Early prediction of orthodontic gingival enlargement using S100A4: a biomarker-based risk stratification model.

Abstract

Orthodontic-induced gingival enlargement (OIGE) affects approximately 15-30% of patients undergoing orthodontic treatment and remains largely unpredictable, often relying on subjective clinical assessments made after irreversible tissue changes have occurred. S100A4 is a well-characterized marker of activated fibroblasts involved in pathological tissue remodeling. This was a cross-sectional precision biomarker study that analyzed gingival tissue samples from three groups: healthy controls (n = 60), orthodontic patients without gingival enlargement (n = 31), and patients with clinically diagnosed OIGE (n = 61). Immunohistochemical analysis quantified S100A4-positive fibroblasts, type I collagen synthesis, and microvascular density. Advanced statistical analyses included multivariate logistic regression, machine learning-based validation, causal mediation analysis, and survival modeling for risk stratification. The density of S100A4-positive fibroblasts was significantly higher in OIGE patients (245.8 ± 38.7 cells/mm²) compared to orthodontic controls (165.3 ± 29.4 cells/mm²) and healthy individuals (98.2 ± 18.5 cells/mm²) (p < 0.001; η² = 0.891). Multivariate analysis confirmed S100A4 as an independent predictor of OIGE (OR = 1.028 per cell/mm²; 95%CI 1.021-1.035; p < 0.001). Machine learning validation demonstrated high predictive accuracy (AUC = 0.946). Survival analysis identified distinct risk strata: individuals with S100A4 densities > 180 cells/mm² had a 78% probability of developing OIGE within 24 months, compared to 12% for those with < 130 cells/mm². S100A4 demonstrates 95% predictive accuracy for OIGE, supporting its role in personalized risk stratification and early preventive interventions during a defined therapeutic window. This study presents the first validated precision biomarker in orthodontics with the potential to prevent an estimated 180,000-360,000 OIGE cases globally each year.