Adhesive Strategies for Erosion-Affected Dentin: A Systematic Review and Meta-Analysis of Bond Durability

To systematically evaluate the durability of resin bonding to erosion-affected dentin following adhesive strategies and dentin pretreatments, prioritizing independently replicated aged microtensile bond-strength (µTBS) comparisons.

Methods: PubMed/MEDLINE, Scopus, Web of Science, and Embase were searched through 21 August 2026.

Eligible preclinical studies evaluated bonding to permanent dentin subjected to erosion before adhesive procedures and reported quantitative bond-strength outcomes. Aged µTBS (measured in megapascals [MPa]) was prioritized. Pairwise random-effects meta-analysis was restricted to sufficiently comparable independent studies; Paule-Mandel heterogeneity estimation and safeguarded Hartung-Knapp confidence intervals were used. Single-study and overlapping-cohort effects were reported separately. RoBDEMAT was applied at the item level.

Among 1,163 identified records, 58 retrievable eligible primary reports were retained. At 12-18 months, interface-applied chlorhexidine (CHX) showed numerically lower bond strength than matched erosion-affected dentin controls, but the difference was not statistically significant (k = 3; MD -0.65 MPa; 95% CI -1.83 to +0.52; τ = 0.00). Six-month CHX/protease protocols were not pooled; digitized 72-month DMSO data were kept separate from 48-month data. A hydrophobic resin coating showed a favorable 24-month conditional within-source estimate (MD +17.50 MPa; 95% CI +15.49 to +19.51) from an overlapping publication family. Four-year DMSO and two-year cross-linker findings were favorable in individual studies but lacked independent long-term replication.

Current evidence does not establish a consistently effective bond-preserving pretreatment for erosion-affected dentin. CHX did not demonstrate a detectable benefit at 12-18 months, whereas favorable long-term effects for DMSO, hydrophobic coatings, and selected collagen biomodifiers derive predominantly from single independent studies or one source cohort. Methodological heterogeneity and sparse replication preclude clinical recommendations.

Laboratory bond-strength outcomes should be interpreted in terms of aging resistance rather than immediate adhesion alone; they do not establish improved restoration survival or justify treatment-specific clinical recommendations.