INTRODUCTION

As adhesive technology continues to evolve and newer and improved adhesive and composite systems are developed, the use of posterior composites continues to gain popularity. A number of the early drawbacks of posterior composite restorations no longer represent a challenge with currently available materials. Improved dentin adhesive systems offer increased bond strengths and enhanced marginal sealing.1 Stronger, more wear-resistant composite systems have been developed for use in stress-bearing areas. However, polymerization shrinkage remains an unavoidable problem, which may compromise the integrity and longevity of posterior restorations,2 particularly those with margins in dentin.3 The potential for marginal gaps and microleakage increases with cavosurface margins in dentin due to the biological variability of this tissue.4 The presence of shrinkage-induced gaps at the composite-tooth interface may lead to post-operative complications, such as restoration fracture, leakage, sensitivity, staining and recurrent caries,2 all of which may contribute to early failure of the restoration.

Approaches to minimize the adverse effects of polymerization shrinkage and gap formation have primarily focused on the modification of conventional placement techniques, including incremental placement,5–6 soft-start polymerization7–9 and the use of liners with a low modulus of elasticity as stress relievers.10

The use of liners with a low elastic modulus as the first increment under composite restorations has become well accepted in recent years, with the availability of materials, such as glass ionomers and flowable resins. However, the benefit of using flowable liners to reduce microleakage and improve gingival margin adaptation is still controversial, with studies showing improvement,11–14 no effect15–20 and even deterioration of the marginal adaptation and microleakage.21–22

A number of studies showing improved margin adaptation with flowable liners have attributed their results to reduced viscosity of the material, which allowed it to better wet the preparation walls.1123 The use of flowable liners, however, involves a restorative compromise. Their enhanced flowability may contribute to better wetting of the preparation walls, as well as providing a stress-absorbing layer during polymerization.10 On the other hand, the reduced viscosity is achieved by lowering the filler concentration. Consequently, physical properties, such as polymerization shrinkage and strength, are negatively affected.24–25 The thickness of the flowable resin liner and whether the liner is polymerized prior to or simultaneously with the overlaying resin have also been investigated, with the hypothesis that thinner liners may result in better margin adaptation, since less polymerization shrinkage is expected to occur in thinner increments.1113

A commercially available unit (Calset, AdDent Inc, Danbury, CT, USA), which preheats resin composites prior to their application into the preparation, offers promise. By preheating the composite to one of the preset temperatures, 98°F (37°C), 130°F (54°C) or 155°F (68°C), a transient viscosity reduction comparable to that of flowable resins can be obtained.26 A high strength, highly filled material with less shrinkage could then be used as the initial gingival increment, while still being able to take advantage of the transient viscosity reduction of the material for an enhanced wetting of the preparation walls. Preliminary studies have demonstrated improved flowability and handling characteristics of preheated resins without alteration of their physical properties.27 A strong positive correlation between temperature and monomer conversion has also been demonstrated in-vitro with preheated composites requiring less light exposure to achieve higher conversion values compared to room temperature photopolymerization. 28–29 By achieving higher final monomer conversion values, the amount of residual unreacted monomer, which may potentially leach into the oral cavity, is also reduced.28

Limited information is available on the effect of preheated composites on the microleakage and marginal adaptation of Class II composite restorations. A laboratory study found significantly less microleakage at the cervical margin of Class II restorations with preheated composites, compared to the use of a conventional unheated hybrid or flowable liner.30 With resin composite restorations, it has been widely assumed that microleakage and margin adaptation can be directly correlated.31 However, this relationship may not be as clear-cut as is widely assumed.32–33 Microgaps present at the restoration margin may leak, but this may not always be the case. Conversely, penetration of nano-size particles from the dye may be observed, irrespective of the presence of gaps at the margin. Given that microleakage and margin adaptation represent different outcome measurements, direct comparisons with other studies in the literature reporting on the effect of preheated composites on microleakage29–30 are not appropriate. The authors are not aware of any studies reporting on the effects of preheated composites on the gingival margin adaptation of Class II composite restorations.

The present study evaluated the effect of preheated composites (PHC) and flowable liners (FL) on the gingival margin gap formation of Class II composite restorations as compared to the conventional placement of room temperature composite (RTC) 24 hours after polymerization. Furthermore, this study aimed to evaluate the effect of two different placement techniques with flowable liners and two different temperatures of preheated composite on Class II gingival margin gap formation. The null hypothesis was that there would be no difference in gingival margin gap formation among the placement techniques.

METHODS AND MATERIALS

Forty extracted human mandibular third molars free of caries or restorations were selected, cleaned and stored in a thymol solution at room temperature (74°F/23.3°C) until needed. All preparation and restorative procedures were performed in a temperature-controlled environment (23 ± 2°C) by a single operator who rehearsed material handling a number of times prior to beginning the study.

Restorative Procedures

The prepared teeth were individually fixed to a stone mold by embedding the roots in a quick-set vinyl polysiloxane impression material (Template, Clinician's Choice, New Milford, CT, USA), and a custom matrix band (AutoMatrix Retainerless Matrix System, Dentsply-Caulk, Milford, DE, USA) was adapted to the tooth. All the preparations were etched with 37.5% phosphoric acid (Gel Etchant, Kerr Corp, Orange, CA, USA) and a three-step etch-and-rinse adhesive system was applied (Optibond FL, Kerr Corp) according to the manufacturer's instructions and polymerized for 20 seconds. An LED light-curing unit (G-light, GC America, Alsip, IL, USA) was used for all procedures. An irradiance of at least 800 mW/cm² was ensured at all times by periodically monitoring the intensity of the light (Optilux radiometer, Kerr Corp). The 40 molars were randomly assigned to five study groups with eight specimens (N=8) per group as follows: Group 1: Preheated composite (PHC) to 130°F/54.4°C placed in 2-mm increments. Group 2: Preheated composite (PHC) to 155°F/68.3°C placed in 2-mm increments. Group 3: A 0.5- to 1-mm layer of flowable liner (FL) injected onto the gingival margin and immediately polymerized for 20 seconds; the remaining bulk of the preparation was filled with unheated resin in 2-mm increments. Group 4: A 0.5- to 1-mm layer of flowable liner (FL) injected onto the gingival margin and immediately followed by a 2-mmincrement of unheated resin over the uncured liner; the two layers were polymerized simultaneously. Group 5 (Control): Room temperature composite (RTC) placed in 2-mm increments. All the increments were polymerized for 40 seconds from the occlusal aspect, except for the flowable liner in Group 3, which was polymerized for 20 seconds prior to application of the unheated resin. Groups of five specimens at a time, each corresponding to a study group, were restored. Randomization in the restorative procedures was done to avoid bias relative to sequence of the tooth restoration.

Microhybrid resin (Filtek Z-250, 3M ESPE, St Paul, MN, USA) and flowable resin (Flow-It, Pentron Clinical Technologies, Wallingford, CT, USA), both in shade A2, were used for all the restorative procedures. For groups 1 and 2, a commercially available unit (Calset, AdDent Inc) was used to preheat the resin prior to its application into the preparation. Only two of the heater's available settings were used in the current study: 130°F/54.4°C and 155°F/68.3°C. Two units were used for this purpose; each was kept at the desired temperature during the restorative procedures. An assembly of a composite compule held by a carrying gun was placed in the heating unit and left in place for five minutes to reach the desired preset temperature. A previously completed pilot study found that it takes approximately 10 minutes for the heating unit to reach the desired preset temperature from the moment it is turned on. At this time, the assembly of a composite compule held by a carrying gun is placed into the unit. The same pilot study found that it takes approximately five minutes for the composite material inside the compule to reach the desired temperature from the moment the assembly is placed into the already heated unit. Since fluctuations in flowability are expected to occur rapidly upon cooling of the resin, the same pilot study was used to determine the cooling rate of the preheated resin. Figure 1 shows a graphic representation of the temperature change as a function of time for a resin preheated to 130°F/54.4°C. Resin preheated to 155°F/68.3°C showed the same trends. Based on the results of this pilot study, the time for placement of the first increment of resin for all the study groups was kept under 10 seconds.

View Figure

• Download as Powerpoint
• Download Figure

Replica Technique and SEM Evaluation

Prior to the impression procedure, the specimens were cleaned with 5.25% sodium hypochlorite solution (household bleach) for 30 seconds to remove contaminants and smear layer debris from the tooth-restoration interface. Immediately after cleaning, the specimens were rinsed under copious tap water and air-dried to remove excess moisture. An initial impression was made with a light body vinyl polysiloxane impression material (Extrude, Kerr Corp) to further remove any smear layer debris from the tooth-restoration interface; these impressions were discarded. A second impression with the same impression material followed immediately. These, if accurate upon visual inspection, were then stored in a hermetically-sealed plastic container at room temperature for 24 hours and poured with a low-viscosity epoxy resin (Epoxicure, Buehler Ltd, Lake Bluff, IL, USA) to produce positive replicas of the restorations. A period of at least 24 hours before pouring the impressions with epoxy resin has been recommended to avoid the incorporation of air bubbles in the epoxy replica, which may be caused by the liberation of gases during the setting reaction of the vinyl polysiloxane (personal communication with Dr Blunck). Epoxy replicas were left to set undisturbed for eight hours before gold sputter-coating for four minutes at a current of 30 milliampers.

Using eight to 10 images per replica, the entire gingival margin and at least 0.2 mm into the buccal and lingual wall margins were captured. The images were obtained with a field emission scanning electron microscope (AMRAY 1820D, Amray, Inc, Bedford, MA, USA) at a magnification of 200x. The sequence of images per replica was stitched together with computer software (Photoshop 7.0, Adobe Systems Inc, San Jose, CA, USA) (Figure 2). Cervical margin micromorphology was evaluated by a blinded evaluator who used image software (Image-Pro Plus version 5.1.0.20, BioImaging. Solutions Inc, San Diego, CA, USA) to trace the complete length of the evaluated margin and to quantitatively assess the quality of the margin. The evaluator assigned one of seven “margin quality” (MQ) scores to each of the margin sections according to the modified ordinal scoring criteria described by Blunck34 as indicators of marginal integrity. Table 1 provides definitions for each of the original seven outcome measures (MQ1-MQ7) used for assessment of the margins in-vitro, as well as a description of how the data were summarized for statistical purposes into the derived measures (Score 1–Score 4). The length of the margin that was involved (had a gap) was measured and expressed as a percentage of the total evaluated margin. Repeated scores in different sections of the margin were added to determine the total percentage relative to the complete length of the evaluated margin. After preliminary assignment of the “margin quality” scores in-vitro, the data were summarized for statistical purposes into four “margin scores.” These were further simplified into the two nominal categories of “Gap” and “No Gap” (Table 1). Results of the gingival margin gap formation were expressed as the percentage of open margins (gaps) relative to the complete length of the evaluated margin.

View Figure

• Download as Powerpoint
• Download Figure

Table 1 Gingival Margin Micromorphology Scoring Criteria

View Fullscreen

RESULTS

Intra-examiner Reliability Measurements

Intra-class correlation (ICC) coefficient results for the seven measured outcomes and five derived measures (Score 1–Score 4 and Gap) yielded a high level of agreement between duplicate measurements of percentage of gap formation relative to the complete evaluated length of the composite-dentin interface (ICC=0.956, p<0.0001) (Table 2).

Table 2 Intra-class Correlation Coefficients (ICC) for Both Margin Quality and Score Measurements

View Fullscreen

Although all ICC scores were significant at the conventional 0.05 level, reliability was high for MQ1, MQ2, MQ3 and MQ7 (ICC of 0.947, 0.969, 0.901 and 0.979, respectively), moderately high for MQ5 (0.729) and low for MQ4 and MQ6 (0.313 and 0.364, respectively). Reproducibility was also reflected in the five derived measures, yielding exceptionally good agreement for percentage of gap formation relative to the complete evaluated length of the margin (ICC=0.956, p<0.0001) (Table 2).

Margin Adaptation Scores

Main margin adaptation score results for the dentin-composite interface of all specimens are provided in Figure 3 and Table 3. No significant difference was evidenced in terms of the percentage of gap formation among groups defined by the placement technique (p=0.82). Although the results were not statistically significant, an outlier in group 2 shifted distribution of the mean, median and standard deviation measurements to a level higher than all the other groups. For all groups, at least one specimen exhibited perfect marginal adaptation (zero percent gaps) as reflected in the column labeled “Minimum” in Table 3. Overall, the mean gap percentage for the 40 margins evaluated was 6.3, with a median of 1.1 and a standard deviation of 14.8.

View Figure

• Download as Powerpoint
• Download Figure

Table 3 Percentage of Gaps in Total Length of Evaluated Margin (Average of Two Measurements)

View Fullscreen

The p-values for the Kruskal-Wallis test of the differences in distribution of a given measurement among the treatment groups is provided in Table 4. There was no evidence of treatment group differences for any of the direct (Margin Quality) or derived (Scores/Gap) measurements. The null hypothesis was supported by the results of the study.

Table 4 Differences in Distribution for Marginal Quality and Gap Formation Scores Among Treatment Groups

View Fullscreen

DISCUSSION

In the current study, different techniques with preheated resins and flowable liners were tested to determine their effect on gingival margin gap formation as compared to the conventional placement of room temperature composite. Although no significant differences in terms of gap formation were detected, group 5 (Control) had the smallest mean value, followed by group 1 (PHC, 130°F) and group 4 (FL cured simultaneously with first increment of composite).

With a lower modulus of elasticity and better wetability, flowable resins can be expected to serve as an intermediate elastic layer between the adhesive and overlaying resin, reducing contraction stresses and improving the seal of the restoration.23 Two different techniques involve the use of a flowable liner polymerized either prior to or simultaneously with the overlaying resin. Polymerizing the flowable liner prior to placement of the conventional hybrid resin yields a thicker layer, which may theoretically result in higher volumetric shrinkage.23 On the other hand, a thicker flexible layer may better help to relieve polymerization contraction stresses from the overlaying resin.35 By polymerizing both the flowable and hybrid resin simultaneously, the uncured flowable liner may be able to provide greater stress relief than the flowable liner that has been cured beforehand and already exists in a vitrified state. Also, by placing the stiffer hybrid resin onto the uncured flowable liner, most of the liner is expelled against the lateral walls, resulting in a thinner layer of flowable resin and perhaps less associated shrinkage and improved final adaptation. Excellent results have been reported in the literature with the use of this technique.1113 The results of this investigation showed no significant differences in mean gap formation between the two groups of flowable liners.

Although the present study found that preheated composites did not significantly reduced gap formation at the gingival margin, their use may yield benefits in other ways. By achieving a transient viscosity reduction comparable to that of flowable resins,29 an enhanced adaptation of the resin to all intricacies of the preparation may be obtained. At the same time, no restorative compromise was made, since a hybrid resin with high strength and high filler loading was used as the initial gingival increment. Also, an increased curing rate and degree of conversion may yield a final restoration with improved physical properties2629 and less unreacted monomer.28

Despite the findings of the current study, future research may produce different results, since the use of low viscosity materials is highly technique-sensitive and is associated with a significant learning curve. Low viscosity materials are difficult to manipulate easily, entrapping air upon removal of the syringe.1336 Flowable resins and preheated composites must be carefully applied to the preparation in only one direction, and removal must be done by gently wiping the compule tip against the preparation walls. In the current study, a speculation is that the operator's technique may have had an influence on the observed results.

Group 2 exhibited a single outlier specimen that shifted the entire distribution of the curve. Atypical of the overall gap occurrence, the outlier specimen displayed not only larger gaps (MQ5, MQ6 and MQ7), but also a higher percentage of gap formation. However, these values did not remain significant. Aspeculation is that this outlier could have been the product of human error. Polymerizing the resin while at its maximum temperature would increase the degree of conversion,28 thereby maximizing shrinkage stresses. Another speculation relates to the time when polymerization reaction was started relative to the cooling curve of the resin, which might have resulted in the combined shrinkage stress from both thermal contraction and polymerization shrinkage being potentially maximized in this specimen.

A material with low shrinkage, Filtek Z250,37 was used for this study, as current techniques are moving away from high shrinkage materials. By using a low shrinkage material, the potential confounding effect of resin composite shrinkage was eliminated from the equation. This may help explain why differences among study groups were not detected.

Care should be taken not to overemphasize the results of in-vitro tests, since a number of in-vivo variables that may play a role in the margin gap formation of Class II composite restorations cannot always be accurately replicated in-vitro. This in-vitro investigation was conducted in a temperature-controlled room (23 ± 2°C). However, preheated composites applied in-vivo can be expected to behave differently by retaining heat longer than those applied at room temperature. Although the gingival walls on dentin are a relative contraindication for restoration with composites, the preparation margins in the present study were located 1 mm below the CEJ in dentin to provide a greater adhesive challenge. Even under these circumstances, gap formation was minimal and no significant differences were detected among groups. The rather favorable results of this investigation have to be confirmed in-vivo. Although most approximal margins cannot be polished in-vivo, all margins were finished and polished to remove overhangs that would have otherwise interfered with an accurate assessment of the margins.

The impressions of the specimens were taken after 24 hours of storage in artificial saliva. No thermocycling to simulate the long-term challenge of the interface was performed, as the purpose of this investigation was to identify gaps that may have been caused by the immediate effects of polymerization shrinkage rather than those caused from degradation of the adhesive interface over time. No prolonged water storage was performed based on the same rationale. It is well accepted that composites shrink on curing, but perhaps it is less known that they swell upon water absorption.38 It is possible that water uptake during a longer storage period could compensate for initial polymerization shrinkage of the resin, closing microgaps or minimizing the occurrence of stresses that could induce adhesive failure.39

Although intra-class correlation scores were all significant at the conventional 0.05 level, reliability was high for MQ1-MQ3 and MQ7, moderate for MQ5 and poor for MQ4 and MQ6. A speculation is that the lack of descriptor with a numerical cutoff point of MQ4, MQ5 and MQ6 compared to the other categories made reproducibility of the duplicate measurements harder to achieve in this case. It seems that the replica technique used in this study can provide a reliable measurement for analysis of the margin gap formation, provided that adequate calibration of the examiner takes place prior to the beginning of the study.

CONCLUSIONS

Within the limitations of an in-vitro model, it can be concluded that gap formation at the gingival margin of Class II preparations was not improved relative to the control group by any of the placement methods tested in this study. A high degree of intra-examiner reliability was confirmed using duplicate measurements, indicating that the replica technique may provide an accurate method for the measurement of margin gap formation.