Prognostic Biomarkers of Periodontal Disease

Abstract

Objectives: Previous studies in our laboratory have identified the antimicrobial proteins Human Neutrophil Protein 1 – 3 (HNP1-3), Myeloid Related Protein 8 (S100A8/MRP8) and LL-37 as putative periodontal salivary biomarkers. The aims of the studies reported in this thesis were to investigate the diagnostic and prognostic potential of these markers, together with Matrix Metalloproteinase- 8 (MMP-8), serum markers C - reactive protein (CRP) and Interleukin-6 (IL-6), on the initial outcome of nonsurgical periodontal treatment. Material & Methods: We carried out a cross sectional study which aimed to verify and establish a diagnostic threshold for a group of salivary biomarkers (MMP-8, HNP1-3, S100A8 and LL-37) and to test the validity of the diagnostic utility of these biomarkers. A total of 133 unstimulated saliva samples (46 with chronic periodontitis, 38 with aggressive periodontitis and 49 with gingivitis) were analysed by ELISA. In addition multiple markers were combined to give a single combined cut off point by normalising each biomarker to percentage of cut-off point value, (such that x + y = combined cut off point). These pre-determined cut-off points were applied to salivary AMPs levels in an independent cohort originally collected to investigate the effects of diabetes on periodontitis. To investigate prognostic potential a total of 66 participants were recruited to a longitudinal intervention study of patients with moderate–severe Chronic Periodontitis. 53 subjects completed the protocol and were included in the final analysis. Subjects (28 male, 25 female) age range (23-65 years) with were recruited, with 14 smokers and 3 with type II diabetes, and saliva and serum samples were collected prior to periodontal examination. Patients were then given a course of non-surgical periodontal therapy over 2 visits. 8-10 weeks post-operatively saliva/serum sampling and clinical examination were repeated. Salivary MMP-8, S100A8 and HNP1-3 concentrations were all determined by ELISAs. In addition we measured serum levels of CRP and Interleukin 6. Results: In the cross-sectional study the HNP1-3 and S100A8 could differentiate between gingivitis and chronic periodontitis with high specificity (around 90%) and around 75% sensitivity compared to MMP-8 which was able to discriminate between gingivitis and periodontitis (chronic and aggressive) with both high 3 specificity and sensitivity. LL37 showed no significant diagnostic potential. Within the independent cohort the application of pre-determined thresholds, either individual or combined cut-offs, were able to detect periodontitis with specificity of between 75 – 85 % but with very low sensitivity. In addition diabetic status was found to result in significantly increased MMP-8 and S100A8 concentrations in subjects with periodontal disease. In the intervention study, treatment resulted in reductions in the mean: a) number of deep sites (>4mm) (33.57 ± 20.75 vs 18.51 ± 13.87; mean ± SD, p<0.0001); b) probing pocket depths (5.92 ± 0.47 mm vs 4.74 ± 0.76 mm, p<0.0001); c) bleeding index (0.32 ± 0.20 vs 0.21 ± 0.16, p<0.0001); d) plaque index (0.46 ± 0.20 vs 0.37 ± 0.18, p=0.0003). Only the mean concentrations of MMP-8 and S100A8 showed significant reductions post-treatment (MMP-8: 355.4 ± 319.9 ng/ml vs 216.6 ± 217.2 ng/ml, p<0.0001), (S100A8: 1182 ± 1095 ng/ml vs 693.9 ± 719.6 ng/ml, p=0.0007). Only the change in concentrations of MMP-8 were strongly associated with magnitude of treatment response (MMP-8: r2= 0.1, p=0.02). In addition, the baseline levels of MMP-8 & S100A8 were also associated with treatment response (MMP-8: r2= 0.1, p=0.03; S100A8: r2=0.1, p=0.02). Overall, there were 13 out of 53 participants who did not respond to the treatment (24.5% of cases). MMP-8 baseline concentrations were significantly higher in responders (419.3 ± 343.1 ng/ml) than non-responders (158.8 ± 73.3 ng/ml) (p=0.009). MMP-8 concentrations at baseline that were above the cut-off (<182.8 ng/ml) predicted a good response to periodontal treatment with 77% sensitivity and 70% specificity. There was no effect of the single round of non-surgical periodontal treatment on the levels of systemic markers CRP & IL-6, and also there was no correlation between local and systemic markers. Conclusion: These results of both studies suggest that MMP-8, HNP1-3 and S100A8 may be useful to identify cases of periodontitis with good specificity and moderate sensitivity and may give superior results when combined. In addition the salivary MMP-8 and S100A8 showed promising periodontal prognostic ability to detect the likelihood of a good response to treatment, with MMP-8 showed the best results with moderate sensitivity and specificity. However, further validation studies would be useful in larger, non-diabetic cohorts.

Date of Award	2016
Original language	English
Awarding Institution	King's College London
Supervisor	Mark Ide (Supervisor) & Francis Hughes (Supervisor)