DNA slit‐scan flow cytometry of bladder irrigation specimens and the importance of recognizing urothelial cells

Open Access
Publisher Website
Google Scholar
Abstract
DNA slit-scan flow cytometry was used to analyze 150 bladder irrigation specimens from 83 patients. Specimens were categorized into groups based on cystoscopy, histology, and cytopathology. Cells were stained for DNA with propidium iodide using a whole cell protocol. Nonspecific fluorescence in the cytoplasm of some urothelial cells together with differential DNA staining of cell types in certain specimens was noted. DNA frequency distributions were analyzed using a semi-automated technique. Data were gated using slit-scan morphological features to remove cellular debris, multiple nuclei, and cells exhibiting nonspecific cytoplasmic fluorescence. Specimens were classified abnormal if they were aneuploid or had a hyperdiploid fraction (HDF) greater than 8%. The sensitivity to abnormality was 89% for grade 3 transitional cell carcinoma (TCC), 70% for grade 2 TCC, and 67% for grade 1 TCC. Specificity was 61%. Specimen data were then reprocessed using slit-scan morphological features to enrich for urothelial cells. The urothelial cells were identified by the ratio of nuclear diameter to cell diameter. This method was found to be in good agreement with immunofluorescent labeling of urothelial cells using the urothelium-seleetive T16 monoclonal antibody. The sensitivity to abnormality remained 89% for grade 3 TCC and 70% for grade 2 TCC, but fell to 52% for grade 1 TCC. Specificity for the urothelial cell enriched data increased to 77%. Reprocessing of data to enrich for urothelial elements resulted in 16 fewer specimens with an aneuploid DNA distribution and 2 fewer specimens with increased HDF. These data illustrate the necessity for recognition of urothelial cells in the flow analysis of unfixed bladder irrigation specimens stained with propidium iodide using a whole cell technique. Differential staining of blood and urothelial cells can mimic aneuploidy or contribute to an increased hyperdiploid fraction. These may result in the specimen being classified as abnormal if they are not recognized as such.