To investigate whether immunocytochemical detection of cytokeratin (CK)-20 could serve as a reliable diagnostic marker for transitional cell carcinoma (TCC) of the bladder.

A total of 232 patients were enrolled in the study. Group 1 consisted of 144 patients with histologically confirmed TCC (62 at diagnosis and 82 in follow-up), and group 2 consisted of 88 subjects, including healthy volunteers and individuals with “non-TCC” conditions. Spontaneously voided urine specimens were obtained from each patient and submitted to immunocytologic and standard cytologic examination.

CK-20 immunocytology yielded an overall sensitivity of 65.3%, significantly greater than the sensitivity of urine cytology (54.2%, P = 0.013). A more detailed analysis revealed a sensitivity advantage for the former technique in the detection of primary (61.3% versus 51.6%, P = 0.046), recurrent (68.3% versus 56.1%, P = 0.027), Stage pT1 (81.8% versus 59.1%, P = 0.006), grade 2 (76.2% versus 61.9%, P = 0.031), and grade 3 (82.1% versus 67.9%, P = 0.039) tumors. Moreover, CK-20 immunocytochemistry demonstrated greater overall specificity than cytology (90.9% versus 86.4%, respectively), a difference stemming from the subgroup of lithiasis patients (100% versus 66.7%, P = 0.024). In terms of reliability, the positive and negative predictive values of the immunoassay were greater than those calculated for cytology (92.2% versus 86.7% and 61.5% versus 53.5%, respectively).

CK-20 immunocytology is more sensitive than standard cytology in the detection of TCC, particularly of Stage pT1, grade 2, and grade 3 tumors. In view of its high overall specificity and predictive accuracy, it is conceivable that the proposed immunoassay may progressively replace conventional cytologic screening in the diagnosis of bladder cancer.