Objectives. To evaluate the diagnostic characteristics of the urinary measu
rement of cytokeratin tissue polypeptide-specific antigen (TPS) for the det
ection of bladder cancer.
Methods. Three hundred thirty-five individuals in five groups were studied:
group 1, subjects with microhematuria under suspicion for primary bladder
cancer; group 2, patients being followed up with scheduled cystoscopic exam
inations; group 3, patients in follow-up receiving chemotherapy instillatio
ns; group 4, patients with other urologic diseases; and group 5, healthy su
bjects. Urine samples belonging to subjects from groups 1, 2, and 3 were co
llected immediately before cystoscopy. Additionally, patients from groups 2
and 3 were monitored with urinary TPS for a minimum period between two cys
toscopies, TPS was measured by an enzyme immunosorbent assay.
Results. Receiver operating characteristic analysis gave a sensitivity of 6
4% and a specificity of 84% at a threshold value of 279 U/L. The positive a
nd negative predictive value was 66% and 82%, respectively; accuracy was 77
%. TPS could discriminate the presence of bladder tumor sooner than the sch
eduled cystoscopies in 9 of 19 follow-up patients with recurrence. False-po
sitive results during follow-up were found in 112 urine samples, one third
of which were associated with urinary tract infections. TPS did not appear
to be specific for bladder cancer, with elevated results in 45% of patients
from group 4, which might lead to clinical misinterpretation of urinary TP
S results.
Conclusions. Urinary TPS might provide additional information for the detec
tion of bladder cancer as an adjunct to cystoscopy. Considering the false-p
ositive rates, different urologic diseases should be ruled out before makin
g clinical decisions on the basis of elevated urinary TPS results, UROLOGY
55: 526-532, 2000, (C) 2000, Elsevier Science Inc.