Abstract
INTRODUCTION: Identification of molecular biomarkers in the saliva and serum of oral cavity cancer patients represents a first step in the development of essential and efficient clinical tools for early detection and post-treatment monitoring. We hypothesized that molecular analyses of paired saliva and serum samples from an individual would likely yield better results than analyses of either serum or saliva alone. MATERIALS AND METHODS: We performed whole-transcriptome and small non-coding RNA sequencing analyses on 32 samples of saliva and serum collected from the same patients with oral squamous cell carcinoma (OSCC) and healthy controls (HC). RESULTS: We identified 12 novel saliva and serum miRNAs and a panel of unique miRNA and mRNA signatures, significantly differentially expressed in OSCC patients relative to HC (log2 fold change: 2.6-26.8; DE: 0.02-0.000001). We utilized a combined panel of the 10 top-deregulated miRNAs and mRNAs and evaluated their putative diagnostic potential (>87% sensitivity; 100% specificity), recommending seven of them for further validation. We also identified unique saliva and serum miRNAs associated with OSCC and smoking history (OSCC smokers vs. never-smokers or HC: log2 fold change: 22-23; DE: 0.00003-0.000000001). Functional and pathway analyses indicated interactions between the discovered OSCC-related non-invasive miRNAs and mRNAs and their targets, through PI3K/AKT/mTOR signaling. CONCLUSION: Our data support our hypothesis that using paired saliva and serum from the same individuals and deep sequencing analyses can provide unique combined mRNA and miRNA signatures associated with canonical pathways that may have a diagnostic advantage relative to saliva or serum alone and may be useful for clinical testing. We believe this data will contribute to effective preventive care by post-treatment monitoring of patients, as well as suggesting potential targets for therapeutic approaches.