Raman spectroscopy can discriminate between normal, dysplastic and cancerous oral mucosa: a tissue-engineering approach.

Head and neck cancer (HNC) is the sixth most common malignancy worldwide. Squamous cell carcinoma, the primary cause of HNC, evolves from normal epithelium through dysplasia before invading the connective tissue to form a carcinoma. However, less than 18% of suspicious oral lesions progress to cancer, with diagnosis currently relying on histopathological evaluation, which is invasive and time consuming. A non-invasive, real-time, point-of-care method could overcome these problems and facilitate regular screening. Raman spectroscopy is a non-invasive optical technique with the ability to extract molecular level information to help determine the functional groups present in a tissue and the molecular conformations of tissue constituents. In the present study, Raman spectroscopy was assessed for its ability to discriminate between normal, dysplastic and HNC. Tissue engineered models of normal, dysplastic and HNC were constructed using normal oral keratinocytes, dysplastic and HNC cell lines, and their biochemical content predicted by interpretation of spectral characteristics. Spectral differences were evident in both the fingerprint (600/cm to 1800/cm) and high wave-number compartments (2800/cm to 3400/cm). Visible differences were seen in peaks relating to lipid content (2881/cm), protein structure (amide I, amide III), several amino acids and nucleic acids (600/cm to 1003/cm). Multivariate data analysis algorithms successfully identified subtypes of dysplasia and cancer, suggesting that Raman spectroscopy not only has the potential to differentiate between normal, pre-malignant and cancerous tissue models but could also be sensitive enough to detect subtypes of dysplasia or cancer on the basis of their subcellular differences.