Abstract: Objectives To evaluate the clinical utility of a Multiprobe FISH (multicolor fluorescence in situ hybridization, M-FISH) assay in voided urine specimens for the detection of bladder cancer, comparing the results with those afforded by urinary cytology. Methods Voided urine specimens from 100 patients who had cancer tumor or after surgery and 10 patients who began prostate hyperplasia were obtained for urinary cytology and FISH. FISH was performed using a mixture of fluorescent labeler DNA probes for the centromeric regions of chromosomes 3, 7, and 17 and 9p21 region. Cystoscopy with biopsy or tumor resection was performed in all patients, comparing the pathological results with the cytological and FISH findings. Results The sensitivities for M-FISH were 75.6% for low grade tumors, 100% for high grade tumors, with overall sensitivities of 85.5%. Urinary cytology affords an overall sensitivity of 62.9%, the figure being 33.3% for low grade tumors, 96.0%for high grade tumors. Significant difference in overall sensitivities was observed between M-FISH and Urinary cytology (P<0.05). The specificities of M-FISH and urinary cytology were 84.6% and 87.8%, respectively. No significant difference in specificities was observed between M-FISH and Urinary cytology .Conclusion M-FISH improves the sensitivity rates obtained with urinary cytology for bladder cancer detection, and affords similar specificity, and can detect all high grade infiltrating tumors patients.

Key words: Bladder, Cancer, Fluorescence in situ hybridization, Cytology