Comparison of urinary cytology and fluorescence in situ hybridization in the detection of urothelial neoplasia: An analysis of discordant results.

Background We examine the performance of cytology and FISH in the detection of urothelial carcinoma (UC), and explore the reasons for discrepant results, and potential clinical implications. Methods Urine samples from 89 patients were prospectively collected for simultaneous cytology and UroVysion FISH, and results correlated with concurrent biopsies and/or clinical or histologic follow-up data. Corresponding tissue biopsies, where available, were also evaluated by FISH. Results Sensitivity and specificity of cytology and FISH for the detection of UC was 54.8% and 92% and 50% and 88%, respectively. Only one of seven false-positive urinary FISH results proved to be an "anticipatory positive" on extended follow-up. Five of eight (62.5%) high grade (HG) carcinomas with false-negative urinary FISH, were negative due to the absence/paucity of FISH-detectable changes in the tumor cells. In atypical cytology cases, the FISH result did not assist in identifying UC. There was no significant difference between an atypical cytology result and a positive FISH result, with respect to the identification of patients with UC. Conclusions We found urinary cytology to be more sensitivity and specific than FISH in the detection of UC, though the difference was not statistically significant. Up to 24% of HG UCs are FISH negative due to an absence of FISH-detectable abnormalities in the tumor cells. Paucity of neoplastic cells in the urine also contributes to false-negative FISH results in both HG and low grade tumors. Negative urinary FISH cannot be taken alone as indicating the absence of significant disease in patients with atypical cytology.