MICROHARDNESS EVALUATION AROUND COMPOSITE RESTORATIONS USING FLUORIDE-CONTAINING ADHESIVE SYSTEMS

1DDS, MSc, PhD, Assistant Professor, Department of Restorative Dentistry, Dental School, Federal University of Minas Gerais, Belo Horizonte, MG, Brazil. 2DDS, MSc, Graduate Student (Doctor Degree), Department of Restorative Dentistry, Piracicaba Dental School, State University of Campinas, Piracicaba, SP, Brazil. 3DDS, MSc, PhD, Graduate Student (Doctor Degree), Department of Restorative Dentistry, Piracicaba Dental School, State University of Campinas, Piracicaba, SP, Brazil. 4DDS, MSc, PhD, Associate Professor, Department of Restorative Dentistry, Ribeirão Preto Dental School, University of São Paulo, Ribeirão Preto, SP, Brazil. 5DDS, MSc, PhD, Associate Professor, Department of Restorative Dentistry, Piracicaba Dental School, State University of Campinas, Piracicaba, SP, Brazil.


INTRODUCTION
Adhesive systems have promoted high bond strengths to dental hard tissues 5,7,21 .However, secondary caries is still the reason most commonly cited for the replacement of composite restorations 3,9,14 .The deficient marginal adaptation and lack of cavity sealing may allow fluid and bacteria infiltration, leading to caries lesions development in the tooth structure along the restorative interface 20 .In fact, an in vivo study has shown that at gap-free regions there is no wall lesion development, emphasizing the critical role of marginal sealing in restoration integrity 12 .
Besides the sealing properties, secondary caries incidence can be assigned to the anticariogenic effect of restorative materials.Thus, approaches that inhibit the ingress of metabolic products of biofilm into the microgap between tooth and restoration, increase resistance to demineralization, and/or facilitate remineralization are of considerable clinical interest 6 .
As numerous investigations have shown the ability of fluoride-containing materials to inhibit caries activity around restorations, continuing research into the development of other materials for delivering fluoride in a form that will be effective is ongoing 3,6,9,14,16,18,11,19 .Glass-ionomer and resinmodified glass ionomer cements have been formulated as cavity liners 10,15 .Also, many bonding agents, total-etch dentin adhesives, and one-bottle adhesives containing fluoride have become available.The objective of these materials is to produce long-lasting restorations, providing protection against secondary caries.
Even though fluoride-releasing products directly contact the cavity wall, there is little evidence regarding the amount of fluoride ions released from them, their diffusion into the cavity wall and their effects on surrounding tooth structure 4,9,14,17,18 .
Because bonding agents play a crucial role and are strategically positioned in a restoration, this in vitro study evaluated the enamel microhardness around composite restorations bonded with fluoride-containing one-bottle adhesive systems, after thermal and demineralization/ remineralization cycling protocols.The specific null hypotheses tested were: (1) there are no differences among the enamel microhardness around composite restorations bonded with different adhesive systems and techniques; and (2) there are no differences of microhardness along different distances from the cavity wall.

Experimental Design
The factors under study were adhesive systems/ restorative techniques at seven levels and distance from the restoration cavity wall at three levels (150, 300, 450mm).The experimental units were 175 dental fragments (n = 25), randomly restored in 25 blocks of 7 fragments each.Each block contained one fragment of each one of the seven groups.A randomized complete block design was used to systematically control the variability arising from known nuisance sources.As it was not possible to completely randomize the order in which the distances from the cavity wall were analyzed, this investigation characterized a factorial 7x3 split-plot design 11 .The continuous response variable was Knoop microhardness.

Preparation of samples
Forty unerupted, extracted human third molars, stored in 2% formaldehyde solution (pH = 7.0) for no longer than two months after extraction were used in this study.The teeth were cleaned of gross debris and placed in deionized water for 24h at 8 o C prior to beginning of the experimental procedures.The crowns were sectioned longitudinally to obtain 175 dental fragments (5mm width x 5mm length x 4mm height) using a diamond blade (#7020 -KG Sorensen Ind. e Com.Ltda, Barueri, SP, Brazil).

Cavity preparation and restorative procedures
Standard cylindrical-shaped cavities (2mm diameter and 1.5mm depth) were prepared on the enamel surface of each fragment, using a cylindrical diamond bur (#2096, KG Sorensen) mounted in a high-speed turbine with air-water coolant.Dental fragments were then randomly assigned into 7 groups (n = 25), according to the adhesive systems/ restorative techniques studied.
All materials, adhesives and restoratives, were applied according to manufacturers' instructions (Table 2).Bonded cavities were bulk-filled with microfill composite (Durafill VS, Heraeus Kulzer GmbH) and light-cured for 40 s (Optilux 500, Demetron Corp., Danbury, CT, USA).A group comprised "Sandwich" technique restorations 10 with Vitrebond as liner (0.7 mm thickness) and Durafill Bond / Durafill VS restorative system.Single Bond adhesive was used in two groups: one group was subjected to the thermo and pH-cycling protocol and the other was not.Experimental groups were as follows: Group 1-Restored with Single Bond adhesive + Durafill VS microfilled composite (not submitted to cycling protocols and was kept in 100% relative humidity at 37 o C).
Group Restored samples were stored in 100% relative humidity for 48h at 37 o C and polished with Sof Lex discs (3M ESPE), ensuring that there were no overhangs at margins.

Thermo-and pH-cycling
Except for Group 1, all experimental groups (2 to 7) were submitted to thermal and demineralization/remineralization cyclings 19 .The restored enamel surfaces were covered with nail varnish leaving exposed only the restorations and 2mm of enamel surface around them.These specimens were thermocycled (MCT2 -AMM, Instrumental Ltda, São Paulo, SP, Brazil) for 1,000 cycles in distilled water at 5 o C and 55 o C with a dwell time of 1 min.
Each restored sample was placed in 15mL of demineralizing solution (2.0mM of calcium, 2.0mM of phosphate in a buffer solution of 74mM of acetate at pH = 4.3) for 6h at 37 o C, followed by rinsing with deionized water and immersion in 15mL of remineralizing solution (1.5mM of calcium and 0.9mM of phosphate in a buffer solution of 20.0mM of tris (hydroxymethil)-aminomethane at pH = 7.0) for 18h at 37 o C.After five pH cycles had been performed, each specimen was stored in 15mL of remineralizing solution for 2 days at 37 o C.

Preparation of specimens for microhardness measurements
At the end of each treatment period, restored specimen were rinsed in deionized water and longitudinally sectioned through the center of the restoration with a slow-speed diamond saw (KG Sorensen Ind. e Com.Ltda).A half of the specimen was embedded in polystyrene resin and its restored surface was wet-polished with 600-and 1000-grit SiC (Carborundum Abrasivos Ltda, Vinhedo, SP, Brazil).Final polishing was performed with diamond abrasive pastes (6µm, 3µm, 1µm and 0.25µm -Buehler Ltd., Lake Bluff, IL, USA) on cloths, under mineral oil lubrication.
Microhardness measurements were carried out on the leveled enamel subsurface of each dental slab with a microhardness tester (FM-1, Future Tech, Tokyo, Japan) and a Knoop indenter.Nine indentations with 25g applied for 5s were made in each sample.Three measurements were taken at the distance of 150µm, 300µm and 450µm from the  restoration cavity wall.For each distance indentations were taken at 30 to 90µm from the outer enamel surface.The microhardness values derived by the equation: Knoop Hardness = 14229.K / L 2 , where K is the applied load in grams, and L is the indentation length, in micrometers.

Statistical analysis
After checking of the assumptions of homogeneity of variance (Hartley's test) and normal distribution (Shappiro-Wilks), the response variable cross-sectional enamel Knoop microhardness was analyzed using a two-way analysis of variance (ANOVA) considering the significance level of 5%.The effect of the factors Adhesive System and Distance from the restoration was evaluated, as well as the interactions of those variation sources.Multiple Comparison Duncan's test was applied to detect significant differences among groups (a = 0.05%).

RESULTS
Mean values of microhardness for restorative procedures and distances from the restoration cavity wall are displayed in Tables 2 and 3, respectively.ANOVA did not show significant interaction between adhesive systems and distances from the restoration (p = 0.98).Significant effects were detected for adhesive systems (p = 0.0083) and distance from the cavity wall (p = 0.0188).
Duncan's test showed that the group restored with Single Bond not subject to both, thermal and pH cycling, presented the highest microhardness mean around the restoration (p < 0.05).Groups restored with Single Bond, Vitrebond / Durafill Bond -Durafil VS (NC) and Tenure Quick exhibited higher microhardness than Optibond Solo, Prime&Bond 2.1 and Syntac Sprint.Prime&Bond 2.1 and Syntac Sprint presented the lowest microhardness values and were not statistically different from each other (Table 2).
The comparison of microhardness at different locations (Table 3) indicated that the first two measurements that were closer to the cavity wall (150 and 300µm) presented lower microhardness means than the values obtained at 450µm from the cavity wall.

DISCUSSION
In vitro evaluations of induced caries-like lesions around restored cavities are important to analyze the behavior of restorative materials on the development of caries 3,9,14,[16][17][18][19] .On the basis of this statement, this study was designed to verify the effects of enamel demineralization adjacent to composite restorations bonded with fluoride-containing adhesive systems.The enamel demineralization was produced by a pH-cycling model that tries to simulate highly cariogenic oral conditions.
High quality sealing of the cavity margins is necessary for the establishment of long lasting restorations 20 .Even though bonding of composite restorative materials to enamel has been considered a durable and effective procedure, microleakage may occur and be detrimental for the longevity of the restorations.Thermal cycling regimen was adopted to simulate the temperature extremes conform those found in the oral cavity.Resin adhesive volumetric shrinkage and expansion 2 during thermal cycling may stress and damage the bond.Marginal gaps formation and/or presence of microporosities at the bonded interfaces could favour pH cycling-induced demineralization of the cavity wall 8,19 .In this current study, the analysis of microhardness showed lower values at 150µm and 300µm from restoration, which can be related to restorations effects.
Because dental caries comprise a process of alternating demineralization and remineralization phenomena, enamel microhardness reduction adjacent to bonded restorations were promoted by a dynamic model that simulates in vivo conditions of high caries risk 19 .The effectiveness of the adopted cariogenic challenge in producing demineralization was confirmed, as there was statistically significant difference between cycled (Group 2) and non-cycled (Group 1) specimens bonded with Single Bond.Following thermoand pH-cyclings, all groups presented lower microhardness than Group 1, which represented sound enamel around restoration.Thus, no restorative technique completely avoids the demineralization areas formation around restorations.However, some restorative procedures provided higher enamel microhardness values around   composite restorations than others, suggesting a different behavior of composite restorative/bonding agent systems.The Single Bond one-bottle adhesive that does not contain fluoride, the "sandwich" technique that comprised the association of glass-ionomer cement/composite resin 10,15 and a fluoride-containing one-bottle system (Tenure Quick) provided lower demineralization than the other restorative procedures.Therefore, the reduction of microhardness was not dependent on fluoride presence in the adhesive composition.Cariostatic properties of glass-ionomer cements are related to the fluoride release, which can reduce demineralization on enamel around restorations 1,14,16,19 .However, the marginal sealing promoted by Durafil Bond adhesive resin, which is not a fluoride-containing adhesive may have blocked the direct access to fluoride from the resin-modified glass-ionomer (Vitrebond) to adjacent demineralized areas, when the "sandwich" technique was used.Even though "sandwich" restorative technique did not reach microhardness values obtained for Group 1 (Single Bond) that was not submitted to thermal and pH-cyclings, it showed higher microhardness than three fluoride-containing adhesive systems (OS, PB and SS).
Simplification of bonding procedures led manufactures to combine the primer and adhesive resin into a single bottle.These bonding agents are directly placed on dentin and enamel following the "total-etch technique" with phosphoric acid.One-bottle adhesive systems have also been evaluated regarding the bond strength and interfacial micromorphology 5,713,21 and some of these studies have demonstrated that Tenure Quick provided low bond strength values and incomplete infiltration of monomers into the mineralized dental structures 7,13,21 .However, in this study, it was among the total-etch fluoride-containing adhesives that produced the highest microhardness values, i.e., lowest reduction on enamel demineralization adjacent to composite restoration.Although manufacturers do not show the source of fluoride and do not explain the fluoride-releasing mechanism, it is possible that the concentration of fluoride ions and the mode of releasing may have influenced results 4,8,19 .
Studies have shown that Single Bond adhesive promotes high bond strengths and a consistent adhesion to tooth substrate, however the efficacy of Prime&Bond 2.1, Optibond Solo and Syntac Sprint has not been established 7,13,21 .For Prime&Bond 2.1, the source of fluoride is provided by cetylamine hydrofluoride, while for Optibond Solo, sodium hexafluorosilicate is its source.Both adhesives presented similar results, but higher than Syntac Sprint.Although fluoride presents a dynamic action mechanism on the progression/reduction of incipient caries lesions and a predictable adhesion to enamel has been achieved, the demineralization promoted by caries-like lesions adjacent to restorations bonded with fluoride-containing adhesives was not inhibited.Because of differences among the enamel microhardness around composite restorations and differences of microhardness along different distances from the cavity wall, both null hypotheses tested were rejected.

CONCLUSIONS
Enamel microhardness around composite restorations varied depending on the restorative technique.Fluoridecontaining adhesive systems were not necessarily associated with higher microhardness values.Close to the cavity wall (150 and 300µm), the formation of demineralized areas was more evident.