Comparison of different concepts for interpretation of chromosomal aberrations in urothelial cells detected by fluorescence in situ hybridization

Urine fluorescence in situ hybridization (FISH) has become a broadly used marker for noninvasive detection of bladder cancer (BC). However, it has been discussed whether the interpretation algorithm proposed by the manufacturer could be improved. Aim of the present study was to compare alternative evaluation strategies of FISH for detection of BC. We included 1048 patients suspicious for BC, who underwent urine FISH examination before cystoscopy (diagnostic cohort). Herefrom, we selected 122 patients (prognostic cohort) with a history of non-muscle-invasive BC who were cystoscopically tumor free and received FISH analysis ahead of a follow-up period of 24 months. FISH results were interpreted by the algorithms of UroVysion (TM), Bubendorf et al. and Zellweger et al. In the diagnostic cohort, 228 patients (21.8%) had BC at time of evaluation; in the prognostic cohort 39 patients (32.0%) experienced tumor recurrence. Alterations in chromosome 3, 7 and 17 correlated with the presence of BC. Relative loss of 9p21 was associated with BC and higher risk for progression. The evaluation strategy proposed by Zellweger et al. showed highest accuracy of all FISH assessments. Performance of evaluation strategies differed in voided urine samples and samples obtained after mechanical manipulation. The performance of FISH in BC diagnosis strongly depends on the interpretation criteria. Alternative evaluation methods partly show superior diagnostic performance compared to the manufacturer's algorithm. The introduction of specific cutoffs for tetraploid cells improves specificity. Further modifications of the interpretation algorithm of the Urovysion(A (R)) FISH assay have the potential to positively affect the value of this test in diagnosis and surveillance of BC.