Oral squamous cell carcinoma (OSCC) has high incidence rates in India, with poor survival rates due to late diagnosis. Oral potentially malignant disorders (OPMDs) such as leukoplakia (L) and oral submucous fibrosis (OSMF) present a critical window for intervention. Minimally invasive approaches capturing early biochemical alterations were investigated to improve early detection and stratification. Blood serum samples from 169 subjects, including healthy controls (C), tobacco habitués (TC), L, OSMF, and OSCC, were analyzed using Raman spectroscopy (RS) and liquid chromatography-tandem mass spectrometry (LC-MS/MS). Principal component-based quadratic discriminant analysis (PC-QDA) models were built with RS data and validated on independent cohorts. Multivariate curve resolution-alternating least-squares (MCR-ALS) analysis identified altered spectral features. Global LC-MS/MS metabolite profiles were assessed by multivariate statistics and pathway enrichment. RS-based PC-QDA models achieved ∼95% accuracy in training and testing when C subjects were compared against TC, L, OSCC, and OSMF in a two-group model. A three-group model with C, TC, and oral diseases accurately classified >80% C and 86% oral disease subjects. The model stratifying L, OSMF, and OSCC identified 100% OSCC and 73-84% of L and OSMF in independent test sets. MCR-ALS revealed spectral features of albumin, immunoglobulins, carotenoids, and lipids, corresponding to LC-MS/MS findings of altered albumin-bound metabolites, bile acids, lipid metabolism, and oxidative stress. Serum RS demonstrated efficacy as a rapid, minimally invasive detection tool for oral disease stratification. LC-MS/MS identified metabolites and pathways aligning with RS spectral signatures. This multimodal approach shows promise for early detection and risk assessment of oral cancer.
Saha et al. (Tue,) studied this question.