Abstract:Abstract OBJECTIVE Bladder carcinoma (BC) is a malignant tumor that is formed in the bladder of the genitourinary system. The diagnosis at an early stage is directly associated with the improved overall survival of BC patients because a later stage usually means a poorer prognosis. Current methods to diagnose BC have various limitations, thus urologists call for novel effective non-invasive diagnostic markers. Herein, we identified miRNAs which can be used for the diagnosis of BC. MATERIALS AND METHODS Patients with BC (n = 112) and healthy individuals (n = 112) were recruited and enrolled in this study. The quantitative reverse transcription-polymerase chain reaction (qRT-PCR) was carried out for the measurement of miRNAs expression in serum. A two-phase test was proceeded for the identification, selection, and confirmation of the miRNAs that could be used for BC diagnosis. A backward stepwise logistic regression (BSLR) was conducted to establish a model containing these miRNAs with superior diagnostic performance. In addition, bioinformatics and survival analysis was exerted by analyzing database in Mirwalk, Enrichr, and OncoLnc. RESULTS Five significantly aberrant miRNAs with good diagnostic value were validated, namely miR-129-2-3p, miR-29c-3p, miR-149-3p, miR-138-5p, and miR-194-5p. Then three of them (miR-129-2-3p, miR-29c-3p, and miR-149-3p) were used to establish a diagnostic panel, in which the area under the curve (AUC) was 0.927 (95% CI: 0.876 to 0.962), providing both high sensitivity (92.68%) and specificity (80.49%). CONCLUSION In this study, a panel of three miRNAs (miR-129-2-3p, miR-29c-3p, and miR-149-3p) was developed, which could be used for the diagnosis of BC sensitively and specifically.

Combination of three microRNAs as serum diagnostic panel for bladder urothelial carcinoma