Effect of moisture control on root caries progression adjacent to different restorative materials in an in-vitro biofilm model

Abstract This study aimed to assess the impact of moisture control on the risk of caries progression to root restorations (CAR) made of resin composite (RC) or glass ionomer cements (GIC) under conditions of absolute or inadequate moisture control in an in vitro model. In each of the sixty caries-free human maxillary premolars, standardized mesial and distal cavities were prepared 5 mm apical to the cemento-enamel junction. The specimens were randomly allocated to the experimental groups RC-Z250 (Scotchbond Universal SE & Z250, 3M, Seefeld, Germany), GIC-DF (DeltaFil, DMG, Hamburg, Germany), GIC-EF (Equia Forte, GC, Tokyo, Japan), and GIC-KF (Ketac Fil, 3M, Seefeld, Germany) (n = 15). In each group, restorations were placed either simulating conditions of absolute (Control) or inadequate (Test) moisture control. After water storage for 100d (37°C, 100% humidity), specimens were subjected to thermocycling (10.000 thermal cycles; 5–55°C), and a continuous-culture Lactobacillus rhamnosus biofilm model for 10d. The integrated mineral loss ΔZ was measured along the tooth-restoration interface (dentin wall lesion; DWL) and the outer surface of the root dentin (dentin surface lesion; DSL) using transverse microradiography. Only group RC-Z250 exhibited significant differences in mineral losses ΔZ between Control and Test lesions in DSLs (ΔZ ctrl : 704; 517/1329 vol%×µm; ΔZ test : 1629; 1460/1996 vol%×µm; p < 0.05, Wilcoxon) whereas no significant Control-Test difference was observed for DWLs in RC-Z250 or for DSLs and DWLs in the tested GIC groups. In DSLs, the RC-Z250 Controls demonstrated significantly less mineral loss compared to all GIC groups, while the mineral loss adjacent to the Test-restorations did not differ significantly from that of the other groups (p < 0.05, Kruskal-Wallis/Dunn-Bonferroni). The differences in mineral loss (ΔΔZ = ΔZ test − ΔZ ctrl ) across all groups were not significant (Kruskal-Wallis; p > 0.05). Direct comparison of moisture-associated differences in mineral loss did not reveal significant differences among restorative materials. Under ideal moisture-control conditions, RC-Z250 showed lower mineral loss in DSLs than the investigated GICs, but this finding should be interpreted cautiously due to the simplified model.