Periradicular tissue fluid‐derived biomarkers for apical periodontitis: An <i>in vitro</i> methodological and <i>in vivo</i> cross‐sectional study

Background

Periradicular tissue fluid (PTF) offers a source of diagnostic, prognostic and predictive biomarkers for endodontic disease.

Aims

(1) To optimize basic parameters for PTF paper point sampling in vitro for subsequent in vivo application. (2) To compare proteomes of PTF from teeth with normal apical tissues (NAT) and asymptomatic apical periodontitis (AAP) using high‐throughput panels.

Methodology

(1) To assess volume absorbance, paper points (n = 20) of multiple brands, sizes and sampling durations were inserted into PBS/1%BSA at several depths. Wetted lengths (mm) were measured against standard curves to determine volume absorbance (μL). To assess analyte recovery, paper points (n = 6) loaded with 2 μL recombinant IL‐1β (15.6 ng/mL) were eluted into 250 μL: (i) PBS; (ii) PBS/1% BSA; (iii) PBS/0.1% Tween20; (iv) PBS/0.25 M NaCl. These then underwent: (i) vortexing; (ii) vortexing/centrifugation; (iii) centrifugation; (iv) incubation/vortexing/centrifugation. Sandwich‐ELISAs determined analyte recovery (%) against positive controls.

(2) Using optimized protocols, PTF was retrieved from permanent teeth with NAT or AAP after accessing root canals. Samples, normalized to total fluid volume (TFV), were analysed to determine proteomic profiles (pg/TFV) of NAT and AAP via O‐link Target‐48 panel. Correlations between AAP and diagnostic accuracy were explored using principal‐component analysis (PCA) and area under receive‐operating‐characteristic curves (AUC [95% CI]), respectively. Statistical comparisons were made using Mann–Whitney U, anova and post hoc Bonferonni tests (α < .01).

Results

(1) UnoDent’s ‘Classic’ points facilitated maximum volume absorbance (p < .05), with no significant differences after 60 s (1.6 μL [1.30–1.73]), 1 mm depth and up to 40/0.02 (2.2 μL [1.98–2.20]). For elution, vortexing (89.3%) and PBS/1% BSA (86.9%) yielded the largest IL‐1β recovery (p < .05).

(2) 41 (NAT: 13; AAP: 31) PTF samples proceeded to analysis. The panel detected 18 analytes (CCL‐2, ‐3, ‐4; CSF‐1; CXCL‐8, ‐9; HGF; IL‐1β, ‐6, ‐17A, ‐18; MMP‐1, ‐12; OLR‐1; OSM; TNFSF‐10, ‐12; VEGF‐A) in ≥75% of AAP samples at statistically higher concentrations (p < .01). CXCL‐8, IL‐1β, OLR‐1, OSM and TNFSF‐12 were strongly correlated to AAP. ‘Excellent’ diagnostic performance was observed for TNFSF‐12 (AUC: 0.94 [95% CI: 0.86–1.00]) and the PCA‐derived cluster (AUC: 0.96 [95% CI: 0.89–1.00]).

Conclusions

Optimized PTF sampling parameters were identified in this study. When applied clinically, high‐throughput proteomic analyses revealed complex interconnected networks of potential biomarkers. TNFSF‐12 discriminated periradicular disease from health the greatest; however, clustering analytes further improved diagnostic accuracy. Additional independent investigations are required to validate these findings.