Background and aim:The gold standard diagnosis of oral pre-cancer/cancer is an invasive biopsy followed by a histological examination, which may present psychological trauma and risk of infection to patients. Moreover, multiple biopsies are often required to monitor precancer lesions' progression. With the aim of reducing the need for multiple biopsies, this study was undertaken to identify pre-cancerous lesions using minimally invasive brush biopsy and saliva samples, analysed by Raman spectroscopy.
Method:In this study, brush biopsies and saliva samples were collected from healthy donors (n = 50) and from patients with dysplastic lesions (n = 50). Raman spectra were acquired from the nucleus and cytoplasm of each cell and from liquid saliva.
Results:A discriminant analysis of brush biopsy samples indicated that dysplastic cells could be successfully differentiated from normal cells with 96% accuracy based on changes in lipid content, while a similar analysis of saliva samples achieved 86% accuracy, based on protein spectral changes. In addition, discrimination between high-grade and low-grade dysplasia was achieved with 80% accuracy using cell samples, and discrimination between dysplasia and oral squamous cell carcinoma was achieved with 75% accuracy using saliva samples.
Conclusions:This study successfully demonstrated that Raman spectroscopy and minimally invasive brush biopsy and saliva samples could be used for the early detection of precancerous lesions and could have a role in monitoring patients with dysplasia, reducing the need for multiple biopsies.