Oral squamous cell carcinoma patients can be differentiated from healthy individuals with label-free serum proteomics.

Background: No blood biomarkers to detect early oral cavity squamous cell carcinoma (OSCC) without clinical signs exist - diagnosis is solely based on histology of a visible tumour. Most OSCC patients are diagnosed at advanced stage, which leads to significant morbidity and poor survival. Our aim was to find the serum screening or detection biomarkers in OSCC.

Methods: Serum samples from patients with OSCC treated at the Department of Otorhinolaryngology - Head and Neck Surgery, Helsinki University Hospital (Finland) were collected. Age- and gender-matched healthy individuals served as controls. Quantitative label-free proteomics in high definition MS mode(HDMS) was performed on 13 patients and 12 healthy samples. Various statistical analyses were performed on quantitative proteomics data to obtain the most influential proteins, which classify the patients vs healthy samples.

Results: In quantitative proteomic analysis (HDMS), 388 proteins were quantified in our pilot study. A complete separation between cases and controls was seen in supervised and unsupervised classification techniques such as orthogonal projections on latent structure-discriminant analysis (OPLS-DA) and self-organising maps. Using OPLS-DA S-plot, we identified a set of eight proteins that completely separated OSCC patients from healthy individuals.

Conclusions: Although the tumour stages varied from I to IVa, these potential biomarkers were able to identify all OSCCs demonstrating their sensitivity to detect tumours of all stages. We are the first to suggest a set of serum biomarkers in our pilot study to be evaluated further as a diagnostic panel to detect preclinical OSCC in risk patients.