Identification of miRNAs as prognostic factors for esophageal squamous cell carcinoma

Objective: The prognostic value of microRNAs for esophageal squamous cell carcinoma (ESCC) is still not be well identified. Methods: The microRNA expression profiles of 119 paired ESCC tissue samples and para-carcinoma tissues from GEO database under accession number of GSE43732. A mutation information based feature selection method was applied to identify the discriminative microRNAs between paired ESCC tissues and para-carcinoma tissues. Results: In para-carcinoma tissues, patients had better survival with higher has-miR-410 (log-rank p = 0.0123), has-miR-411-5p (log-rank p = 0.0152), has-miR-193b-5p (log-rank p = 0.0188) and has-miR-4486 (log-rank p = 0.0307) expression levels. When compared with para-carcinoma tissues, there has more correlations between miRNA expression levels and survival in tumor tissues. We identified 20 potential miRNAs associated with prognosis. Besides, a heatmap was draw to explore miRNA expression levels in tumor tissues and survival. Conclusions: The present study identified 24 miRNAs in 119 paired ESCC tissue samples and para-carcinoma tissues, including 4 miRNAs in para-carcinoma tissues and 20 in tumor tissues, respectively. The dysregulation of these miRNAs were associated with different outcomes. We thought this study could provide novel noninvasive early biomarkers for ESCC patients.