Saliva Biomarkers That Differentiate Between Health or Gingivitis and Periodontitis

Objectives: Periodontitis is a prevalent chronic inflammatory disease that causes destruction of the tooth supporting tissues. Earlier detection utilising saliva biomarkers could offer opportunities for prevention and potentially enable better disease management options. The aim of this study was to discover and validate saliva biomarkers that can differentiate between health or gingivitis and periodontitis.
Methods: A total of 190 participants were recruited from two centres (Birmingham, UK and Newcastle upon Tyne, UK) comprising healthy (n=39), gingivitis (n=35), periodontitis (n=80) and edentulous donors (n=36). Participants donated stimulated saliva samples at both sites and gingival crevicular fluid (GCF) at Birmingham. Samples from the Birmingham cohort were analysed by quantitative mass spectrometry proteomics to discover potential biomarkers. The shortlisted candidates were then verified in saliva from both cohorts by enzyme linked immunosorbent assay (ELISA). Leave-one-out cross validation logistic regression analysis was used to determine area under the curve (AUC) for the receiver operator characteristics (ROC).
Results: 95 proteins were found in both saliva and GCF and 15 were selected with potential to differentiate between health or gingivitis and periodontitis. After verification by ELISA, 4 biomarkers (matrix metalloproteinase 9, S100A8, alpha-1-acid glycoprotein and pyruvate kinase) used in combination had AUC 0.960 (95% confidence intervals 0.943-0.977) compared to single biomarkers (mean±SD: 0.732 ± 0.12). Saliva from edentulous donors contained low quantities of the biomarkers and appeared most similar in biomarker profile to healthy donors.
Conclusions: Saliva proteins were discovered and validated that could differentiate between saliva donors with healthy mouths or who had gingivitis and donors who had periodontitis. The inclusion of GCF proteins in the discovery phase appears to have yielded markers specific to the site of inflammation as saliva from edentulous patients, without GCF or inflammation, contained low quantities of the suggested biomarkers.