The 5-year survival rate of patients with head and neck squamous cell carcinoma (HNSCC) was only 40%-50%. To investigate the prognostic and predictive value of specific mircoRNAs (miRNAs) in HNSCC. We identified 19 miRNAs associated with over survival (OS) of patients with HNSCC in different clinical classes between 492 HNSCC tissues and 44 normal tissues from The Cancer Genome Atlas (TCGA) dataset. A signature of six miRNAs was identified by the supervised principal component method in the training set. The AUC of the ROC curve for the six microRNA signature predicting 5-year survival was 0.737 (95%CI, 0.627-0.825) in the testing set and 0.708 (95%CI, 0.616-0.785) in the total dataset. In the multivariate Cox regression analysis, patients with high-risk scores had shorter OS (HR, 2.380, 95%CI, 1.361-4.303) than patients with low-risk scores in the total dataset. Therefore, these results provided a new prospect for prognostic biomarker of HNSCC.
Keywords: HNSCC; TCGA; miRNA; overall survival; prognosis.