Qualitative and quantitative assessment of urinary cytokeratin 8 and 18 fragments compared with voided urine cytology in diagnosis of bladder carcinoma.

Objectives: To evaluate the value of urine tests based on the detection of cytokeratins 8 and 18 for the diagnosis of bladder cancer compared with urine cytology.

Methods: Samples from 112 patients before transurethral resection (group 1), 40 patients before secondary surgical treatment (group 2), 29 healthy control subjects (group 3, controls), and 10 women with acute urinary tract infection (group 4, controls) were examined with the UBC Rapid and UBC II enzyme-linked immunosorbent assay (ELISA) tests and voided urine cytology.

Results: Of the 112 patients in group 1, 90 had transitional cell carcinoma. For the UBC Rapid, UBC ELISA, and cytology, the sensitivity and specificity was 64.4%, 46.6%, and 70.5% and 63.6%, 86.3%, and 79.5%, respectively. The cytology had the greatest accuracy (72.3%) compared with both cytokeratin tests (54.4% and 64.2%). For all three tests, sensitivity increased with tumor grade and stage. In group 2, 16 of 40 patients had residual carcinoma. The sensitivity was similar for all three tests, and the specificity of cytology was lower compared with its specificity in group 1 (47.9% versus 70.5% in group 1). In the controls with or without urinary tract infection, the specificity of cytology was greater than that of both other tests. The combination of the UBC ELISA test with cytology increased the sensitivity to 83% (specificity 68%).

Conclusions: Both cytokeratin tests detected patients with transitional cell carcinoma, but were inferior to voided urine cytology in test quality.