Is the performance of urinary cytology as high as reported historically? A contemporary analysis in the detection and surveillance of bladder cancer

Abstract

ObjectivesThe goal of this study was to evaluate sensitivity and specificity of urine cytology during a contemporary period at our institution in comparison with historical analysis and other reported urinary biomarkers. Materials and methodsData from 1,114 consecutive patients corresponding to 3,251 specimens (2,979 cytologic and 272 histologic specimens) between January 2006 and July 2006 were retrieved. Subsequent cytologic and surgical specimen reports were examined with a minimum 2-year follow-up period. Collected parameters included the date of collection, reason for urinary evaluation, type of specimen, and tumor grade. Atypical diagnosis was considered negative. ResultsOn cytologic examination, 71% of specimens were benign, 23% atypical, and 6% suspicious or positive for urothelial carcinoma. Reason for collection was surveillance in 61% and new symptoms in 28%. Depending on the tumor grade, sensitivity results ranged from 10% for low-grade to 51% for high-grade tumors. Importantly, specificity of urine cytology ranged from 83% to 88% (depending on the type of urine collection and type of clinical presentation). Anticipatory positive rate was 44% after a median time of 15 months. Specificity of other reported urinary markers ranges from 40% to 90%. ConclusionOur institution's experience with regard to specificity of urine cytology is lower than reported historically. Whether this is a consequence of heterogeneous study designs and parameters is open to debate. As the anticipatory positive rate was high, close surveillance remains recommended in patients with positive urine cytology and negative workup. Other institutions are encouraged to evaluate whether there remains a significant advantage for urine cytology over other urinary marker assays within their own clinical setting.