Salivary metabolite-based prediction of radiotherapy-induced oral mucositis and candidiasis

Objectives: Oral toxicities affect a high proportion of head and neck cancer (HNC) patients receiving radiotherapy, lacking reliable predictive biomarkers for incidence and severity. We investigated salivary metabolites and metabolic pathways as non-invasive predictors of radiotherapy-induced oral mucositis (OM) and candidiasis. Material and methods: HNC patients undergoing radiotherapy were prospectively studied. Saliva samples of 45 HNC patients and 30 controls were analysed using NMR spectroscopy, and patients were assessed for oral complications during treatment. Results: Significant metabolite differences were identified between patients with and without complications. An increased concentration of lactate (p < 0.000), and a reduced concentration of formate (p = 0.018) and pyruvate (p = 0.045), was found in HNC patients at risk for OM during RT. Comparison of salivary metabolites between patients at risk or not risk for candidiasis presented increases in acetate (p = 0.005), alanine (p = 0.009), proline (p = 0.002), tyrosine (p < 0.000), trimethylamine (p = 0.007) and fucose (p < 0.000), whereas the level of citrate (p = 0.008) was decreased. Of the metabolic pathways, glucose-alanine cycle and thyroid hormone synthesis, were significantly different in patients at risk for candidiasis. Conclusions: Salivary metabolomics is a promising predictive tool for radiotherapy-induced oral complications, potentially enabling early intervention and personalized supportive care. Clinical relevance: Elevated level of salivary lactate seems to predict the development of OM, while fucosis was a sign of candidiasis. Multicentre validation studies are needed for its clinical implementation.