ABSTRACT
UroVysion FISH detects chromosomal aberrations associated with urothelial carcinoma. In our laboratory, UroVysion FISH was initially evaluated manually with a change to image-aided interpretation using the BioView Duet imaging system. This retrospective study examined diagnostic findings over an 8.6 year period, with 1,869 manual interpretations over 4.8 years and 3,936 image-aided interpretations over 3.8 years. Although the initial goal was to evaluate possible impacts of the imaging system on diagnostic interpretations, the most important finding was that the demographics of the test population changed significantly. Female specimens increased incrementally from an average of 29% compared to 43% of the samples during periods of manual interpretation versus image-aided interpretation, respectively. The shift may reflect a gradual increase in the percentage of low-risk hematuria patients being evaluated for initial diagnosis of bladder cancer, rather than bladder cancer recurrence. Interpretation rates, evaluated separately for males and females, changed significantly over the test period. Male interpretation results were negative (75.1 vs. 67%), positive (18.6 vs. 14.6%), unsatisfactory (5.0 vs. 16.9%), and equivocal (1.4 vs. 1.5%) during periods of manual versus image-aided interpretation, respectively (Fisher Exact Test P-value = <0.0001). For females, results were negative (86.1 vs. 79.3%), positive (9.2 vs. 11.1%), unsatisfactory (2.8 vs. 8.9%), and equivocal (1.8 vs. 0.7%) over the same periods (Fisher Exact Test P-value = <0.0001). Logistic regression analysis identified the change in test population as the variable with the greatest impact on observed interpretation rate changes.
