Gingival Crevicular Fluid Biomarker Identification for Diagnosing Periodontitis by SWATH-MS

Objectives: To identify new molecular biomarkers with a high diagnostic capacity to detect untreated and treated periodontitis in gingival crevicular fluid (GCF).
Methods: GCF samples were collected from 44 periodontally healthy subjects and 41 with periodontitis (stage III-IV). In the latter, 25 underwent non-surgical periodontal treatment and showed clinical improvement after two months. Samples were analysed using the sequential window acquisition of all theoretical mass spectra (SWATH-MS) and proteins were identified using UniProt human-specific database. The diagnostic capability of the proteins was determined with generalised additive models (GAM) to distinguish untreated periodontitis from periodontal health (first modelling), treated periodontitis from periodontal health (second), and untreated periodontitis from treated periodontitis (third). The area under the curve (AUC) and other classification measures were calculated.
Results: A total of 250 proteins were quantified. For distinguishing untreated periodontitis from periodontal health, 29 biomarkers had AUC values ≥0.850, and for 48 dual combinations was ≥0.990 (sensitivity/specificity≥95%). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) achieved the highest values (bias-corrected -bc- AUC=0.972; bc-accuracy=92.4%), followed by zymogen granule protein 16 homolog B (bc-AUC=0.962; bc-accuracy=89.4%). Their combination with matrix metalloproteinase-9 and cornulin, respectively, obtained a bc-AUC=0.999 and bc-accuracy=99.9%. For detecting treated periodontitis, 47 proteins had an AUC≥0.850 and for 46 pairs was ≥0.995 (sensitivity/specificity≥95%). Carbonic anhydrase 1 achieved a bc-AUC=0.930 and bc-accuracy=91.2%, while three combinations had a bc-AUC=1 and bc-accuracy=100% (haemoglobin subunit -Hb- delta with actin cytoplasmic 1, Hb alpha with GAPDH, and annexin A1 with peroxiredoxin-1). In the third modelling, no protein individually or combined achieved an accuracy of ≥60%.
Conclusions: New single and dual-combination biomarkers in GCF were identified with outstanding capability to distinguish untreated and treated periodontitis from periodontal health using SWATH-MS. Pre-treatment and post-treatment periodontitis after two months were indistinguishable, even associating two proteins. However, these findings need to be validated using other techniques.