RESUMO
BACKGROUND: Screening of high-risk patients using bladder tumor markers can offer an advantage of early detection and saving medical costs. For these purpose many tumor markers have been developed to supplement invasive cystoscopy. Our study evaluated the NMP22 point-of-care test (NMP22 POCT), which is one of the tumor makers, comparing with the standard urine cytology for the diagnosis of bladder cancer. METHODS: From January to September 2005, 232 patients who had undergone a cystoscopy due to bladder cancer associated symptoms including hematuria and dysuria were enrolled in this study. Urine specimens were collected for NMP22 POCT and cytology. NMP22 POCT and urine cytology were compared for sensitivity and specificity. In addition, we evaluated urine stick test and microscopy to explain some false-positive results in NMP22 POCT. RESULTS: Superficial transitional cell carcinoma was diagnosed in 10 patients. The sensitivity of NMP22 test was 60% (95% confidence interval [CI], 26.2-87.8%), whereas that of cytology was 33.3% (95% CI, 7.5-70.1%); however, the difference was not significant. The specificity of NMP22 test was 69.8% (95% CI, 63.3-75.8%), compared with 99.0% (95% CI, 96.5-99.9%) for cytology (P<0.001). The presence of microscopic RBCs in urine specimen was significantly associated with the lower specificity of NMP22 POCT (P=0.02). CONCLUSIONS: NMP22 POCT was significantly less specific than urine cytology. To be useful as a bladder cancer screening test, the NMP22 test should have a higher specificity.