Specimen Preparation and Scanning

For this study (exempt approved by the institutional review board), 24 extracted human molars kept in 10% formalin acetate were selected in eight sets of three teeth that matched in shape and size. The sample size was based on previously published studies with similar methods; a sample size of eight has 95% confidence to detect a standard deviation difference of 0.7 between groups. The teeth were secured by their roots in stainless steel rings, pumiced to remove residual biofilms, and etched with 37% phosphoric acid solution to achieve a matte surface suitable for optical scanning. During the experiment, all teeth were kept hydrated in water, except during scanning.

Large mesio-occluso-distal slot-shaped cavities were prepared with a 245 carbide bur in a high-speed handpiece under copious amounts of water. During preparation, cavity width was checked with a digital caliper and cavity depth with a periodontal probe. After sectioning, later in the process, the cavity and cuspal dimensions were measured with image analysis software under a stereomicroscope (see the “Dye Penetration” section). The buccolingual width was 4.23 ± 0.13 mm and depth was 3.80 ± 0.32 mm (mean±standard deviation). Widths and depths were not statistically significant different among the three groups (analysis of variance [ANOVA], p=0.9373 and p=0.5798, respectively). The resulting cuspal widths were also not significantly different among buccal and lingual sides (ANOVA, p=0.5452) or between groups (ANOVA, p=0.6413). About 10 to 15 minutes after preparation, the prepared teeth, along with reference spheres embedded in the steel rings, were digitized with an optical scanner (Comet xS 3D Optical Digitizing System, Steinbichler Vision Systems, Neubeuern, Germany). Each point on the scanned surfaces was measured with 5-μm accuracy; the measured points were 60 μm apart (resolution). The scans of the prepared teeth were used as the baseline (Figure 1A).

View Figure

• Download as Powerpoint
• Download Figure

Restorative Procedures

Materials used in the study and their compositions are shown in Table 1. A total etch (etch and rinse) bonding technique was used. The cavities were etched with 34% phosphoric acid gel (Dentsply Caulk, Milford, DE, USA) for 15 seconds, rinsed with water, and blotted dry using Kimwipes (Kimberly-Clark, Roswell, GA, USA). Prime & Bond Elect (Dentsply Caulk) was applied and agitated for 20 seconds and gently air-dried for 5 seconds. The mesial and distal halves of the preparation were light-cured for 10 seconds each (VIP Junior, Bisco, Schaumburg, IL, USA). The light intensity was 600 mW/cm², measured with a radiometer (Model 100, Demetron Research Corp, Danbury, CT, USA) periodically during the experiments. The three matched teeth from each set were randomly assigned to be restored as follows: 1) two 2-mm horizontal increments using Esthet X HD (Dentsply Caulk) as a control; 2) two 2-mm horizontal increments using Surefil SDR Flow (Dentsply Caulk); and 3) one 4-mm bulk-filling using Surefil SDR Flow. The mesial and distal halves of each increment or bulk material were light-cured for 20 seconds each from the occlusal direction. The composite surfaces were wiped with an alcohol pad to remove the oxygen inhibition layer. After 20 minutes in water, the restored teeth were digitized with the optical scanner.

Table 1:  Materials Used in the Study^a

View Fullscreen

Cuspal Flexure Analysis

The baseline and restoration scans were accurately aligned by fitting the stainless steel spheres as unchanged references (Figure 1A) using Cumulus software (Copyright Regents of the University of Minnesota).¹⁸ The amount of cuspal flexure was calculated using a custom-developed CuspFlex software. Flexure was determined for the buccal and lingual surfaces from the area between the cusp ridges and the level of the pulpal floor (Figure 1C,D). The average deflection of each surface (buccal or lingual) was calculated from the notional volume change (integrated difference across the selected areas) divided by the selected surface area. The cuspal flexure was the sum of buccal and lingual cuspal deflections.

Dye Penetration

Dye penetration of the occlusal interface was assessed to verify the bond integrity. After the restored teeth had been scanned, restoration margins were finished with composite finishing burs (LogicSet, Axis Dental Corp, Coppell, TX, USA). Root apices were obstructed with utility wax, and the roots were painted with nail polish. The teeth were then placed in 0.5% basic fuchsin dye solution overnight (16 hours), embedded, and sectioned buccolingually into two halves. One half was used for the bond integrity assessment and the other half for the depth of cure evaluation. For the bond evaluation, the samples were sectioned buccolingually into 1-mm slices, yielding 4-6 slices per sample. Each slice was imaged using a stereomicroscope with a charge coupled device camera (SZX16 & UC30, Olympus, Tokyo, Japan). The distances of dye leakage at the occlusal margins along the buccal and lingual interfaces were measured by two independent evaluators using image analysis software (Stream Basic, Olympus Soft Imaging Solution GmbH, Münster, Germany) (Figure 2). If the distances measured by the two evaluators differed more than 10%, a consensus value was reached. The dye penetration distances were averaged for each tooth. In addition, the wall height of each cavity interface was measured from the occlusal margin to the pulpal floor.

View Figure

• Download as Powerpoint
• Download Figure

Depth of Cure Measurement

The other tooth half was used to assess the depth of cure by measuring hardness on the cross-sectional surface.^19,20 Thus, the hardness was measured one day after the restoration had been cured. The tooth halves were embedded in acrylic resin, and the cross-sectioned surfaces were serially polished with 400- and 600-grit silicon carbide paper followed by 1 μm and 0.05 μm alumina pastes. Vickers hardness (QV-1000 Micro Hardness Tester, Qualitest USA LC, Fort Lauderdale, FL, USA) was measured from indentations made across the restoration at 0.5 mm increments from the occlusal surface to the bottom of the restoration (pulpal floor) (Figure 3A). The indentation load was 200 g at 15 seconds dwell time.

View Figure

• Download as Powerpoint
• Download Figure

Statistical Analysis

The differences in cuspal flexure and dye penetration among groups as well as differences in hardness values among various depths within each group were statistically analyzed with ANOVA followed by Student-Newman-Keuls post-hoc tests (significance level 0.05).

Three Factors

The three selected factors could be considered most affected by a bulk fill technique. Of the three, the extent of cure is obvious because it determines the feasibility for bulk cure. Cure also affects polymerization shrinkage stresses. Stress is a local condition that is calculated by taking into account the properties of the restored tooth, transient properties of the composite, shapes of the tooth-restoration complex, and the state of bonding. In this study stresses were not determined because that required complex calculations.²¹ Instead, cuspal flexure was determined.^22-24 Cuspal flexure is caused by stresses, and thus inherently encompasses all attributes involved in the shrinkage stress, including the transient composite properties.^17,25 Marginal bond, the third factor, was examined because shrinkage stress is generated when composite polymerization contraction is constrained by the bonding to the tooth structure. The state of the restoration bonding must therefore be considered when assessing shrinkage stresses.

Cuspal Flexure

The cuspal flexure for the bulk-fill flowable composite used in this study did not show a significant difference between the incremental or bulk-filling techniques. The cuspal flexure with the bulk-fill flowable base composite restorations was significantly lower than the conventional nonflowable restorative composite that was used as a reference. This suggests that the new bulk-fill flowable base composite generated lower shrinkage stresses in the tooth than those of a conventional, well-accepted restorative composite system.^5,26 It also suggests that for controlling shrinkage stresses within a clinical context (ie, restoration of a tooth), the difference between incremental or bulk-fill techniques is probably marginal.^13,20

Extent of Cure

Interpretation of cuspal flexure results requires caution. As noted previously, differences in cure also cause differences in flexure. For example, inadequate cure leads to lower elastic modulus and less polymerization contraction, which results in less cuspal flexure and hence lower shrinkage stresses in a restored tooth. The trade-off between less cure for lower shrinkage stress is undesirable, and therefore obtaining lower cuspal flexure due to less cure is not necessarily a favorable outcome. To ensure that the cure was adequate or comparable, the extent of cure was determined throughout the depth of the restorations using hardness measurements.

The hardness measurements of the bulk-fill flowable base composite indicated that a homogeneous cure through the entire 4-mm depth of the composite was achieved with both the incremental and bulk cure. If the cure at the surface is assumed to have been adequate, the cure throughout the 4-mm-deep restorations must have been adequate also, ensuring optimal mechanical properties, biocompatibility, and chemical stability regardless of whether incremental or bulk curing techniques were used. Unlike the bulk-fill composite, the conventional composite showed a decrease in hardness (and thus extent of cure) at the bottom of each 2-mm increment, demonstrating that composites that are not designed for bulk curing should be cured in increments to avoid compromising their mechanical and biophysical properties. It is important to mention that no direct comparisons were made about the degree of cure of the bulk-fill and conventional composites using the hardness data because hardness is a material-specific property that also depends on composition (eg, filler distribution, resin type).¹⁹ The lower hardness values of the bulk-fill composite, consistent with previous studies,²⁶ were likely due to the lower filler loading of this flowable restorative.⁴

Marginal Bond

The third factor that was examined was the state of the bond at the occlusal margins. Clinically, good bonding between composite and tooth structure is critical for restoring the structural integrity of a restored tooth²⁷ and for sealing the inside of the tooth structure from bacteria and their nourishments. A restoration bond is challenged by polymerization shrinkage as well as stresses caused by thermal and mechanical loading. Resisting those challenges requires sufficient bond strength. Although strength or longevity of the bond are important for the performance of a restoration, it was not the objective of this study. For the present study design, intact bonding needed to be verified to ensure that the measured cuspal flexure was due to polymerization shrinkage and not modified by bond failure. Debonding would relieve shrinkage stresses and reduce the cuspal flexure, wrongly suggesting lower shrinkage stresses. Clinically, reduction in shrinkage stress due to debonding is also undesirable. Penetration of the fuchsin dye revealed that the bond remained largely intact for all groups, thus indicating comparable quality of the bonds among the three groups. Therefore, comparison of the cuspal flexure among the three groups was justified and could be used to draw conclusions about the effect of bulk filling on the stresses in restored teeth.

Interpretation of Cuspal Flexure

Various studies have confirmed the ability of bulk-fill composites to cure much deeper than conventional composites,^4-6 and in vitro cuspal flexure and marginal gap studies have not found compelling evidence for avoiding bulk-fill techniques.^7-10 Cuspal flexure measurements are a convenient technique for comparing the state of stresses in teeth restored with different composites or filling techniques, provided that the tooth properties, shapes, and bonding conditions are comparable.²¹ However, the performance of a restoration is not determined by shrinkage stress. Stress cannot be extrapolated to performance without the context of strength. A weak interface will fail at low stress values, and thus a low shrinkage stress state can still have little relevance for clinical performance. Moreover, polymerization shrinkage stress is a temporary condition because it is likely to be compensated by hygroscopic expansion and stress relaxation.^28-30 If a bond survives the initial shrinkage stresses, other properties may turn out to be more important for the life expectancy of a restoration.

Based on these considerations and the results of this study, it can be postulated that the most important aspect of a composite restoration technique is maintaining acceptable cure and intact interfaces. The choice between an incremental or bulk-filling technique should be guided by those requirements, rather than by shrinkage stress considerations. From cross-sectioned samples, we observed that the low-viscosity flowable composite provided better adaptation to the cavity walls, floor, and between composite layers in the incremental fillings than the conventional composite. The lower microhardness of the flowable composite compared with the conventional composite supports the recommendation that under clinical conditions this flowable material should be used as a base/liner or dentin replacement.³¹ The potential to attain adequate polymerization in deep layers offered by the new generation of bulk-fill composites is an advantage in critical areas where curing can be compromised, such as in the gingival portion of a proximal restoration where a bulk-fill flowable base composite like Surefil SDR Flow is recommended under a universal or posterior composite.