A prospective comparison of UroVysion FISH and urine cytology in bladder cancer detection

BACKGROUND: UroVysion fluorescence in situ hybridization (uFISH) was reported to have superior sensitivity to urine cytology. However uFISH studies are limited by varying definitions of what is considered a positive result, absence of histopathology and small sample size. The aim of our study was to better determine the performance characteristics of uFISH and urine cytology by overcoming some of the deficiencies of the current literature. METHODS: Intraoperative bladder wash cytology and uFISH were collected prospectively on all patients. Strict definitions for positivity of uFISH and cytology were determined before initiating the study. A re-review of false-negative uFISH specimens was performed to analyze potential sources of error. Sixteen bladder tumors embedded in paraffin were analyzed by uFISH and compared with the result in the urine. RESULTS: One hundred and twenty-nine specimens were analyzed. Sensitivity was 67% and 69% (p = 0.54); specificity was 72% and 76% (p = 1.0), for uFISH and cytology, respectively. Thirty-two false negative uFISH samples were re-reviewed. Low grade tumors often showed cells with abnormal morphology and patchy DAPI staining but diploid chromosomal counts and a few high grade tumors had tetraploid counts but less than needed to interpret uFISH as positive. uFISH study of the tumors revealed three categories; positive in both tumor and urine (9), negative in both tumor and urine (5) and positive in tumor but negative in urine (2). CONCLUSION: In a pathologically-confirmed analysis of bladder washed urine specimens, uFISH does not outperform urine cytology in cancer detection.