marginal integrity of cervical restorations with caries-affected dentinal walls: effect of contamination with hemostatic agents

Objectives: the aim of this study was to compare the microleakage in normal and caries- affected dentin (cad) and to investigate the effect of three hemostatic agents on the microleakage of class v composite resin restorations in cad. materials and methods: ninety-six class v non-beveled cavities were prepared in 48 third molars at 1 mm below the cementoenamel junction (cej) in the cervical margin with the occlusogingival size of 2 mm, mesiodistal dimension of 3 mm, and a depth of 1.5 mm. next, the teeth were divided into 8 groups (n=12): g1-4 included normal dentin (n) substrate, while g5-8 were exposed to mineralization/demineralization cycles to produce cad substrate. groups 1 and 5 were the controls. viscostat was used in groups 2 and 6, viscostat clear was used in groups 3 and 7, while trichloroacetic acid (tca) was used in groups 4 and 8. the cavities were restored with composite resin. the samples were sectioned after thermocycling and immersion in 2% fuchsin for 24 hours. the degree of dye penetration was evaluated under a stereomicroscope at 40× magnification. data were evaluated using kruskal-wallis and mann-u-whitney tests in spss 15 software (α=0.05). results: significant differences were recorded on the mean microleakage of different groups (p=0.047). there was a significant difference in the mean dentinal microleakage between n and cad groups (p=0.014). the dentinal microleakage in group 3 was significantly higher than that in groups 4 to 8. conclusions: according to the results, cad showed less microleakage in comparison with intact dentin. viscostat clear caused a greater microleakage than did viscostat or tca.