Classification of cytological samples from oral potentially malignant lesions through Raman spectroscopy: A pilot study.

ABSTRACT

The potential of Raman microspectroscopy of exfoliated cells has been demonstrated for oral cancer diagnosis. In this study, brush biopsies were collected from the buccal mucosa/tongue of healthy donors (n = 31) and from oral mucosal dysplastic lesions (n = 31 patients). Raman spectra were acquired and subjected to partial least squares-discriminant analysis (PLS-DA). The patient samples could be differentiated from healthy donor samples with 96% sensitivity and 95% specificity. Furthermore, PLS-DA models were developed based on cytopathological and histopathological assessment. Low and high grade dysplasia could be discriminated with 64% sensitivity and 65% specificity based on cytopathological assessment, while 81% sensitivity and 86% specificity could be achieved when histopathological assessment was within six months of the brush biopsy sampling. Therefore, this explorative study has successfully demonstrated that Raman spectroscopy may have a role in monitoring patients with dysplasia and may reduce the need for multiple biopsies.