Urine Biomarkers for the Diagnosis of Bladder Cancer: a Network Meta-Analysis.

Abstract

To identify effective urine biomarkers for bladder cancer diagnosis. This meta-analysis was conducted following the guidelines of the Meta-Analyses (PRISMA) statement. Relevant studies were searched from the PubMed, Embase, and Cochrane Library databases. Heterogeneity tests were performed using Q statistics and I2 tests to determine the use of the random or fixed effects model. A direct comparison meta-analysis and network meta-analysis were conducted. The effect values are presented as odds ratios and 95% confidence intervals. Sensitivity analysis and consistency tests were performed. Fifty-eight studies with 12,038 participants were included. Direct comparison meta-analysis showed statistically significant differences in bladder cancer antigen (BTA) trak vs. nuclear matrix protein 22 (NMP22), BTA stat vs. urine cytology (UC), and fluorescence in situ hybridization (FISH) vs. UC, among the sensitivity indicators. Among the specificity indicators, there were statistically significant differences in BTA trak vs. UC, ImmunoCyt (immunocyte) vs. NMP22, and BTA stat vs. FISH. Among the positive predictive indicators, NMP22 vs. UC, BTA stat vs. UC, and FISH vs. NMP22 showed statistically significant differences. Among the negative predictive indicators, the differences in FISH vs. UC, FISH vs. NMP22, and hyaluronidase 1 (HYAL-1) vs. UC were statistically significant. Among the accuracy indicators, FISH vs. NMP22, FISH vs. UC, and HYAL-1 vs. UC showed statistically significant differences. Network meta-analysis showed that HYAL-1, urothelial carcinoma associated 1 (UCA1) and survivin had the highest sensitivity, while UC had the lowest sensitivity. The specificity of UC, FISH, and HYAL-1 was the highest, while that of UCA1 was the lowest. In terms of positive predictive indicators, UC, FISH, and HYAL-1 had the highest positive predictive value, while the BTA group had the lowest positive predictive value. In terms of negative predictive indicators, HYAL-1, UCA1, and survivin had the highest negative predictive value, while UC had the lowest negative predictive value. In terms of accuracy indicators, HYAL-1, UCA1, and survivin had the highest accuracy, while UC had the lowest accuracy. HYAL-1 and survivin are suitable urine biomarkers for bladder cancer diagnosis.