RESUMO
The Paris System (TPS) for reporting urinary cytology attempts to unify the terminology in this field. OBJECTIVES: To analyze the impact of adopting TPS by measuring nomenclature agreement and cytohistological correlation. MATERIALS AND METHODS: Voided urine liquid-based cytology samples corresponding to 149 biopsy-proven cases (76 high-grade carcinomas, 40 low-grade carcinomas, and 33 benign lesions), were reclassified by the same pathologist using TPS. Diagnostic agreement and sensitivity for both nomenclature systems was measured. RESULTS: When using TPS, the rate of atypical samples increased 8 times (from 3 to 24.2%) in benign cases, 10 times (from 2.5 to 25%) in low-grade carcinomas, and 2.4 times (from 6.6 to 15.8%) in high-grade carcinomas. The false-positive rate (abnormal cytology in negative or low-grade carcinoma cases) increased from 11 to 34.2%. Sensitivity was higher (63 vs. 49%) with TPS at the expense of a lower specificity (73 vs. 91%). The agreement between both nomenclatures was moderate for negative and high-grade carcinoma cases (k = 0.42 and 0.56, respectively) and weak for low-grade tumors (k = 0.35). CONCLUSIONS: Adopting TPS for reporting urine cytology results in a considerable increase in atypical diagnoses, improving sensitivity but lowering specificity. Appropriate management recommendations for patients with an atypical cytological diagnosis are required.
