Host markers of periodontal diseases: Meta‐analysis of diagnostic accuracy studies

Abstract

To identify host markers with optimal diagnostic performance for clinical implementation in the diagnosis of periodontal diseases and prediction of future disease progression and/or disease resolution. Cross-sectional and prospective studies with ≥ 20 participants per group, reporting diagnostic accuracy (e.g., area under the curve [AUC]) of host markers for periodontal diagnosis (focused question 1 [FQ1]), periodontitis progression/relapse (FQ2) or resolution (FQ3) were searched in three electronic databases. Meta-analyses estimating diagnostic accuracy (DA) for individual host markers and for grouped salivary and gingival crevicular fluid (GCF) markers independently were performed whenever two or more studies were identified. Sixty-one eligible studies were identified, of which 13 were included in meta-analyses for FQ1 (discrimination between health and periodontitis). Matrix metalloproteinase-8 (MMP-8) was the most reported biomarker in both saliva and GCF, with comparable AUC (0.70–0.90), sensitivity (0.49–0.84) and specificity (0.62–0.79) in both sample types. Cytokines had good ability for discrimination of periodontitis/gingivitis versus health, although they were substantially less accurate for periodontitis versus gingivitis. Combinations of cytokines and MMPs tended to increase overall diagnostic accuracy but without significant improvement in the case of periodontitis/gingivitis discrimination. Bone markers were the best performing group of salivary markers (AUC = 0.91) when compared to cytokines (AUC = 0.86) and MMPs (AUC = 0.77). GCF microRNAs (MiRs) were a singly meta-analysed group of biomarkers demonstrating AUC = 0.79.

Sandro Radovanović

Assistant Professor at University of Belgrade

My research interests include machine learning, development and design of decision support systems, decision theory, and fairness and justice concepts in algorithmic decision making.