Restoration of teeth in the posterior area of the oral cavity is considered a challenge for dentists today. It is necessary to fulfill the increasing cosmetic demands and meet the expectations of patients while ensuring the functional aspect and maintaining dental tissues. Restorations are expected to provide excellent long-term durability and prognosis. The restoration options of class I and II cavities include direct composite restorations, indirect composite inlays and onlays, and ceramic partial coverage restorations.
When compared to posterior composite restorations especially in larger cavities/preparations, ceramic partial coverage restorations are considered the best option both aesthetically and in terms of durability. These restorations are considered excellent alternatives for complete coverage restorations that cause an extensive loss in dental hard tissues. The amount of the preparation of a complete crown is 67.5%-75.6% of the whole dental tissue, while the partial preparations remove merely 5.5%-27% (Edelhoff & Sorensen 2002). This can influence the longevity of the restoration and the tooth at the same time (Christensen, 2008).
Thanks to the development of the materials and strategies of adhesion, partial coverage ceramic restorations have become a clinical option. Different systems of ceramics have been used for the fabrication of these restorations. The most important one is the system of heat-pressed lithium disilicate glass-ceramic, as this ceramic has distinctive optical and mechanical properties and a high fracture strength (Stappert et al. 2007). Due to its microstructure, it can be acid-etched. Consequently, the bond strength with resin cements is high (Rocca et al. 2015). High survival rates have also been recorded for partial coverage restorations made of lithium disilicate glass-ceramic (Sasse et al. 2015; Al-Akhali et al. 2017).
The success of dentin bonding with resin materials is the cornerstone in the success of onlays, due to the fact that the final strength of the tooth-restoration complex mainly depends on the adhesion procedures, especially with the high possibility of dentin exposures during the preparation of such restorations. In this context, the management of dental tissues after the completion of the preparation is considered an important matter in increasing the chances of successful treatment. The immediate dentin sealing method was proposed to improve the bonding in the early 1990s by Dr. Shigehisa Inokoshi, the Godfather of contemporary “Immediate Dentin Sealing” and “Cavity Design Optimization” techniques using resin materials. He published an article in Japanese, in which he described the dentin sealing and pulp protection under the name of “Resin Coating” technique (Inokoshi, 1992). Pashley in 1992 had approached the subject as well; he and his team proposed to seal dentin by only using an adhesive (Pashley et al. 1992), while Inokoshi used a flowable with an adhesive. The technique was further developed by establishing a resin sealing layer on the surface of dentin by applying a bonding system plus a low-viscosity composite resin. It is believed that this base layer of the composite resin insulates and protects the underlying hybrid layer, and consequently this should help to preserve the dentin seal (Duarte et al. 2006).
With immediate dentin sealing, it has been observed that there is a better relief for the patient with less bacterial microleakage and dental sensitivity during the temporization phase and after cementation procedure (Hu & Zhu 2010; Qanungo et al. 2016). Therefore, the technique of immediate dentin sealing was proposed when there are widely exposed dentin areas, like the preparations for indirect restorations, such as inlays, onlays and full crowns. Moreover, some investigations also observed that the bond strength is much better with freshly cut dentin than with the dentin contaminated with temporary cement residues (Lee & Park 2009). The immediate dentin sealing technique allows for pre-polymerization of the dentin bonding agent, which contributes to improving durability of the bonding layer (Magne et al. 2005; Okuda et al. 2007; Magne, 2014; Gresnigt et al. 2016).
Clinically, the most common reason for ceramic partial coverage restorations failure is restoration fracture, followed by failure of endodontic treatments and secondary caries (Morimoto et al. 2016). In a study about the clinical success of inlays and onlays made of glass-ceramic over a 12-year period, bulk fracture was the main reason of failure, due to the fragility of the ceramic structure (Frankenberger et al. 2008). Hence, fracture strength is regarded a crucial issue in the ceramic partial restorations and a qualitative criterion for evaluating the durability of prosthesis. However, there appears to be insufficient data on the effect of immediate dentin sealing technique on onlay fracture strength.
The aim of this study was to investigate the fracture strength of lithium disilicate ceramic onlays after applying the immediate dentin sealing technique and comparing it to delayed dentin sealing. The null hypothesis was that the immediate dentin sealing technique would not affect the fracture strength of the ceramic onlays under the conditions of this study.
Materials and methods
The sample of the study consisted of 20 maxillary premolars that were restored with onlays. The onlays were fabricated using heat-pressed lithium disilicate glass-ceramic. Twenty premolars freshly extracted for orthodontic reasons were collected. The teeth were sound, free of lesions and cracks and had close dimensions (ISO/TS 11405:2003). They were cleaned of any tissue remnants or calculous residues using dental curettes, hand scalers and air scaler tips. After that, they were preserved in an aqueous solution of 0.05% of thymol at room temperature until the start of the study. Then, they were stored in an aqueous solution of 0.9% of sodium chloride. The premolars were randomly divided according to the dentin sealing technique into two stratified groups (n = 10, each): immediate dentin sealing (IDS) and delayed dentin sealing (DDS).
A putty condensation silicone mold was used to make cuboid acrylic bases (18×18×13 mm3), where the teeth roots were immersed in a transparent self-cured acrylic resin (Orthocryl®, Dentaurum, Ispringen, Germany) perpendicular to the horizontal surface of the mold and 1 mm below the cemento-enamel junction (CEJ). A silicone guide was made for each premolar to copy the anatomical shape in order to check later the thickness of the preparation. Furthermore, twenty-perforated custom trays were made of self-cured acrylic resin (RESPAL® NF, SPD, Mulazzano (LO), Italy) for making the final impression.
20 standardized mesio-occlusal-distal (MOD) preparations with reduction of the palatal cusp were performed using the set of burs for ceramic inlays and partial crowns (Expert Set 4562; Komet Dental, Gebr. Brasseler GmbH&Co. KG, Lemgo, Germany). The dimensions of the preparations were standardized using gingival probe, iwanson caliper and split silicone guide (Fig. 1). The walls after the preparation were smooth and relatively flat, the gingival-occlusal divergence of the axial walls was 6 to 10 degrees per wall and all lines and point angles, internal and external were rounded and the cavosurface margins were 90 degrees. The palatal cusp was reduced by 2 mm and beveled in an angle of 45 degrees along with the longitudinal axis of the premolar with a depth of 2 mm. The palatal margin was in the form of a rounded shoulder with a depth of 1 mm and placed 2 mm away from the tip of the palatal cusp. The axial wall of the resulting shoulder was reduced by 1.5 mm and was occlusally convergent by 10 degrees and had the same path of draw as the main portion of the preparation (Fig. 1).
Immediate dentin sealing
The IDS technique was applied upon completion of the preparation of the first group and before temporization and making the final impression. Acid etching of the freshly cut dentin was performed for 15 seconds by phosphoric acid (H3PO4) (UNI-ETCH® (32%), BISCO, Inc., IL, USA) with the extension of the application 1 mm towards enamel margins. Then, it was rinsed thoroughly and dried, leaving the preparation visibly moist, and a coat of dual-cured universal dental adhesive system (ALL-BOND 3®, BISCO, Inc., IL, USA) was applied to the preparation and agitated for total of 10 seconds and air-dried. It was then light-cured for 10 seconds (LEDition, Ivoclar Vivadent AG, Schaan, Liechtenstein). After that, a thin coat of light-cured low viscosity resin (ALL-BOND 3® RESIN, BISCO) was applied to the preparation and light-cured for 10 seconds. Subsequently, a thin layer of flowable composite resin (Te-Econom Flow, Ivoclar Vivadent AG) was applied and light-cured for 20 seconds, the unpolymerized superficial oxygen-inhibited layer (OIL) was wiped using an alcohol-moistened cotton pellet. Light polymerization was repeated for 20 seconds after covering the sealed surface with glycerine jelly (‘‘air blocking’’). The preparation was finally redefined and the enamel margins were re-beveled. They were finished and polished using fine diamond burs and rubber polishing tips.
The impression was immediately made in the DDS group after the completion of preparation, whereas in the IDS group it was made after the completion of the immediate dentin sealing technique. Twenty impressions were made of the 20 prepared premolars using perforated custom trays, by following the putty-wash one-step technique, using a condensation silicone putty and light body (zetaplus/oranwash L/indurent gel, Zhermack SpA, Italy).
A coat of provisional separating material (PRO-V COAT®, BISCO, Inc., IL, USA) was applied to the entire preparation and air-dried for 15 seconds. The temporary restoration was created using a light-cured provisional packable composite (PRO-V FILL®, BISCO, Inc.) which was light-cured for 10 seconds. Then, a layer of light-cured low viscosity liquid polish (BisCoverTM LV, BISCO, Inc.) was applied over the entire surface of the provisional restoration and light-cured for 30 seconds.
The premolars were stored in an atmosphere of 100% relative humidity at a temperature of 37 °C, where the sample was placed in dry heat sterilization unit (TS TAU STERIL 2000 Automatic, Fino Mornasco, Italy) to match the oral environment conditions for a period of 14 days.
The impressions were cast using type IV dental stone (ISO 6873, Type 4; Die Stone Premium, South Korea). The gypsum master cast scanning was performed (SMART (Big), Open Technologies Optical 3D Scanner, Italy) and the type and material of the restoration were chosen by the Exocad DentalCAD software (Exocad DentalCAD, Exocad GmbH, Germany). The thickness of the cement was assigned at 20 µm, along with a distance of 1 mm from the margins of the preparation. The wax patterns of the onlays were designed in a fully anatomical form and milled using a wax molding disc (Wax Molding Disc, Shenzhen Vasden Bioceramics Co., Ltd., China) by an in-lab CAD/CAM system (Roland DWX-51D, Roland DGA Corporation, CA, USA). These patterns were invested in the phosphate-bonded investment (Maruvest Speed, Megadental GmbH, Germany). The ceramic onlays were fabricated by heat-pressed technique using lithium disilicate glass-ceramic in HO 1 color according to the manufacturer’s instructions (IPS e.max Press, Ivoclar Vivadent AG, Schaan, Liechtenstein), after that the margins were stained using khaki stain.
The ceramic onlays were placed into the preparations after removing the temporary restorations and cleaning the preparations, where their stability and the adaptation of their margins were tested using the probe. Their internal fit was checked using a condensation silicone wash (Speedex Light Body/Universal Activator, Coltène/Whaledent AG, Altstätten, Switzerland). The ceramic onlay was disinfected with ethyl alcohol. Then the internal surface was cleaned with phosphoric acid (H3PO4) (ETCH-37TM (37%), Bisco Inc., IL, USA) for 30 seconds. 9% hydrofluoric acid was applied for 20 seconds (Porcelain Etch, Ultradent Products Inc., UT, USA). It was put in an ultrasonic cleaning device (Clean-02; Ultrasonic cleaner, Runyes Medical Instrument Co., Ltd., Ningbo, China), and dried using a hot air dryer. After that, two layers of a silane solution (Silane, Ultradent Products Inc., UT, USA) were applied, and they were left for 60 seconds. The restoration was heated by applying warm air current for 30 seconds, then a coat of universal dental adhesive system (All-Bond 3®) was applied and gently air-dried without light-curing. The premolars were cleaned with a brush and a polishing paste containing fluoride-free pumice powder. For the abutment in an immediate dentin sealing group a sandblaster (CoJetTM Prep, 3M ESPE AG, Seefeld, Germany) with a silica-coated particle size of 30 µm (CoJetTM Sand; CoJetTM System, 3M ESPE AG) under a pressure of 30 psi from a distance of 10 mm and at a 45° angle was used for 3 seconds. Then phosphoric acid (UNI-ETCH® (32%)) was applied for 15 seconds to the preparation surfaces for etching the enamel margins and cleaning the conditioned sealed surfaces. A layer of silane solution (ESPETM Sil; CoJetTM System, 3M ESPE AG) was applied to the entire conditioned sealed surfaces and gently air-dried. Then, 2 coats of dual-cured dental adhesive system (All-Bond 3®) were applied to the entire preparation after which gently air-dried and left without light-curing.
As for the teeth in the delayed dentin sealing group, phosphoric acid (UNI-ETCH® (32%)) was applied to the entire preparation for 15 seconds with rinsing and drying, and a coat of dental adhesive system (All-Bond 3®) was applied, gently air-dried, and light-cured for 10 seconds.
The onlays were cemented with dual-cure conventional resin cement (Variolink N, Ivoclar Vivadent AG, Schaan, Liechtenstein). Sonic tips (SF1981/SF12; Komet Dental, Brasseler, Lemgo, Germany) mounted on a sonic hand-piece (AS2000 M4; NSK, Nakanishi Inc., Tochigi, Japan) were used to apply a pressure of 1.5 N for 4 seconds to ensure the gentle and precise positioning of the onlays. Light-curing was done for 5 seconds. Then, the margins were cleaned and the curing was carried out for 40 seconds from all sides. The margins were finished with rubber flame-shaped tips, cones and discs. Fig. 2 shows two ceramic onlays from the two groups after final cementation on the teeth.
All the restored teeth were subjected to thermocycling at 500 cycles in water baths between 5°C and 55°C and in 15 seconds at each temperature.
Fracture load measurement
The fracture load of the specimens was determined using a universal testing machine (Testometric M350-10KN, The Testometric Company Ltd., Rochdale, UK), where the compressive loads were applied in a perpendicular way to the center of the occlusal surface of the ceramic onlays with a 6 mm stainless steel round-ended rod that contacted the buccal and palatal triangular ridges and along the axis of the restored teeth at a 1 mm/min crosshead speed until fracture (Fig. 3). The maximum compressive load (N) required for causing fracture was recorded for each specimen. The failure types were examined under a stereomicroscope at 10x magnification and the failure mode was categorized as follows; Mode (I): fracture of the restoration only, Mode (II): fracture of the restoration with a small portion of the tooth, Mode (III): fracture of the restoration with more than half of the tooth, and Mode (IV): fracture involving the root.
The data was analyzed statistically using (IBM SPSS Statistics v.25 software, IBM Corp., Chicago, IL, USA). The Shapiro-Wilk test was used to test distribution of the data. The mean and standard deviation were calculated for each group. Then, an independent-samples t-test was run to determine if there were significant differences in fracture strength between the two groups. The selected level of statistical significance was p < 0.05.
Data obtained for the fracture load for each group were normally distributed, as assessed by Shapiro-Wilk test (p > 0.05). Tab. I shows the descriptive statistics of the fracture strength (N) in the two groups, where the highest mean fracture strength was recorded at 1785N.
The t-test revealed that there were statistically significant differences in the fracture strength between the two groups, where the mean fracture strength in the immediate dentin sealing group was (1335 ± 335N) higher than it when compared to the delayed dentin sealing group (931 ± 274N), at a significance level of 95% and p < 0.05 (Tab. II).
Tab. III shows the mode of failure within each group. The restoration fracture with a small portion of the tooth (Mode II) was the most common mode of failure in both groups, while root fracture (Mode IV) appeared in 30% of the cases in the DDS group and in 20% of the cases in the IDS group. Fig. 4 shows some failure modes that occurred.
To guarantee good bonding of indirect adhesive restorations to dentin, some researchers recommended the application of a dentin bonding system to the freshly cut dentin immediately after tooth preparation and before making the impression. This is called immediate dentin sealing (IDS). This technique provided many advantages, especially in relation to improving the bond strength (Magne et al. 2005; Okuda et al.2007; Magne 2014). Due to the fact that fracture is the most common failure mode (Morimoto et al. 2016), the aim of this study has emerged.
In the current bench-top study, the premolars were embedded in acrylic blocks to facilitate the various working procedures, especially when determining the fracture load of the specimens using later a universal testing machine. Owing to the elastic modulus of acrylic close to the alveolar bone (Scherrer & de Rijk 1993). In IDS group, the HEMA-Free, hydrophobic, filled bonding resin (All-Bond 3®) was applied to the entire sealed dentinal surface to obtain a homogeneous resin layer that has the ability to reduce microleakage, make the bonding layer less prone to water sorption, protect the hybrid layer during the temporization phase and prevent the possibility exposure of mineralized dentin during pre-cementation cleaning of the tooth preparation (Stavridakis, et al. 2005). A thin layer of flowable composite resin (Te-Econom Flow) was also applied to the entire IDS surface as a stress-absorbing layer with a low coefficient of elasticity to protect the dentin-resin adhesive interface from the fatigue caused by stresses, especially the overlaying resin cement shrinkage and occlusal forces (Baroudi & Rodrigues 2015). The enamel margins were re-prepared just before final impression to remove excess adhesive resin, to provide ideal taper for onlays, and to make direct bonding of the restoration to the enamel.
As for the temporary restoration, a water-soluble separating agent (PRO-V COAT®) was used to prohibit bonding of the provisional to the adhesive interface or tooth structure, protect the dentin surface from contamination, and clean up effortlessly without any residue (Magne 2005).
The wax patterns were milled by computer-aided design and computer-aided manufacturing (CAD-CAM) technique, where the immediate dentin sealing and its effectiveness were verified with CAD/CAM restorations. It was found that there is an improvement of the bond strength of ceramic and composite onlay restorations fabricated with this technique (Ishii et al. 2017; Murata et al. 2018).
In this study, the mean fracture strength in DDS group and IDS group were 931 N and 1335 N, respectively. The force required to fracture ceramic onlays was higher than the value of human masticatory forces. The average masticatory forces do not exceed 270 N (Naeije & Loon 1998), while the fracture strength in this study ranged between the lowest value (420 N) and the highest value (1785 N). Thus, the average human masticatory force was under this range. According to Bakke (2006), the value of bite forces in the premolar’s region ranges between 200-300 N. In any case, there is no consensus in the dental literature on the average values of masticatory forces, bearing in mind that the force is clinically distributed over more than one dental component, which reduces the load on a single tooth.
The strength of glass-ceramic restorative materials depends on the bond strength of resin cement to the internal surface of the restoration and to the various dental tissues of enamel and dentin (Peumans et al. 2000). The mean fracture strength in this study was close to that mentioned in similar study (Spohr et al. 2013); where the mean fracture load of the IPS empress 2 ceramic crowns in IDS and DDS group was 1300 ± 230 N and 1001 ± 186 N respectively. Also Yazigi et al. (2017) concluded that the mean fracture load values of thin CAD/CAM-fabricated occlusal glass-ceramic veneers were 1275 N without immediate dentin sealing and 1804 N with immediate dentin sealing.
On the other hand, the mean fracture strength of molars restored with lithium disilicate overlays was 2011 N in IDS group and 1837 N in delayed sealing (Hofsteenge et al. 2020). The difference might be attributed to the difference nature of the samples, which are bulkier and the lack of marginal ridge and specific morphology of the premolars.
The IDS group showed better fracture strength than of DDS group, as the resin materials were cured before impression, leading to a formed thickness that had been a positive effect on the fracture strength values in the IDS group. The resin materials were applied to seal the dentin, smoothen the internal preparation surfaces and create rounded internal line angles. These are required because of the shapes of the burs used to mill the wax molding disc or ceramic block and to achieve excellent cavity wall adaptation of the pressed and milled ceramic restorations (Hopp & Land 2013). Following the IDS technique appears to protect the dentin-resin adhesive interface during pre-cementation cleaning of the tooth preparation after removing of temporary restoration and when using the intraoral microblasting device before final cementation. Furthermore, the dentin-resin adhesive interface is protected from stresses caused by the polymerization shrinkage of resin cement during final cementation (Stavridakis et al. 2005). Furthermore, the use of a low-modulus flowable composite resin material acts as a stress-relieving layer.
The reason for the higher fracture load in IDS group could also be attributed to the fact that the IDS with the adhesive system, filled resin and low-viscosity microfilmed resin has significantly improved the bond strength of the ceramic restorations bonded to dentin using the resin cement (Okuda et al. 2007; Sultana et al. 2007).
Yazigi et al. (2017) proved that a higher fracture strength of thin CAD/CAM-fabricated occlusal lithium disilicate glass-ceramic veneers can be obtained when using immediate dentin sealing, regardless of the bonding technique used, whether it was total or selective etching protocol. El-Damanhoury & Gaintantzopoulou (2016) showed that the immediate dentin sealing was not improved the fracture strength of CAD/CAM-fabricated endocrowns when using the conventional feldspathic ceramic (Vita Mark II, Vita Zahnfabrik, Bad Säckingen, Germany), where immediate dentin sealing had a negative influence on the flexure resistance of the restored teeth, with a significant decrease in the obtained values. The reason for that was explained by the presence of optical powder residues.
The failure mode in this study varied from the fracture of the restoration only to the fracture of the root. Root fracture can occur when the loading is strong enough to exceed the borne limits of the tooth (Stappert et al. 2005; Schmidt et al. 2011), especially with the roots that are thinner. Root fracture was seen in 20% of the cases in the IDS group, while when not applying the immediate dentin sealing, the fracture of the roots reached to 30%. The force required to fracture the restoration is much higher than the common masticatory forces in oral cavity. Therefore, the destructive fracture mode is not seen in clinical studies of onlays (Beier et al. 2012). Most of failure modes in the two groups in this research were represented by the fracture of the restoration with a small part of the tooth, despite the bonding failure mode of the restoration with the tooth was not seen that can be explained by the high bonding strengths of a glass-ceramic to the dental tissues. In these cases, the adhesion was done in a dry, contamination-free medium. Consequently, a cohesive failure happened in the restoration or in the tooth itself. This is consistent with many previous studies (van den Breemer et al. 2017; Hofsteenge et al. 2020). However, in other previous studies, there were other failure modes, such as the adhesive failure mode between the restoration and the tooth, especially in laminate veneers (Gresnigt et al. 2016).
Some limitations of this study are represented by the fact that it was performed in a dry medium and under static conditions. Consequently, it was not completely similar to wet clinical conditions. In addition to that, the intraoral forces vary in quantity, application, speed and direction. However, the forces in this experiment were increasingly applied until fracture occurred. Therefore, this may not be the same as for oral conditions. However, the aim of the fracture strength studies is to determine the structural integrity of the restored tooth and restoration. Therefore, these types of studies are only considered as a tool for making a preliminary comparison in order to assess the relative effectiveness of examined technique.
Within the limitations of this in-vitro study, the following can be concluded:
The onlays in both techniques were strong enough to withstand physiological masticatory forces.
Dentin sealing affected the fracture strength of ceramic onlays; the onlays bonded using immediate dentin sealing technique showed higher fracture strength than those bonded using delayed dentin sealing.
Most of the failure modes in both groups were restoration fracture with a small portion of the tooth.
Bearing in mind the limitations of this study, it can be recommended to use the immediate dentin sealing technique in clinics by applying the dentin bonding system and low viscosity resin materials to the freshly cut dentin immediately after tooth preparation and before making the final impression.
The content of the manuscript has been entirely edited by Prof. Loai Aljerf (ORCID ID: 0000-0002-1132-9659; firstname.lastname@example.org)
Conflict of interest
The authors declare that there are no conflicts of interest related to this study.