EFFECTS OF AGING METHODS ON MICROLEAKAGE OF AN ADHESIVE SYSTEM USED AS A SEALANT ON CONTAMINATED SURFACES

1MSc, PhD student, Department of Dental Materials, Dental School, University of São Paulo, SP, Brazil. 2PhD Professor, Department of Dental Materials and Operative Dentistry, University of Oeste de Santa Catarina (UNOESC), Joaçaba, SC, Brazil. 3PhD, Professor, Department of Statistics, Institute of Mathematics and Statistics, University of São Paulo, SP, Brazil. 4PhD, Associate Professor, Department of Dental Materials, Dental School, University of São Paulo, SP, Brazil.


INTRODUCTION
Whereas clinical trials are the best way to provide valuable information about the performance of dental materials 17 , they are expensive and time-consuming.Furthermore, it is likely that the products under investigation might be discontinued by the time the study is published.
Microleakage tests are often relied upon predictors of clinical performance for new generation adhesive-restorative approaches and materials.These tests usually include thermal cycling to simulate intraoral conditions 1,17 , which requires exposure of specimens to water baths set at temperatures that resemble those found in the oral cavity.The rationale behind this is that marginal percolation can be caused by differences in the coefficient of thermal expansion between dental tissues and the restorative material.The polymerization shrinkage, poor bonding of restorative materials to tooth tissues and coefficient of thermal expansion are blamed as the main factors for leakage 17 .One concern of the thermal cycling regimen is that the temperature variations between 5 and 55°C do not necessarily mean that this is the temperature variation that causes damages to the bonding interfaces 11 .
Apart from thermal stressing, the presence of water has been generally overlooked in microleakage studies.Bond breakdown may be enabled or facilitated by hydrolysis of the bonding agent and its union with tooth tissue 17 .It is widely known that water acts as plasticizer of polymers and reduces the marginal integrity of restorations 2,3 being a stress-raising factor.
Pit and fissure sealing is a successful clinical procedure to prevent caries development and it is routinely indicated for teeth under caries risk 5,12,24,26,28 .Sometimes, the success of this technique is dependent on the application of the material during the teeth erupting stages.Under such clinical conditions the use of rubber dam 6 , to reduce the risk of contamination by saliva, crevicular fluid and/or blood is almost impossible 10,15 .
Microleakage and partial/total loss of sealants due to salivary contamination are the main concerns regarding their long-term success 15 .Therefore, adhesive systems have been proposed as an intermediate layer between enamel and sealants for minimizing deleterious effects of non detected moisture contamination, and such combination has being analyzed via in vitro 15,16,22 and in vivo 8 studies.
A more recent approach indicates the sole use of filled adhesive systems for sealing, mainly when contamination is likely to occur 4,13,29 .It was observed that OptiBond dual cure 29 and OptiBond FL 4 showed less microleakage values whether applied on contaminated enamel surfaces or not, compared to two conventional sealants (Delton and Fluroshield) and three adhesive systems (Scotchbond MP, All Bond and SOLO).This result was also maintained even when the specimens were subjected to thermal and/or load cycling 4,29 .
To date, no study has separately addressed the effects of water and temperature changes on the marginal sealing, which was the main objective of the present investigation.
Therefore the objective of this study was to analyze the effects of water contact and aging methods on the marginal sealing of an adhesive system used as a sealant for contaminated condition.

MATERIALS AND METHODS
The Ethics Committee at Dental School, University of São Paulo, approved this study.Sixty caries-free extracted human molars and premolars, free of defects/fractures, were cleaned and immersed in a 10% formaldehyde solution for some days.After that they were stored in distilled water for no longer than 4 months.The roots and cervical region of crowns were embedded in acrylic resin and the teeth were pumiced and rinsed thoroughly with water using an ultrasonic cleaner (Thornton, Inpec Electronics, Vinhedo, SP, Brazil).
After surface conditioning with 32% phosphoric acid for 30 seconds (Uni-etch, Bisco Inc, Schaumburg, IL, USA; batch 9900010368), the sample was contaminated with 1 microliter of human plasma (Pharmaceutical School -University of São Paulo) for 30 seconds, followed by material application, as shown in Table 1 a -etch (30 s); b -rinse with distilled water (30 s); c -air-dry (15 s); d -contamination with 1mL of human plasma (30 s); e -primer application (brushed for 30 s and air-dry for 5 s at 30 cm distance); f -bonding agent; g -light activation (30 s).

EFFECTS OF AGING METHODS ON MICROLEAKAGE OF AN ADHESIVE SYSTEM USED AS A SEALANT ON CONTAMINATED SURFACES
intensity increasing from 100 to 1000 mW/cm 2 in a period of 20s.The light intensity was checked using the own light curing unit radiometer.The specimens were then allocated into the following groups: 1) thermocycling-4,000 cycles, 200 per day, at 5 and 55 o C with a dwell time of 60 s in water environment; 2) thermocycling-4,000 cycles, 200 per day, at 5 and 55 o C with a dwell time of 60 s without water contact; 3) storage in distilled water for 20 days at 5 o C (thermocycling duration time); 4) storage without water contact for 20 days at 5 o C; 5) storage in distilled water for 20 days at 37 o C; 6) storage without water contact for 20 days at 37 o C; 7) storage in distilled water for 20 days at 55 o C; 8) storage without water contact for 20 days at 55 o C; 9) immediate or control group, with no aging treatment.All specimens, except those from the control group, were maintained immersed in water for 24 h.The experimental design is depicted in Figure 1.
The specimens contact with water was avoided by means of an individual custom tray, fabricated for each tooth from a 0.2 mm vacuum-formed soft plastic (Nite White, Discus Dental, LA, CA, USA).The custom tray was then glued to the acrylic resin surface with cyanoacrylate resin Super Bonder (Henkel Ltda., Itapevi, SP, Brazil), in order to avoid water penetration during the aging method and only allow temperature changes within tooth.In order to confirm if the temperature changed during thermocycling, a thermocouple was placed within the pulp chamber in some teeth and they were subjected to the thermocycling procedure, validating the method employed.
Subsequently, the crowns of the specimens from all groups were coated with three layers of nail varnish, except from the area 1 mm away from the adhesive interface, and immersed in a 50% AgNO 3 aqueous solution (pH > 7.0) in darkness for 2 hours.Afterwards, they were rinsed off, immersed in a photodeveloping solution and exposed to fluorescent light for 8 hours 20 .
Each specimen was sectioned twice, through the sealed fissure system, with a microtome (Labcut 1010, Extec Co., Enfield, MT, USA) in buccal-to-lingual direction generating 4 cross-sections per tooth that were digitized (ScanJet 4c, HP, Palo Alto, CA, USA) to obtain 600dpi TIFF images.On each cross-section the total linear interface between adhesive and enamel in the buccal and lingual slopes, as well as the extent of the dye penetration were measured (mm) with IMAGELAB (image analysis software developed at FOUSP, Brazil) and expressed as a percentage of the total adhesive interface length.For each tooth, the percentages of dye penetration along the cross-section interfaces were averaged.
An unbalanced ANOVA with empty cells was employed to evaluate the effects of the main effects as well as their interaction with respect to the mean value of leakage 18 .The significance level considered was 5%.

RESULTS
Table 2 contains the descriptive measurements per group and Figure 2 shows the mean profile plots with the respective standard errors.
The analysis of variance indicated significant aging methods x water interaction (p=0.033).It is worthwhile to note that the significant water x aging interaction suggests that the differences between the overall dye penetrations for the different methods depend on the storage condition (with water contact or not).The main factors, water (p=0.448)and aging (p=0.186), were not statistically significant.
The Tukey test for multiple comparisons was applied to the different combinations of aging methods and water.Eight of the thirty-six comparisons were statistically significant (considering the level of significance as 5%), five of them corresponding to storage at 5 o C with no water.Results of the Tukey test suggested that the overall dye penetration percentage for teeth stored at 5 o C with no water contact is  smaller than that for the other methods, which were statistically equivalent.These findings were incorporated into a linear model, which confirmed the above results with statistically significant effect (p=0.015).Under this approach, the overall dye penetration percentage from the group at 5 o C under no water contact was 15.9 ± 11.0 (mean ± standard error).For the other groups, the overall dye penetration percentage was 44.9 ± 3.7.
The measurements were repeated, some time later, for sixteen specimens and the intra-class correlation coefficient was calculated as a measure of reliability; the result was 0.81, indicating excellent agreement 9 .

DISCUSSION
The marginal integrity of the tooth/material interface is dependent upon many factors like the physical properties of the material and dental tissue, as well as their interactions with the environment 23 .It is well accepted that a clean, uncontaminated bonding field is mandatory for achieving effective adhesion to tooth structures 5 .The shrinkage that occurs as the material polymerize has been considered the main factor for bonding failures and it may lead to leakage around the restoration and subsequent secondary caries 17 .The linear coefficient of thermal expansion has been noted as other important contributing factor to leakage.However, in a clinical situation, the dimensional change may not be as great as the linear coefficient of the material would predict, once extreme temperatures seem to occur in short periods of time 23 .The clinical relevance of the linear coefficient has not been clarified yet 11 since some time is required for a non uniform temperature body to approach the equilibrium 11,23 .
The results of many clinical trials and in vitro studies depicted the longevity and effectiveness profile of sealant technique 12,15,24 , although there are reports showing decreased retention rates over time 12,15 .The application of hydrophobic materials to acid-etched enamel susceptible to moisture contamination is likely the major cause of reduced retention of fissure sealants 6,7,10,13,15 .Therefore, the use of hydrophilic adhesive systems to seal fissures, followed 15 or not 13 by hydrophobic sealants, is of special interest since they could overcome the negative effects of some casual contamination during the bonding process 6 .
As in previous studies no significant differences between microleakage values for dry and contaminated condition were observed when OptiBond was used 4,13,29 .Based on that and in the fact that the use of an adhesive system as sealant is proposed in cases where there is risk of contamination, the authors of the present study decided not to use dry enamel condition.
The present results show that aging at 5 o C, without water contact, has an overall dye penetration value of 15.9 ± 11.0%.It is likely that the diffusion coefficient is lowest at 5 o C than at 35 o C and that the mobility of the polymer molecules are reduced 25 , which in turn may restrict the tracer penetration along polymer/enamel interface, since there is no water contact.Although at 5 o C in a dry condition lower microleakage was observed, it does not clinically simulate the oral environment.All other experimental groups, except the group maintained at 5 o C in a dry condition, showed overall leakages that were statistically similar (44.9 ± 3.7%).This implies that the microleakage degree taken from the control group was statistically similar to the specimens that were thermocycled or stored in one of the aging methods.
For instance, in a previous study 29 , the microlekage values of three adhesive systems applied for sealing occlusal surfaces after thermal cycling (4,000 times; 5-55 o C) were compared.The authors also kept some specimens in water at 37 o C for 72 hours, as control, to verify the interface integrity after a short storage time (to allow possible water sorption and residual polymerization).Under the control condition, dye penetration was observed for all materials regardless the contamination condition; OptiBond showed less overall leakage (11 ± 5%) than the other adhesives (21 ± 6%).After thermal cycling, OptiBond dual cure continued showing the least overall leakage value (28 ± 6%) than All Bond 2 and Scotchbond MP that showed similar values (38 ± 6%).This clearly indicates that microleakage adversely affected early defective interfaces; however, the overall conclusions were not affected by the thermocycling procedures.
As in the study described above, none of the experimental conditions from the present study achieved an initial complete seal.Therefore, it seems that there is no point in subjecting the specimens to thermal cycling.Apart from the thermal changes, other factors such as water sorption may take place and lead to misinterpretation of microleakage results.The effect of water as a polymer plasticizer has been currently overlooked in microleakage studies 11,21 and its effects on thermal cycled specimens should be further evaluated.
The results regarding microleakage coincide with others described in the literature.Neither the storage at a constant temperature or under thermal cycling, nor the dwell period or the number and temperature of cycles seem to significantly affect microleakage 11,14,20 .No difference in microleakage between thermocycled and non-thermocycled groups were detected; breakdown and leakage have also been identified on non-thermocycled specimens interface 19,23,27 .The middleterm storage (3 months) and thermal cycling (500x, 10-60 o C) did not enhance the severity of the leakage since the worst

CONCLUSIONS
Considering the experimental design and the limitations of this study regarding a clinical situation simulating a contaminated sealed surface, it was concluded that: 1) all groups showed dye penetration, even the control group; 2) storage teeth at 5 o C on dry condition provided the lowest degree of dye penetration; 3) in the presence of water, variations seem not to affect the microleakage values.

TABLE 2 -
Mean dye penetration and standard errors (%) Same superscript letters indicate no statistically significant difference (p>0.05)FIGURE 1-Experimental design