Comparative predictive values of urinary cytology, urinary bladder cancer antigen, CYFRA 21-1 and NMP22 for evaluating symptomatic patients at risk for bladder cancer.

PURPOSE

We study the potential diagnostic use of urinary bladder cancer antigen, CYFRA 21-1 and NMP22*; for evaluating symptomatic patients who present with microscopic hematuria and are at risk for bladder cancer.

MATERIALS AND METHODS

Urinary tumor markers were determined in 187 samples from 112 patients symptomatic of bladder cancer (group 1), and 75 with benign and other urological conditions (group 2). Immunoassays were used to measure the 3 selected biomarkers. Sensitivity and specificity were established by previously defined cut points. Biomarker results were reported as corrected and uncorrected for urinary creatinine. Urinalysis was performed in all samples.

RESULTS

Positive and negative predictive values were 85.5%, 80.5% and 81.1%, and 80.8%, 79.2% and 76.5% for urinary bladder cancer antigen, CYFRA 21-1 and NMP22, with the cutoffs 9.7 microg./l., 5.4 microg./l and 10.0 units per ml., respectively. These predictives values were 85.2% and 72.5%, respectively, for urinary cytology. The combination of biomarkers decreased the positive predictive values to 72.3% to 78.6% and increased negative predictive values to 84.2% to 86.1%. Urinary tract infection, inflammation and malignancy associated with other genitourinary organs were the primary cause for false-positive test results in the 3 assays evaluated.

CONCLUSIONS

With a single biomarker, around 80% of the positive results would have correctly identified symptomatic patients for cystoscopy. Of the negative results 75% would have correctly reduced the number of cystoscopies. Sensitivity and negative predictive values could be improved with the combination of biomarkers but with a loss of specificity and positive predictive values. Urinary tract inflammation and other genitourinary malignancies might contribute to the reduction in specificity of these tests.