Urinary Biomarkers for Diagnosis of Bladder Cancer A Systematic Review and Meta-analysis

Background: Urinary biomarkers may be a useful alternative or adjunct to cystoscopy for diagnosis of bladder cancer. Purpose: To systematically review the evidence on the accuracy of urinary biomarkers for diagnosis of bladder cancer in adults who have signs or symptoms of the disease or are undergoing surveillance for recurrent disease. Data Sources: Ovid MEDLINE (January 1990 through June 2015), Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and reference lists. Study Selection: 57 studies that evaluated the diagnostic accuracy of quantitative or qualitative nuclear matrix protein 22 (NMP22), qualitative or quantitative bladder tumor antigen (BTA), fluorescence in situ hybridization (FISH), fluorescent immunohistochemistry (ImmunoCyt [Scimedx]), and Cxbladder (Pacific Edge Diagnostics USA) using cystoscopy and histopathology as the reference standard met inclusion criteria. Case-control studies were excluded. Data Extraction: Dual extraction and quality assessment of individual studies. Overall strength of evidence (SOE) was also assessed. Data Synthesis: Across biomarkers, sensitivities ranged from 0.57 to 0.82 and specificities ranged from 0.74 to 0.88. Positive likelihood ratios ranged from 2.52 to 5.53, and negative likelihood ratios ranged from 0.21 to 0.48 (moderate SOE for quan-titative NMP22, qualitative BTA, FISH, and ImmunoCyt; low SOE for others). For some biomarkers, sensitivity was higher for initial diagnosis of bladder cancer than for diagnosis of recurrence. Sensitivity increased with higher tumor stage or grade. Studies that directly compared the accuracy of quantitative NMP22 and qualitative BTA found no differences in diagnostic accuracy (moderate SOE); head-to-head studies of other biomarkers were limited. Urinary biomarkers plus cytologic evaluation were more sensitive than biomarkers alone but missed about 10% of bladder cancer cases. Limitation: Restricted to English-language studies; no search for studies published only as abstracts; statistical heterogeneity present in most analyses; few studies for qualitative NMP22, quantitative BTA, and Cxbladder; and methodological shortcomings in almost all studies. Conclusion: Urinary biomarkers miss a substantial proportion of patients with bladder cancer and are subject to false-positive results in others. Accuracy is poor for low-stage and low-grade tumors.