Comparison of multitarget fluorescence in situ hybridization in urine with other noninvasive tests for detecting bladder cancer


M.G. Friedrich, MD, Norris Comprehensive Cancer Center, University of Southern California, 1441 Eastlake Avenue, Room 7336, Los Angeles, California 90089-9176, USA.



To present a single-centre study investigating aneuploidy at chromosomes 3, 7, 17 and 9p21 (e.g. loss at 9p21) using a multitarget fluorescence in situ hybridization (FISH) system, as identifying genetic alterations in urine specimens is a promising approach for the noninvasive detection of bladder cancer.


Urine samples from 103 patients were evaluated, including those from 46 with histologically confirmed urothelial carcinoma, two with other urological malignancies, and 55 who acted as controls. The urine samples were taken before any manipulation. The validity of FISH (Urovision, Vysis, Downers Grove, Ill, USA) was compared with other noninvasive urine tests, including the BTA-Stat test, the nuclear matrix protein (NMP)-22 test, and immunocytology against 486p3/12 and LewisX. Those evaluating the tests were unaware of the clinical and histopathological data. FISH was considered positive if five or more urinary cells had gains of two or more chromosomes. The threshold for the urine tests were 10 U/mL (NMP-22), 30% positive cells (486p3/12), or 5% positive cells, respectively (LewisX).


The sensitivity was 69% (FISH), 67% (BTA-Stat), 69% (486p3/12), 96% (LewisX) and 71% (NMP22), respectively; the respective specificity was 89%, 78%, 76%, 33% and 66%.


Multitarget FISH had a better specificity than the other urine markers but because of its inadequate sensitivity it does not seem to be powerful enough to replace endoscopy. Optimizing the marker panel could provide a higher sensitivity.