A four-miRNA signature in serum as a biomarker for bladder cancer diagnosis

Abstract

Background: Urinary bladder cancer (BCa) is globally the 10th most frequent cancer. As a novel diagnostic tool, miRNA in serum screening is non-invasive. This project aimed to determine particular serum miRNAs as novel biomarkers for diagnosing urinary BCa.

Methods: We designed a three-phase study with 122 healthy controls (HCs) and 132 BCa patients. The 30 miRNAs' expressions in serum from HCs and BCa patients were detected during the screening phase. The miRNAs with the most dysregulation were tested in the training (HCs vs. BCa, 30 each) and validation (80 HCs vs. 82 BCa) phase further. The diagnostic ability of these candidate miRNAs was estimated by the receiver operating characteristic (ROC) curves as well as the area under the ROC curve (AUC). The miRNAs' target genes and their annotations to functions were predicted utilizing bioinformatic assays.

Results: Six serum miRNAs (miR-124-3p, miR-182-5p, miR-1-3p, miR-196a-5p, miR-23b-3p and miR-34a-5p) had significantly different expression between BCa patients and HCs in the training and validation phase. The four-microRNA panel improved the diagnostic value, with AUC =0.985. The result of bioinformatic analysis showed that these miRNAs' target genes in the panel may be related to the MAPK signaling pathway in bladder cancer.

Conclusions: Our study identified a four-miRNA panel that is a non-invasive new biomarker for diagnosing BCa.

Keywords: Bladder cancer; bioinformatics; biomarker; diagnosis; miRNA.

Figure 1

Study design overview. Study design overview. HCs: healthy controls; BCa: bladder cancer.

Figure 2

Heatmap of the 30 miRNAs during the screening phase. During the screening phase, 30 miRNAs’ expressions are shown in the heatmap. 12 candidate miRNAs were chosen for further study with 1.5 FC and P<0.05. The red means up-regulated and the blue means down-regulated.

Figure 3

The 12 candidate miRNAs’ expression during the training phase. The 12 candidate miRNAs’ expression during the training phase with 30 BCa patients and 30 HCs. *represents P<0.05, **represents P<0.01, ***represents P<0.001.

Figure 4

The expression and ROC curve analyses of six candidate miRNAs in the validation phase. In this phase, 82 BCa patients’ and 80 HCs’ serum were involved. The relative expression level of (A) miR-1-3p, (C) miR-23b-3p, (E) miR-34a-5p and (G) miR-124-3p were remarkably downregulated in BCa patients’ serum. (I) miR-182-5p and (K) miR-196a-5p were significantly upregulated. The AUC were (B) 0.682 for miR-1-3p with 95% CI: 0.604 to 0.753, (D) 0.748 for miR-23b-3p with 95% CI: 0.674 to 0.813, (F) 0.804 for miR-34a-5p with 95% CI: 0.734 to 0.862, (H) 0.688 for miR-124-3p with 95% CI: 0.611 to 0.758, (J) 0.867 for miR-182-5p 95% CI: 0.805 to 0.915 and (L) 0.731 for miR-196a-5p with 95% CI: 0.656 to 0.757. *represents P<0.05, **represents P<0.01, ***represents P<0.001.

Figure 5

Four-miRNA panel’s ROC curve analyses. The four-miRNA panel’s AUC (miR-182-5p, miR-196a-5p, miR-124-3p and miR-34a-5p) was 0.985 (95% CI: 0.952 to 0.998; sensitivity =98.78%, specificity =93.75%).

Figure 6

KEGG pathway enrichment analysis and GO annotation of the target genes of miR-182-5p, miR-196a-5p, miR-124-3p and miR-34a-5p. A. Biological process analysis; B. Cellular component analysis; C. Molecular function analysis; D. KEGG analysis.