[447] Inclusion of the Uniform Tetraploid Cells Reduces the Specificity of the Urine FISH Assay

NA Moatamed, SK Apple, F Moatamed. David Geffen School of Medicine at UCLA, Los Angeles; VA Greater LA Healthcare System, Los Angeles, CA

Background: There are approximately 50,000 new cases and 10,000 deaths from urinary bladder urothelial neoplasia each year in the United States. Cytology alone has high specificity but low sensitivity for bladder cancer detection. To increase the sensitivity, a FISH assay has been utilized to detect chromosomal abnormalities. Existing literature shows a wide variability in sensitivity and specificity of the FISH assay as performed in different centers. In this study we have identified a feature that deteriorates the specificity of the test.
Design: Among 215 cases who had urine FISH (UroVysion) test at the VAMC, 45 had associated histopathology. According to the FDA criteria, for a positive diagnosis of urothelial neoplasia, a minimum of 4 cells with hyperploidy of at least 2 of the 4 chromosomes or a minimum of 12 cells with homozygous loss of P16 gene are required when a minimum of 25 cells are examined. The four probes are for CEP (chromosomes 3, 7, 17) and P16 on chromosome 9. In this study, a cell with four signals for all four chromosomes was classified as a uniform tetraploid cell (UTC); a presumed reparative cell. Positive and negative diagnoses by the FISH were made before and after subtracting the UTCs.
Results: Before subtracting the UTCs, 22 cases were positive and 13 negative with 9 false positive and 1 false negative findings as compared with the histopathologic diagnoses. In this category, the sensitivity and specificity were 95.7% and 59.1% respectively. After subtraction of the UTCs, 21 cases became positive and 20 negative with a significant reduction in the false positive results. By excluding the UTCs, the assay performance significantly improved for all the four parameters.

Summary of the urine FISH assay before (+) and after (-) exclusion of the UTCs

	True Positive	True negative	False Positive	False Negative	Sensitivity	Specificity	PPV - NPV
+Tetraploid	22	13	9	1	95.7%	59.1%	71.0%-92.9%
-Tetraploid	21	20	2	2	91.3%	90.9%	91.3%-90.9%

Conclusions: Although this FISH assay has offered a much higher sensitivity than routine cytopathological examination, it has suffered from its variably lower specificity as used in different centers and laboratories. If, however, the UTCs are excluded as positive cells from the FDA criteria, this FISH assay would offer a more solid performance than its usage with the current criteria.
Category: Cytopathology

Tuesday, March 23, 2010 11:45 AM

Platform Session: Section F, Tuesday Morning