Diagnostic value of voided urine and bladder barbotage cytology in detecting transitional cell carcinoma of the urinary tract

In this article we report on our experience with the use of urine cytology for the screening and diagnosis of transitional cell carcinoma (TCC) of the bladder and upper urinary tracts at our institution between January 1987 a nd December 1995. A total of 76 patients were included in the study. All pa tients had voided urine cytology studies read as positive or highly suspici ous for malignancy and no prior history of TCC of the urinary tract. All th ese patients subsequently underwent cystoscopy, bladder/ureteral barbotage cytology, random bladder biopsies, and radiographic studies of the upper tr acts. Of the 76 patients with positive urine cytology, 53 also had barbotag e urine cytolcogies which were positive. Six of these patients were found t o have cystoscopically evident TCC of the bladder, and 1 patient had upper tract TCC. Three other patients subsequently went on to develop TCC of the bladder at 52, 89 and 111 months of follow-up. An additional patient was di agnosed with upper tract TCC at 12 months of follow-up. Among the 23 patien ts with negative bladder/ureteral barbotage cytology, 3 patients, 2 at the time of initial cystoscopy, and one 15 months later, showed evidence of TCC . Median patient follow-up was 97 (range 35-132) months. Thus of 76 patient s with initial positive voided urine cytology studies, only 9 proved to hav e TCC at initial work-up, while 5 other patients were diagnosed with TCC du ring a median follow-up of 97 months. The statistical diagnostic values of the. bladder/ureteral barbotage urine cytology studies at the time of cysto scopic work-up were: sensitivity 77%; specificity 31%; positive predictive value 13%, and negative predictive value 91%. Our data suggest that in pati ents without a previous history of TCC, the diagnostic value of bladder bar botage urine cytology is insignificant, and therefore not cost effective to be included as part of the routine work-up of TCC. Moreover, in patients w ith initially positive voided urine cytology and negative work-up, if the c ytology subsequently becomes negative, the likelihood of the development of TCC is low. However, if the initially positive cytology continues to remai n positive, there is a much higher probability of TCC being detected in thi s population.