Urine cytology plays an important role in monitoring patients with a history of urothelial carcinoma. Because it is difficult to reliably discriminate artifacts induced by instrumentation, inflammation, or therapy those of from malignant cells, many of these specimens are categorized as atypical. The objective of the current study was to study the prevalence and significance of atypical urine cytology with regard to the effect of instrumentation and prior biopsy.
All urine cytology cases seen during a 4-year period (2001-2004) with a diagnosis of atypical urothelial cells (AU) were obtained from the cytopathology computer database. In all cases with available surgical follow-up, the following data were extracted: total number and type of urine specimen, the primary histologic diagnosis, and follow-up histologic diagnosis.
In all, 1653 voided and 3502 instrumented urine specimens were examined. A diagnosis of AU was rendered in 115 (6.9%) of the voided urine specimens and in 277 (7.9%) of the instrumented specimens. Follow-up histology was available in 70 cases, including 55 instrumented and 15 voided urine specimens. A nonbenign follow-up diagnosis was observed in 18 of 55 (32.7%) cases in the instrumented group and in 7 of 15 (46.6%) cases in the voided group. Voided urine was marginally associated with a worse subsequent biopsy diagnosis (Pexact Monte Carlo = .09)
An AU diagnosis is more predictive of a subsequent adverse biopsy diagnosis in voided urine specimens compared with instrumented urines. In the absence of a benchmark for the atypia rate, it is prudent to keep the atypia rate low to keep it more meaningful. This important category should be used by the pathologist to convey concern and recognize the difficulty in interpretation of specimens that may require close follow-up.