RESUMO
INTRODUCTION: The Paris System (TPS) has recently been used in classification of urinary tract cytological specimens. Upper urinary tract (UUT) specimens are cytologically challenging. The utility of TPS was investigated in evaluating UUT specimens and its correlation with subsequent histological follow-up. METHOD: From 2014 to 2017, 324 cytology cases of UUT from 179 patients were retrieved. Concurrent or subsequent biopsy or resection within a 2-month period was available in 125 cases from 74 patients. RESULT: None of the cases with a cytology of low-grade urothelial neoplasm was found to have a high-grade urothelial carcinoma (HGUC) on biopsy. Among the 19 atypical urothelial cells (AUC) cytology cases, the histology is heterogeneous (seven benign, one atypia, five low-grade lesion, and six HGUC). The risk of HGUC for each cytological diagnostic category are 0% for non-diagnostic/unsatisfactory, 6% for negative for HGUC, 27.3% for AUC, 0% for low-grade urothelial neoplasm, 48% for suspicious for HGUC and 95% for positive HGUC. When we considered cytology cases with suspicious or positive for HGUC interpretations as positive, the performance of TPS in predicting high grade urothelial carcinoma on histology had values of: 78.6% sensitivity, 86% specificity, 80.5% positive predictive value and 84.5% negative predictive value. CONCLUSION: More than one-third of the UUT cytological cases were classified as AUC and approximately 1/15 as suspicious or positive for HGUC. Based on UUT cytology specimens, the risk of malignancy of each cytological diagnostic category of TPS was comparable to those reported in the literature. The use of TPS in evaluating UUT cytology specimens was specific and sensitive in identifying patients with HGUC by histology.