Bacterial and salivary biomarkers predict the gingival inflammatory profile.

Background: The aim of this human investigation is to explore the relationship of gingivitis with salivary biomarkers, periodontal pathogens, and interleukin (IL)-1 polymorphism after a transient inflammatory burden. Methods: Thirty healthy human participants were randomized by IL-1 genotype status to control for potential influences of this particular single nucleotide polymorphism on the inflammatory profile. Oral hygiene practices ceased for 21 days to induce gingivitis (induction), after which home care was reinstated until 35 days (resolution). Clinical parameters included plaque (PI) and gingival (GI) indices and papillary bleeding score (PBS). Levels and proportions of 40 subgingival bacteria were determined using checkerboard DNA-DNA hybridization. Saliva was analyzed using a multiplex protein array for 30 biomarkers associated with host defense, inflammation, tissue destruction, and angiogenesis. Results: Mean PI, GI, and PBS values were significantly increased during induction and decreased during resolution as measured at 35 days (P<0.01), although no differences were observed between IL-1 groups. Participants were stratified as either "high" or "low" responders based on inflammatory response (high: GI >1.5; low: GI <= 1.5). Baseline levels of salivary IL-6 and IL-8 demonstrated the highest ability to discriminate between high and low responders (area under the curve [AUC] of 0.81 and 0.72, respectively). Salivary biomarkers, matrix metalloproteinases (MMPs), and bacterial biofilm were combined to generate receiver operating characteristic curves. High levels of IL-6 and MMP-1 at baseline demonstrated the strongest ability to predict high responders (AUC of 0.89; odds ratio of 17.0; 95% confidence interval, 1.7 to 171.7). Conclusion: In this proof-of-concept investigation, we identified specific biomarker and microbial signatures that are associated with gingival inflammation (ClinicalTrials.gov number NCT00980525). J Periodontol 2012;83:79-89.