology smears were independently evaluated by two experienced cytopathologists and categorized according to TPS. Suspicious for HGUC (SHGUC) and HGUC were considered cytology-positive. Histopathology (transurethral resection of bladder tumor) served as the reference standard where available; radiologic findings with clinical follow-up were used otherwise. Interobserver agreement, diagnostic performance, and risk of high-grade malignancy (ROHM) for each TPS category were calculated with 95% confidence intervals.

Interobserver agreement was excellent across TPS categories (κ = 0.985, p < 0.001) and perfect after binary classification (κ = 1.000). Using the SHGUC threshold, urine cytology demonstrated high diagnostic accuracy against the combined reference standard, with sensitivity of 92.86% (95% CI, 80.99-97.54), specificity of 100% (95% CI, 97.09-100.0), and overall accuracy of 98.24% (95% CI, 94.94-99.40). ROM was 100.0% for both SHGUC and HGUC, 57.1% for atypical urothelial cells, and 0.8% for negative for HGUC. Haematuria, male sex, and increasing age were significantly associated with malignancy.

TPS provides highly reproducible, clinically meaningful urine cytology reporting with excellent specificity and strong sensitivity for HGUC. ROM stratification reinforces the clinical utility of SHGUC and HGUC categories and supports integrated interpretation of cytology with clinical and radiologic findings in routine practice.