Urine cytology: does the number of atypical urothelial cells matter? A qualitative and quantitative study of 112 cases

Abstract

This study presents a detailed and systematic morphological and quantitative analysis of urine cytology specimens in order to determine which qualitative and quantitative features are mostly associated with high-grade urothelial carcinoma (HGUCA). This study included 112 urine cytology cases with a surgical follow-up within 1 year that were originally reported as "atypical," "suspicious for HGUCA," or "positive for HGUCA." The morphological characteristics as well as the number of abnormal cells were correlated with a diagnosis of HGUCA on follow-up biopsy. Multivariate analysis showed that the presence of hyperchromatic atypical cells with nuclear-cytoplasmic ratio ≥ 0.7 was an independent predictor of HGUCA. Similarly, irregular nuclear membranes, single cells, and pleomorphism correlated with surgical outcome whereas eccentric nuclear location, prominent nucleoli, nuclear-cytoplasmic ratio between 0.5 and 0.7 did not. Cases with ≤10 atypical cells had significantly lower rates of subsequent HGUCA than did those with >10 atypical cells (58% versus 77%). Cases with ≤5 atypical cells (n = 26) showed similar prediction rates (58%) for HGUCA than did those with 6 to 10 atypical cells (n = 12). The number of atypical urothelial cells is an important criterion that should be taken into account when assigning cases to the "positive" or the "suspicious" categories. A preliminary cutoff of 10 cells appears to be easily applicable and valid from the clinical standpoint.