Characterization and Comparative Analysis of Voids in Class II Composite Resin Restorations by Optical Coherence Tomography
This study aimed to characterize and analyze the number of voids and the percentage of void volume within and between the layers of class II composite restorations made using the bulk fill technique or the incremental technique by optical coherence tomography (OCT). Class II cavities (4×4×2 mm) were prepared in 48 human third molars (n=24 restorations per group, two class II cavities per tooth). Teeth were divided into four groups and restored as follows: group 1 (FOB), bulk filled in a single increment using Filtek One Bulk Fill (3M Oral Care); group 2 (FXT), incrementally filled using four oblique layers of Filtek Z350 XT (3M Oral Care); group 3 (FBF+FXT), bulk filled in a single increment using Filtek Bulk Fill Flowable Restorative (3M Oral Care) covered with two oblique layers of Filtek Z350 XT (3M Oral Care), and group 4 (FF+FXT), incrementally filled using Filtek Z350 XT Flow (3M Oral Care) covered with two oblique layers of Filtek Z350 XT (3M Oral Care). After the restorative procedure, specimens were immersed into distilled water and stored in a hot-air oven at 37°C. Forty-eight hours later, thermal cycling was conducted (5000 cycles, 5°C to 55°C). Afterward, OCT was used to detect the existence of voids and to calculate the number of voids and percentage of voids volume within each restoration. Data were submitted to chi-square and Kruskal-Wallis tests (α=0.05). Comparisons were made using the Dunn method. Voids were detected in all groups, ranging from 0.000002 (FBF+FXT and FF+FXT) to 0.32 mm3 (FBF+FXT). FF + FXT presented voids in all of the restorations and had a significantly higher number of voids per restoration when compared to the other groups (p<0.05), but restorations with the presence of voids were significantly higher only when compared to FXT (p<0.05). FBF + FXT presented a significantly higher percentage of voids volume than that of FXT (p<0.05). When comparing restorations made using high-viscosity resin-based composites (FOB and FXT), no significant differences regarding number of voids or percentage of voids volume were detected (p≥0.05). The use of flowable resin-based composites can result in an increased number of voids and percentage of voids volume in restorations, and this appears to be more related to voids present inside the syringe of the material than to the use of incremental or bulk fill restorative techniques.SUMMARY
Purpose:
Methods and Materials:
Results:
Conclusions:
INTRODUCTION
Resin-based composites (RBCs) are materials indicated for use in different types of dental treatments in all areas of the mouth, and due to their versatility, their use is growing continuously.1 Since their first appearance, several investigations have been developed to achieve improvements concerning their clinical behavior,1 but there are still controversies regarding which is the best restorative technique that would minimize the incorporation of voids during the restorative procedures in order to avoid reducing the mechanical properties of the final restoration.2-4
The incremental restorative technique, which consists of inserting RBC increments up to 2 mm thick,5 has been commonly executed to ensure a restoration with an appropriate degree of polymerization with reduced cytotoxicity and improved physical properties.6 However, it may also present some disadvantages related to the formation of voids and the absence of union between the layers as well as the difficulties associated with restoring large, deep, and difficult-to-access cavities, which may increase the time needed for the restorative procedure.7 Therefore, a new generation of these materials has been developed, the so-called bulk fill RBCs, designed to restore medium to deep cavities in posterior teeth with large increments of up to 4- to 5-mm depth.8 Manufacturers claim that bulk fill RBCs minimize the incorporation of voids,9 and studies have reported the same pattern, declaring a lower presence of voids within the mass of the RBC restoration with the use of the bulk fill restoration technique due to decreased handling during the restorative procedure,3,10,11 especially in deeper cavities.12 In contrast, a higher presence of porosity in greater thicknesses of the material has also been reported,13 as has a lower incidence of gaps in a larger amount of small increments,2 supporting the use of the incremental technique.
Voids within a restoration, also referred to as porosities or bubbles, are caused by air entrapment and can be incorporated into RBCs during their manipulation when performing a restorative procedure or even while being manufactured. Such voids can be located as follows: on the restoration surface, leading to a greater retention of organic residue and causing an appearance similar to marginal staining; in the restoration's bulk, negatively affecting its aesthetic and mechanical properties;14 or at the tooth-restoration interface, forming gaps at the restoration's margin.15
Diverse techniques have been performed for the evaluation of porosity, including conventional methods, such as microscopy on sectioned samples,16 or radiographic analyses of specimens.11 However, in an attempt to avoid destruction and improve the resolution of the internal structure images, new technologies have been introduced.17 Optical coherence tomography (OCT) is a nondestructive18 technique that allows noncontact cross-sectional and optical imaging.19 Since its introduction in dentistry in 1998,20 OCT has expanded its application in laboratory research as well as clinical practice, enabling the characterization of oral structures and restorative materials.20-23 In this way, OCT continues to be used in research for the characterization of composites, aiming at the evaluation of internal and marginal adaptation24,25 and, furthermore, the detection of defects in restorations.21,25 In fact, OCT analysis has also been presented as a valid method for the characterization of voids in RBCs by previous work that compared OCT results with data obtained by microcomputed tomography,25 another technique recognized for the evaluation of porosities in restorations.6
It is known that the presence of voids within restorations negatively affects the flexural strength of the material and determines a lower resistance to fatigue and wear26,27 due to their contribution to the initiation and propagation of cracks, leading to failure of the restoration as it is subjected to mechanical and external loads when functioning.28 However, the advantages of using a certain restorative technique, improving the quality of a restoration by reducing the presence of voids, are still unclear and must be verified.
Thus, the purpose of this in vitro study was to characterize and analyze the presence of voids inside class II restorations made using the bulk fill technique compared with restorations made using the incremental technique through OCT, a nondestructive technique. The null hypotheses of this study were 1) that there is no significant difference in the number of voids among restorations made using incremental technique compared with those made with bulk fill technique and 2) that there is no significant difference in the percentage of voids volume among restorations made using incremental technique compared with those made with bulk fill technique.
METHODS AND MATERIALS
Specimen Selection and Preparation
Forty-eight freshly extracted, sound, caries-free human third molars were selected, cleaned, and stored at 37°C. In order to obtain a regular enamel surface, proximal, mesial, and distal surfaces of the third molars were polished using a polishing machine (023-011263, EcoMET, Buehler, Lake Bluff, IL, USA). Teeth were divided into four groups (n=24 restorations per group, two class II cavities per tooth, where mesial and distal restorations were observed separately), according to the RBC and restorative technique used (Table 1). Prior to any preparation, high-density vinyl polysiloxane silicone (VPS) was manufactured to serve as a matrix for each tooth, which aided the material placement in the restorative procedure. A VPS matrix was used since by positioning the tooth inside a silicone mold, it was possible to standardize the restorative procedure, especially the composite's cervical adaptation. After that, teeth were removed from their matrices, and two slot class II preparations were made at each proximal surface with a 4-mm depth, 4-mm length, and 2-mm width (Fig. 1). Measurements of the preparations were checked with a digital caliper (500-196-30, Mitutoyo Corp, Kanagawa, Japan).
Citation: Operative Dentistry 45, 1; 10.2341/18-290-L
Restorative Procedures
Prepared teeth were repositioned into the VPS matrix for the restorative procedure, which was performed by a single operator in order to standardize this procedure. All restorative procedures were performed by the same operator as well. The self-etch adhesive system ClearFill SE Bond (Kuraray, Tokyo, Japan) was applied according to manufacturer's instructions. The restorative materials and techniques were as follows: group 1 (FOB), Filtek One Bulk Fill (3M Oral Care, A2 shade), bulk filled using a single 4-mm increment; group 2 (FXT), Filtek Z350 XT (3M Oral Care, A2 Body shade), incrementally filled using four oblique 2-mm layers; group 3 (FBF+FXT), Filtek Bulk Fill Flow (3M Oral Care, A2 shade), bulk filled using a 2-mm single layer covered with two 2-mm oblique layers of Filtek Z350 XT; and group 4 (FF+FXT), Filtek Z350 XT Flow (3M Oral Care, A2 shade), incrementally filled using a 2-mm layer, also covered with two 2-mm oblique layers of Filtek Z350 XT. Each RBC increment was photoactivated for 20 seconds with the light-curing unit Radii-cal (SDI, Victoria, Australia), which presents a radiant exitance of 1200 mW/cm2. After restoration, in order to ensure an adequate polymerization of the RBCs, teeth were removed from the VPS matrix, and afterward a polyester matrix was placed on the occlusal and proximal surfaces of restorations; each occlusal, mesial, and distal surface was photopolymerized for 20 seconds. Once the restorative procedure was completed, specimens were immersed into distilled water and stored in a hot-air oven at 37°C (Fanem 502, Fanem, São Paulo, Brazil). Forty-eight hours later, thermal cycling was conducted using 5000 cycles (5°C to 55°C, 30-second dwell time), equivalent to approximately six months of thermal alterations in the oral cavity,28 using a thermal cycling machine (Nova Etica 521/4D, Nova Etica, São Paulo, Brazil) to simulate intraoral thermal alterations.29
OCT Analysis and Definition of Voids
After thermal cycling, OCT was performed (OCT930 SR, Thorlabs Inc, Newton, NJ, USA). The equipment's light source emitted a central wavelength of 930 nm and 2 mW of optical power with a maximum image depth in resin composite of about 0.7 mm with axial resolution of 0.4 and lateral resolution 6.0 μm. In order to compensate for the limited penetration and to make the evaluation of all the RBC increments possible, scanning of a representative portion of the restoration was made with teeth positioned with the proximal surface facing up. The SR Scan program (Thorlabs Inc) was used to perform the tomographic scan, with parameters of 2000 A-scan (columns) for a width of 2.5 mm (range). Based on the low-coherence light scattered from the sample micron/level resolution subsurface, cross-sectional images of 2000 × 512 pixels, corresponding to 2.5 × 0.7 mm, were obtained by analyzing the interference pattern.19 To standardize the evaluation, specimens were always scanned starting from the cervical to the coronal portion of the restoration and double-checked without considering either the restoration interface or the adhesive layer. Voids in this study were defined as spherical or ellipsoidal defects, and two types of voids were observed: 1) voids inside the restoration, as a defect with a surrounding light margin, which is produced when the light transverses the air trapped inside the void, reflecting a portion of light and producing a higher signal intensity,25 or 2) voids on the proximal surfaces, as a semispherical or semispheroidal irregularity on the top of the image, which corresponded to air trapped between the silicone matrix and the composite during the restorative procedure (Fig. 2). When a void was visually detected, a scan was made around it to define the largest visible diameter. The volume of each one was also calculated by establishing a model in which it was assumed that two types of voids could be observed: those with a similar height and width, close to a spherical shape, and those with one of the two dimensions of greater size (height or width), corresponding to the ellipsoidal format. Therefore, the volume of the spherical voids was calculated using the formula to obtain the volume of the sphere (4/3πr3) and the volume of the ellipsoidal voids with the formula to obtain the volume of the ellipsoid (4/3πa2b). Moreover, total volume of the voids was calculated by restoration, and with these values, the percentage of voids volume of the restoration was determined, considering 11.2 mm3 as the total restoration volume. This volume corresponds to the preparation's dimensions (4-mm length and 4-mm height) and the penetration depth of the OCT (0.7 mm). Images were formatted keeping the original pixel distance (512) using ImageJ software (1.8.0_172, National Institutes of Health, Bethesda, MD, USA). Then measurements used to obtain the dimension values for volume calculation of voids were taken using the same software. To calculate the measurements of the X-axis, the measurement was performed directly on the formatted image, and on the Y-axis, image distortion was considered by the material's refractive index. The value was corrected, dividing it by the refractive index of the composite resin: 1.6.30
Syringe Samples and Characterization of Restorative Materials
One sample of each resin material used in the different restorative techniques was extracted directly from their manufactured syringe, avoiding manipulation. High-viscosity materials were removed from the syringe with a clean cut made with a number 10 surgical blade (Solidor, Lamedid, Suzhou, China), and flowable materials were dispensed on a glass slab. Then syringe samples were light cured according to the manufacturer's instructions, followed by an individual scan by OCT. This step was performed to evaluate the presence of any flaws in the material when directly extracted from the syringes with no further manipulation.
Statistical Analysis
After a descriptive analysis of the number, percentage of voids volume, and shape and location of voids per restoration, data regarding the presence or absence of voids inside each restoration per group was analyzed by the chi-square test to verify if there were differences between restorative techniques (α=0.05). The number of voids and the percentage of voids volume per restoration and per group were organized into tables and analyzed with the Liliefors test for normality. Due to a nonnormal data distribution, the Kruskal-Wallis test and comparisons with the Dunn method were applied (α=0.05).
RESULTS
OCT analysis detected the presence of voids in all groups. The percentage of restorations with voids and the total number of voids per group are presented in Table 2. FF + FXT had the highest number of voids, registering a total of 176. The shape and location of the voids are described in Table 3. All groups presented more spherical voids than ellipsoidal voids, with the exception of FXT, which presented the same number of spherical and ellipsoidal voids. Moreover, the number of voids inside restorations was greater than that of voids located on the surface of proximal surfaces. Voids with diameters of 0.01 mm (FF+FX and FBF+FXT) to 1.07 mm (FBF+FXT) were observed. Individual volume observed among the voids remained within 0.000002 to 0.32 mm3 (Table 4). The percentage of voids volume determined ranged from 0.001% (FOB and FXT) to a maximum of 2.9% (FBF+FXT) of the total restoration.
FF + FXT presented voids within all restorations (100%), being significantly greater than FXT (p<0.0149) and presenting no significant differences when compared with the other groups (p≥0.05). The percentage of restorations with porosities was at least 75% per group, and considering all 96 restorations evaluated, voids were detected in 84.4% of them.
The Kruskal-Wallis test detected significant differences among the groups regarding number of voids (p<0.0001) and percentage of voids volume (p=0.0158). The Dunn multiple comparison test determined that considering the number of voids, FF + FXT was significantly higher than the other three groups (p<0.05), while considering the percentage of voids volume, FBF + FXT was significantly higher only when compared to FXT (p<0.05). There were no significant differences when comparing with the other groups (p≥0.05).
The images of syringe samples (Fig. 3) showed that high-viscosity RBCs FOB and FXT were homogeneous and void free. In contrast, nonmanipulated samples of flowable RBCs, FBF, and FF presented defects in their structure. In FBF, a large defect could be observed in the material's internal region. In addition, there were several smaller defects within the FF material, presenting characteristic void images inside.
Citation: Operative Dentistry 45, 1; 10.2341/18-290-L
DISCUSSION
The first null hypothesis was rejected since FF + FXT number of voids was significantly higher than the other three groups. The second null hypothesis was also rejected since the FBF + FXT percentage of voids volume was higher than FXT.
All groups presented voids within restorations in the OCT evaluation. The high presence of voids within restorations (84.4%) has already been reported, describing values of presence of voids ranging from 86.4% to 100% of examined samples scanned by an electron microscope.2 Regarding the shape of the voids, although a different study measured only spherical voids,25 two types of shapes were observed and defined in this study—ellipsoidal and spherical—in order to get a more accurate volume measure. Some studies associate an ellipsoidal shape with the voids located within increments of composite and a spherical shape with those located inside an increment of the material.2 This study could not verify this characterization since ellipsoidal voids were observed in restorations made with both incremental and bulk fill techniques for both high viscosity and flowable materials. In addition, voids inside the restoration were more frequent compared with voids located on the surface of the proximal surface.
The percentage of voids volume varied from 0.001% (FOB and FXT) to a maximum of 2.9% (FBF+FXT), in accordance with results obtained by a previous study in which values ranging from 0.1% to 2% were registered.25
The analysis of number of voids did not confirm that a greater number of voids is responsible for the greater percentage of voids volume in the restorations. However, those discrepancies can be explained by comparing the difference in the size of voids detected in each group since the FBF + FXT presented voids with an average size of 0.01 mm3, much larger than those detected in FF + FXT, with an average size of 0.0005 mm3. This discrepancy demonstrates the importance of determining the volume of the voids and not only their quantity when evaluating the porosity of the restorations since large discrepancies in size could lead to misinterpretation of the results. Moreover, clinically, a large void is mechanically worse for a restoration than a small void even if the actual number of voids is the same, differing only in volume.
The presence of voids in restorations may be related to the operative technique, the operator's ability, and the consistency of the material.31,32 Since in the present study all the restorations were made by the same operator, the possibility of incorporating voids was equal for all groups, thus avoiding influencing, negatively or positively, the results of any group. However, regarding the operative technique factor, it was observed that the groups that presented the highest quantity of voids and percentage of voids volume, respectively, belonged to different types of composites (incremental technique with conventional flowable RBC [FF+FXT] and bulk technique fill with bulk fill flowable RBC [FBF+FXT]), although the type of manipulation was the same: 2 mm of a flowable RBC followed by incremental layering of a conventional high-viscosity RBC. Both techniques were performed with the same amount and size of increments since cavity preparation depth did not allow for greater thickness of bulk fill flowable composite, as it had to be covered with two oblique layers of high-viscosity RBC.33
The results of this study disagree with the premise that the bulk fill technique allows a decreased void incorporation within restorations, attributed to the possibility of inserting the material into cavities using a single portion. Nevertheless, they agree with the results of studies that showed that a similar percentage of voids volume was detected when incremental and bulk fill techniques were evaluated.34 Although new insertion techniques for bulk fill RBCs, such as sonic insertion of the material, continue to be developed, the previously mentioned technique has not been confirmed to be advantageous, as a recent study performed with microcomputed tomography showed that restorations performed with the sonicated technique presented greater void formation when compared with the conventional technique.4
Considering the viscosity of the materials used in the present study, it can be observed that restorations made with lower-viscosity RBCs presented a higher number of voids and a higher percentage of voids volume (FF+FXT and FBF+FXT) than restorations performed with high-viscosity RBCs. Studies have also reported a higher incidence of voids and percentage of voids volume attributed to the intrinsic porosity of low-viscosity RBCs that cannot be controlled or modified by the operator because it is limited by their stipulated restorative protocols.35
In order to help understand the porosity results obtained in this study, syringe samples of the resin materials used in the different restorative techniques were extracted directly from their syringes, avoiding manipulation and photoactivation. After being individually analyzed using OCT, it was possible to observe that the high-viscosity RBCs FOB and FXT were homogeneous and void free. In contrast, the flowable RBCs presented defects in their structure. In the FBF, a defect can be observed in the internal region of the material; although it does not have the defined characteristics of the definition of a void, it has the potential to cause a defect of rounded shape and of great size once manipulated for its use as a restorative material. In addition, there are several smaller defects in the structure of the FF material, presenting characteristic void images inside. This fact might be related to how the flowable materials are disposed inside their syringes or to how they leave the syringe in the needles, which is directly related to the manufacturer and not to the operator and can jeopardize the resin composite restoration mass by incorporating voids inside the material.
Although conclusions cannot be made from the analysis of only one sample of material, the images suggest a correlation with the results obtained in this study. Since the FF + FXT group presented the highest number of voids and the FBF + FXT group presented the worst behavior when analyzing the percentage of voids volume, it can be assumed from the voids detected by the OCT that for low-viscosity materials, characteristics of the restorative material are more determinant for porosity than the restorative technique.36 The explanation for such an assumption can be attributed to the manufacturing process of these resins, which may generate porosities intrinsic to the material. As the presence of voids in a restoration can result in less fatigue resistance, reduction of flexural strength, and surface irregularity,26,37 it becomes even more critical in the case of flowable bulk fill RBCs since the reduced time required to fill a large cavity using the bulk fill restorative technique demands larger thickness of the material. In contrast, flowable conventional RBCs are indicated for preventive restorations and liners or to restore small lesions due to their reduced mechanical properties.38
High-viscosity materials were homogeneous in the syringe samples, suggesting that the technique, whether incremental or bulk fill, was responsible for the incorporation of voids within restorations. Further studies need to be developed to determine the intrinsic porosity of composite resin materials inside their syringes. Likewise, RBCs must be used following the technique recommended by their manufacturers to avoid failures in the restorative technique, with special caution regarding the use of flowable resins given their potential to present greater initial porosities.
CONCLUSIONS
Within the limitations of this laboratory study, it is possible to conclude that the use of flowable RBCs can lead to an increased number and percentage of voids volume in restorations. This appears to be more related to voids present inside the syringe of the material than to the incremental or bulk fill restorative technique performed.
Tooth cavities. Two slot class II preparations made at each proximal surface, with 4-mm depth, 4-mm length, and 2-mm width.
Figure 2 Void detection. (A): OCT image representing what was considered a void inside the restoration. (B): OCT image representing what was considered a void on the proximal surface of the restoration. (C): Image of a restoration with a superficial void on the proximal face. OCT, optical coherence tomography.
OCT images of syringe samples representing each restorative material used in the study. FOB- and FXT-labeled figures show homogeneous and void-free characteristics. By comparing figures labeled FBF and FF, a higher number of small voids can be observed in FF, while a larger, single void can be seen in FBF, representing a higher void volume but a small number. OCT, optical coherence tomography; FOB, Filtek One Bulk Fill; FXT, Filtek Z350XT; FBF, Filtek Bulk Fill Flowable; FF, Filtek Z350XT Flow.
Contributor Notes