LAUSR

ABSTRACT The pathophysiology of primary Burning mouth syndrome (BMS) remains controversial. Targeted analyses or "omics" approach of saliva provide diagnostic or pathophysiological biomarkers. This pilot study's primary objective was to explore the pathophysiology of BMS through a comparative analysis of the salivary metabolome among 26 BMS female cases and 25 age and sex matched controls. Secondary objectives included comparative analyses of inflammatory cytokines, neuro-inflammation markers and steroid hormones among cases and controls, and among BMS patients according to their clinical characteristics. Salivary metabolome, neuro-inflammatory markers, cytokines and steroids were respectively analysed by liquid chromatography coupled to mass sperctrometry, ELISA and protease activity assay, and multiparametric Luminex method. Among the 166 detected metabolites, univariate analysis did not find any discriminant metabolite between groups. Supervised multivariate analysis divided patients into two groups with an accuracy of 60% but did not allow significant discrimination (permutation test, p=0.35). Among the metabolites contributing to the model, three belonging to the tyrosine pathway (L-dopa, L-tyrosine and tyramine) were involved in the discrimination between cases and controls, and among BMS patients according to their levels of pain. Among the detectable molecules, levels of cytokines, steroid hormones and neuro-inflammatory markers did not differ between cases and controls and were not associated with characteristics of BMS patients. These results do not support the involvement of steroid hormones, inflammatory cytokines or inflammatory neurogenic mediators in the pathophysiology of pain in BMS, whereas the observed shift in tyrosine metabolism may indicate an adaptative response to chronic pain or an impaired dopaminergic transmission.