Three-year evaluation of different adhesion strategies in non-carious cervical lesion restorations: A randomized clinical trial*

Abstract

Objective: To evaluate non-carious cervical lesions (NCCLs) restored with different adhesion strategies. Methodology: This is a prospective, randomized, double-blind, split-mouth study. An adhesive restorative system (Single Bond Universal/Filtek Z350XT – SBU) was evaluated both without and with selective enamel conditioning (E-SBU), resin-modified glass-ionomer cements (Vitremer; RMGIC), and ethylenediaminetetraacetic acid pretreatment (EDTA; E-RMGIC). In total, 200 restorations, placed in 50 patients, were evaluated at baseline and at a 3-year follow-up using the modified United States Public Health Service (USPHS) criteria. Data were analyzed using the two-proportion equality test, multinomial logistic regression, Wilcoxon test, and Kaplan-Meier survival curves. Results: In total, 42 (84%) patients returned for the 3-year follow-up. SBU showed restoration losses statistically different from RMGIC. Retention was also statistically different in SBU between baseline and the 3-year follow-up. Marginal defects and surface texture were statistically significant for all groups in the period studied, except for the surface texture of SBU and the marginal integrity in E-RMGIC. We observed no statistically significant difference in wear, secondary caries, anatomical form, surface staining, and color over time. Recession degree was the only factor to influence retention rates. Cumulative survival (%) was 89, 98, 98, and 95.3, for SBU, SE-SBU, RMGIC, and E-RMGIC, respectively, without significant differences among them. There was a statistically significant difference between survival curves; however, multiple comparison procedures found no statistical differences. Conclusion: Selective enamel etching affected the retention of non-carious cervical restorations. Adhesion using EDTA and resin-modified glass-ionomer cements delayed marginal defects over time. The degree of gingival recession influenced retention rates. Resin composite restorations showed initial marginal defects, and ionomer restorations, reduced surface luster. EDTA pre-treatment followed by resin-modified glass-ionomer cements may be a promising adhesion strategy for NCCL restorations.