A pilot study for early detection of oral premalignant diseases using oral cytology and Raman micro‐spectroscopy: Assessment of confounding factors. Issue 11 (23rd August 2020)

• Record Type: Journal Article
• Authors: Behl, Isha; Calado, Genecy; Malkin, Alison; Flint, Stephen; Galvin, Sheila; Healy, Claire M.; Pimentel, Marina Leite; Byrne, Hugh J.; Lyng, Fiona M.
• Is Part Of: Journal of biophotonics. Volume 13:Issue 11(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-08-23
• Subjects: oral brush biopsy cytological samples -- oral potentially malignant lesions -- partial least squares‐discriminant analysis -- Raman micro‐spectroscopy -- sensitivity and specificity; Photonics -- Periodicals; Optical materials -- Periodicals; Optics -- Periodicals; Medical instruments and apparatus -- Periodicals; 621.3605
• DOI: 10.1002/jbio.202000079
• Languages: English
• ISSNs: 1864-063X

Abstract

Abstract: This study demonstrates the efficacy of Raman micro‐spectroscopy of oral cytological samples for differentiating dysplastic, potentially malignant lesions from those of normal, healthy donors. Cells were collected using brush biopsy from healthy donors (n = 20) and patients attending a Dysplasia Clinic (n = 20). Donors were sampled at four different sites (buccal mucosa, tongue, alveolus, gingiva), to ensure matched normal sites for all lesions, while patient samples were taken from clinically evident, histologically verified dysplastic lesions. Spectra were acquired from the nucleus and cytoplasm of individual cells of all samples and subjected to partial least squares‐discriminant analysis. Discriminative sensitivities of 94% and 86% and specificity of 85% were achieved for the cytoplasm and nucleus, respectively, largely based on lipidic contributions of dysplastic cells. Alveolar/gingival samples were differentiated from tongue/buccal samples, indicating that anatomical site is potentially a confounding factor, while age, gender, smoking and alcohol consumption were confirmed not to be. Abstract : A common drawback of brush biopsy samples for screening oral potentially malignant lesions is that they may not contain morphologically abnormal cells, potentially leading to false‐negative results. In this study, it has been demonstrated using Raman micro‐spectroscopy that morphologically normal cells are biochemically abnormal. Furthermore, although choice of oral site is a potential confounding factor while developing classification models, age, gender, smoking or alcohol consumption are not.