A study of the reproducibility of a fluorescence in situ hybridization bladder cancer detection assay

OBJECTIVE: To assess the reproducibility of the UroVysion fluorescence in situ hybridization (FISH) bladder cancer detection assay. STUDY DESIGN: Thirteen specimens (2 negative, 3 low-level positive [1-10% abnormal cells], 5 mid-level positive [11-75%], and 3 high-level positive [>75%]) were analyzed by 7 cytotechnologists. Each cytotechnologist rendered an overall diagnosis of positive or negative and determined the percentage of abnormal urothelial cells for all positive specimens. RESULTS: The interobserver reproducibility of the assay was 100% for mid-level and high-level positive specimens, 93% for negative specimens, and 78% for low-level positive specimens. The range of percent abnormal determinations was highest for mid-level positive specimens, with mean SDs of 1.8%, 16.4% and 10.1% for the low-, mid-, and high-level positives, respectively. CONCLUSION: There was a high level of reproducibility among the mid- and high-level positive specimens. The reproducibility for low-level positive specimens was lowest, suggesting that such specimens should be reviewed by a second technologist to ensure an accurate diagnosis. The findings of this study are important for further elucidating the clinical value of quantitative FISH analysis in the management of patients undergoing FISH testing for bladder cancer.