Abstract
BACKGROUND: An important reason for the high health care costs associated with bladder cancer is the need for frequent cystoscopy for detection and surveillance of this disease. Cytologic analysis of voided urine specimens can assist, but is too inaccurate to replace cystoscopy. In an effort to create reliable, objective, noninvasive mechanisms for detecting bladder cancer, a number of urine-based molecular tests have been developed with the ultimate goal of reducing the frequency of cystoscopy. OBJECTIVE: To summarize the performance of urine-based biomarker tests, currently commercially available in the US, as part of the initial workup for hematuria and for bladder cancer surveillance. METHODS: In accordance with PRISMA guidelines we performed a systematic review of the literature on the performance of NMP22, BTA, UroVysion, ImmunoCyt/uCyt, CxBladder, and Bladder EpiCheck. Median sensitivity, specificity, negative (NPV) and positive predictive values (PPV) were calculated for each test based on the included studies. RESULTS: Twenty-eight studies met inclusion criteria for the performance of five urine-based biomarker tests in the setting hematuria workup. Median sensitivity ranged from 65.7% -100% and specificity ranged from 62.5% -93.8%. Median NPV ranged from 94.2% -98.3% and PPV ranged from 29% -58.7%. Fourteen studies met inclusion criteria for the performance of six tests in the setting of bladder cancer surveillance. Median sensitivity ranged from 22.6% -92.0% and specificity from 20.5% -97.9%. Median NPV ranged from 52.9% -96.5% and PPV ranged from 48.1% -75.7%. CONCLUSIONS: Our analysis finds that while these tests may provide some clinical utility, none of the assays have thus far demonstrated objective evidence to supplant the gold diagnostic standard.